SOCIETy, Journal o Research

Volume 3 August 1994

ISSN 1070-9428

CONTENTS

119 ANTROPOV, A. V. A review of the agile group of Pison (Hymenoptera: Sphecidae: Trypoxylini) Kieffer from and Ecuador AZEVEDO, C. O. Descriptions of two new species and notes on the Bakeriella 145 (Hymenoptera, Bethylidae)

... 5 BELSHAW, R. and B. BOLTON. A survey of the leaf litter fauna in Gahana, West Africa (Hymenoptera: Formicidae)

in the BITONDI, M. M. G., Z. L. P. SIMOES, A. M. do NASCIMENTO and S. L. GARCIA. Variation haemolymph protein restoration of titre hormone composition of confined Apis mellifera and potential vitellogenin by juvenile analogue treatment 107

BOHART, R. M. A review of North American Belomicrus (Hymenoptera, Sphecidae, Crabroninae) 207

for nest habitats of some social in Mato Grosso State, DINIZ, I. R. and K. KITAYAMA. Colony densities and preferences 133 Brazil (Hymenoptera, )

FIELD, S. A. and M. A. KELLER. Localization of the female sex pheromone gland in Cotesia rubecula Marshall (Hymenoptera: Braconidae) 151 Ichneumonidae: 157 GUPTA, V. K. A review of the world species of Orthomiscus Mason (Hymenoptera: Tryphoninae)

B. in HERATY, ]. M., ]. WOOLLEY and D. C. DARLING. Phylogenetic implications of the mesofurca and mesopostnotum Hymenoptera 241

KERR, W. E and E. B. SOARES REZENDE. Genetic characters of African bees that have high adaptive value in the tropics . 1 17 KUGLER, C. Revision of the ant genus Rogeria with descriptions of the sting apparatus (Hymenoptera: Formicidae)

L. of of the tribe Ashmead 175 NIEVES-ALDREY, J. Revision West-European genera Aylacini (Hymenoptera, Cynipidae)

a of OVRUSKI, S. M. Immature stages of Aganaspis pelleranoi (Brethes) (Hymenoptera: Cynipoidea: Eucoilidae), parasitoid 233 Ceratitis capitata (Wied.) and Anastreplm spp. (Diptera: Tephritidae)

POLASZEK, A. and K. V. KROMBEIN. The genera of Bethylinae (Hymenoptera: Bethylidae) 91

L. nov. a larval of stem-borer QUICKE, D. J. Myosomatoides gen. (Hymenoptera: Braconidae), Neotropical parasitoid pests, 227 Diatraea (Lepidoptera: Pyralidae)

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 The Ac gene has been observed by Brazilian bee biologists since 1957. In 1969 the information There is a constant search in tropical research published that it is male sex limited: it confers for characters that give to the bearer species a bronze color to the abdomen, especially to the higher adaptive value when compared to Euro- tergites, while workers are yellow (Kerr 1969). It pean ones. This happened with corn, wheat, cab- was present in 100% of the 145 queens collected in bage, apples, cattle, goats, etc. Many morphologi- Africa and brought to Brazil in 1956. There was cal, behavioral and physiological characters are some information (Prof.V. Portugal Araujo, p.c.) being studied in European and African bees, and that yellow males were occasionally present in in their hybrids under different ecological condi- Angolan populations. However, they were never tions. Among them, two of the African characters seen by W. E. Kerr in his 1956 trip to Africa. It is an are particularly interesting because within a few allele of b(=black) (Woyke and Kerr 1989), segre- generations they became predominant in a tropi- gates lAc: lac in different genetic background, cal environment, but not in a temperate one. These and is not linked to five components of agressive characters are the gene Ac and the African-mt- behavior (Stort 1978). DNA (that acts as a single gene). The population of this area is being well studied as far as their MATERIAL AND METHODS enzymes are concerned and show that the Africanized honeybee is an admixture of Apis Between 19 July 1990 and 17 September 1991, mellifera mellifera (19.5%), Apis mellifera ligustica 14 swarms of Africanized bees (Apis mellifera Linne) (3.8%) and Apis mellifera scutellata (76.7%) (Lobo et entered emptv stingless bee hive boxes located in al. 1989; Del Lama et al. 1990). the Apiary of the Universidade Federal de The African-mt-DNA has been studied by Uberlandia, MG, Brazil. Hive boxes varied in size, Hall and Muralidharan (1989), Sheppard et al. from 15 to 45 litres in volume, located 420 km from (1991), Sheppard et al. (1991), Soares (1992) and Camaqua, the point of introduction of African others. It shows indications of high adaptive value bees, in 1957. All swarms contained many males. in the tropics and low in southern A sample of males was taken from each swarm (that has European-like climate), where it is domi- and the numbers expressing the Ac and ac alleles nated by European mt-DNA. were determined. Journal of Hymenoptera Research

Males from an additional 17 colonies of (Kerr and Bueno 1970). Africanized bees were sampled from commercial In 1956, there were 400 Italian hives located in honeybee hives that were occupied by Africanized the same Eucalyptus forest into which 26 swarms swarms. Frequencies of the ac and Ac genes were of Africanbees escaped. All Italian colonies had ac calculated and compared with estimates of the and all African colonies had Ac genes. Therefore, frequencies of the same alleles in Camaqua in assuming equal contributions of all escaped colo- 1957. nies to the breeding population, a frequency of 0.939 for the gene ac and 0.061 for Ac in 1957 was RESULTS obtained for that original population. According to Winston (1992 pg. 40-42) an individual Six hundred and ninety (690) Ac drones Africanized colony swarms about 16 times per (0.9623) and 27 ac drones (0.0377) were sampled year. Of course, this can only happen in the expan- from the 31 colonies. No data for 1992 and 1993 sion phase of the population, before the popula- were collected because all swarms (22 and 9, re- tion approaches the carrying capacity of the envi- spectively) had only Ac drones. Drones in a swarm ronment. According to Nascimento (1981 page come from several colonies. 166), however, this figure, obtained in 1980, is 1.5 Some additional observations made are im- swarms per year. For our estimate a conservative portant to note: 1. Of the 14 swarms, only one had "intermediate" figure of two successful swarms two queens (the same proportion found in Kerr et per colony per year will be used. 1 " 57 al. 1970), 2. in three swarms many bees with wax The original frequency F of the gene ac scales were seen, which indicates that bees of (0.939) in 1957, multiplied, per generation, by its different ages were in the swarms, 3. in 1992 two relative adaptive value (w) will give us the fre-

1 *" 1 and in 1993 two swarms arrived without males. quency F for ac in 1991, that is 0.0377, 34 years 1991 = n The lack of ac drones in 1992 and 1993 swarms later, or after 68 generations. Therefore: F w . 11^ 7 is assumed to be a consequence of the population F . In this formula n, the number of generations, 1957 ,MM1 reaching fixation (100% Ac genes) or near fixation. is 68, F is 0.939, F 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 DISCUSSION the reason for the almost universal presence of Ac in feral populations of Apis meUifern in tropical Since the gene Ac segregates independent of South America. It may be a similar reason for the the xo gene, is independent of genes for defensive the high frequency of African-mt-DNA found in behavior and is an allele of black (b), the hypoth- Africanized populations, that is colonies of Euro- esis that it was linked and continues to be linked pean-mt-DNA queens x African-mt-DNA drones with genes for high fitness after 68 generations of should be less fit than colonies of African-mt- meiosis was discarded. DNA queens x European-mt-DNA drones. Several traits of Africanized bees have been The bronze color of the Ac drones may not be and are being selected for Brazilian conditions the cause of its fitness, since the workers are yel- (Page and Kerr 1991; Kerr 1992). The bees are low and heat preservation by dark color would be becoming less aggressive; they are being selected better in temperate climate; the physiological rea- for greater resistance to the Varroa jacobsoni mite son are being studied. (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 ACKNOWLEDGMENTS selection, since the degree of natural crosses be- We thank the FAPEMIG (Stateof Minas Gerais Research tween Africanized x Italian and Italian x Foundation) and CNPq (Brazilian National Research Coun- Africanized are about when and for financial equal queens cil) help and Dr. Robert E. Page, Jr. forcorrecting drones of both races use the same mating ground our English and for sound suggestions. Volume 3, 1994

LITERATURE CITED

E. S. N. Del e no desenvolvimento de colmeias africanizadas. Del Lama, M. A., J. A. Lobo, A. E. Soares, Lama, 1990. Genetic differentiation estimated by isozymic M.Sc. Thesis presented to the University of Sao Paulo at Bra- Ribeirao Brazil. analysis of Africanized honeybee populations from Preto,

W. E. Kerr. 1991 . bee and zil and from Central America. Apidologie 21: 271-280. Page, R. E, Jr., and Honey genetics 157 Hall, H. G. and K Muralidharan. 1989. Evidence from the breeding. 8th article of The "African" Honey Bee, pp - 186. Ed. David C. Fletcher and Michael mitochondrial DNA that African honey bees spread as Maria Spivak, J. D. Breed. Studies in Westview continuous maternal lineages. Nature 339: 211-213. Westview Biology. Kerr, W. E. 1969. Genetica e melhoramento de abelhas. In: Press, Boulder, USA. E. and H. Shimanuki Melhoramento e Genetica. Organized by W.E.Kerr, in Sheppard, W. S., A. E. Soares, D. De Jong of bee near the homage to Prof. F.G. Bneger. EDUSP, Melhoramentos, 1991. Hybrid status honey populations of afncanization in Brazil. 22: USP, Sao Paulo. Cap. XIV pg. 263-295. historic origin Apidologia Kerr, W. E. 1992. Abejas africanas: su introduccion y expan- 643-652. T.E. A. A. Steiner and sion en el continente americano. Subespecies y ecotipos Sheppard, W.S., Rinderer, J. Mazolli, J. 1991. flow between African and africanos. Industria Apicola N" 13: 12-21. H. Shimanuki. Gene Kerr, W. E. and D. Bueno. 1970. Natural crossing between European derived honey bee population in . Evolu- Nature 349: 782-784. Apis mellifera adansonii and Apis mellifera ligustica. molecular e tion 24(1): 145-148. Soares, A.E.E. 1992. A utilizacao da Genetica da de abelhas Kerr, W. E., L. S. Goncalves, L. F. Blotta and H. B. Maciel. 1970. morfometna na caractenzacao de populacoes africanizadas. Naturalia 14 a 18 de Biologia comparada entre abelhas italianas (Apis mellifera (Edic;ao Especial, e suas setembro de 117-125. ligustica), africana (Apis mellifera adansonii) 1992) pg. of of two hibridas. Anais do V Congresso Brasileiro de Apicultura Stort, A. C. 1978. Genetic study the aggressiveness of in Brazil. VII. Correlation of (Flonanopolis, SC) pg. 151-185. subspecies Apis mellifera A. Mestnner. 1989. the various characters each other Lobo, J. A., M. A. Del Lama, and M. aggressiveness among for abdominal color. Ciencia e Cultura Population differentiation and racial admisture in the and with the genes 492-496. Africanized honeybee (Apis mellifera L.). Evolution 43(4): 30(4): 794-802. Winston, Mark L. 1992. Killer Bee - The Africanized Honey Bee

. En- Moretto, G., L. S. Goncalves and D. De Jong. 1991 Africanized in the Americas. Harvard University Press, London, bees are more efficient at removing Varroa jacobsoni. gland

I. test between a sex Preliminary data. American Bee Journal. 131: 434. Woyke, and W.E. Kerr. 1989. Linkage and sex alleles in the bee. Nascimento, A. F., Jr. 1981. Estudo da mfluencia de fatores limited color gene honey ambientais no comportamento enxameatono, migratono 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° 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

Natural Cromwell London SW7 U.K. Biodiversity Division, Department of Entomology, Historv Museum, Road, 5BD,

Abstract. —Leaf litter samples were taken from 34 sites scattered across the moist tropical forest zone in Ghana. They were extracted Winkler included areas of primary forest, secondary forest and cocoa. Over 40,000 individual using bags a of their distribution. A total of 176 and identified. The species found are listed together with their abundance and summary of arboreal and almost two-thirds of which were species was found (excluding stray workers surface-foraging species), that in other forests The . The composition of the fauna is discussed and compared with found tropical species or with distance. composition at the different sites showed little variation either between the different forest types geographic

INTRODUCTION Voucher specimens of all taxa are deposited in the Natural History Museum, London. In the West African forest belt there has been little quantitative sampling of the ant fauna; stud- METHODS ies have been carried out in the Tai Forest Reserve, — Cote d'lvoire (see Levieux 1982 and included ref- Sites. The locations of the sampling sites in erences) and the Reserve de Campo, Cameroun Ghana are shown in Figure 1, with brief descrip- (Halle and Pacal 1992: 65-109). In Ghana there has tions and sampling dates given in Table 1. Sites letter but with different been no quantitative sampling except in the main designated by the same in detail are within of each tree-crop, cocoa. This has been studied numbers (e.g. jl and j2) 3km the other. With one the sites are within the (e.g. Majer 1976 and included references) but exception, extent to which it resembles the original forest moist semi-deciduous forest zone of Hall and fauna is not known. Nevertheless, this research Swaine (1976). has led to the of West African ants being We sampled in a wide range of the forest more advanced than that of most tropical ant habitats found in Ghana, including 14 areas of forest the broad sense of forest with a faunas (e.g. Bolton 1987 and included references). primary (in In consequence we can survey elements of this closed high canopy), 10 areas of secondary forest fauna with the hope of accurately identifying much (of varying age, mostly on agricultural land) and of it to species. 10 cocoa farms. Sampling was carried out between Ghana has two main terrestrial biomes, sa- December 1991 and November 1992. vannah and forest, and these have distinct ant — faunas. In turn, the forest zone is readily divisible Sampling. At each site an area of approxi- : into a canopy and a ground fauna. In this paper we mately 1000m was measured out. Within this area 2 survey the leaf litter element of the forest ground ten lm quadrats were placed at random. All the fauna by sampling at different localities across leaf litter inside a quadrat was collected, shaken Ghana. In addition to identifying the species through a 1cm sieve, and then left for three days in present, we discuss the composition of the fauna a Winkler bag. The extracted ants were combined and compare it with that from other tropical for- to form a single total for each site, each site being ests. We also examine how the species composi- sampled on only one occasion. All sampling was tion at the sites varies geographically and between done between 9.30 a.m. and 3.00 p.m. soil the different forest types. At three sites (h, ql and cj2) an additional Journal of Hymenoptera Research

with dates and habitat was taken from each This was Table 1. Sampling sites description. sample quadrat. in Sacred groves are small pieces of forest left agricultural the soil from a 25cm 25cm done collecting by that Ghanaian name by areas for religious reasons. (Note place 2 to a of (= 0.0625m ) depth approximately quadrat spellings are often variable.) 5cm. This soil was then sieved and left for three in the same manner as the days in Winkler bags a Sui River Forest Reserve, 1.x. 1992, primary forest, overlying leaf litter. b Mabang, 18.xii.1991, secondary forest, forest, The Winkler bag (Besuchet el al. 1987) oper- c Tinte Bepo Forest Reserve, 31.iii.1992, primary d Forest Reserve, 11. in. 1992, primary forest, ates in a similar fashion to a Berlese Funnel except Mankrang e Poano, 9.1x1992, cocoa, that the material is left hanging in a mesh bag to f near Ofinso, 2.X1.1992, cocoa. in air rather than to a heat source. for- dry exposed g Jachie, 20. iv. 1992, sacred grove (28 acres), primary and easier to use Winkler bags are much cheaper est, Funnels. Litter-sift- h Effiduase, 17.xi.1992, cocoa. and to transport than Berlese 1 - Forest Reserve forest); 2 in Winkler records i Bobiri. 6.1V.1992, (primary ing followed by extraction bags 8. iv. 1992, Forest Reserve forest but all mature in (primary which do not turn pitfall traps - many species up trees killed with sodium arsenide in 1947); 3 13. iv. 1992, Our extraction of three in (Olson 1991). period days secondary forest (farmland left in 1982 and burnt was chosen on the basis of a trial extraction, with 1983). a two week Juaso. 1 - 21. ix. 1992, secondary forest (area of Dome daily sorting of a sample over period. I River Forest Reserve burnt in 1983); 2 - 23.ix.1992, sec- We found that within the first three 86% of days - ondary forest (farmland left for c.20 years); 3 8.ix.l992, the individuals and 88% of the had species cocoa emerged. k Southern Scarp Forest Reserve (North-West of Mpraeso and in We ignored winged reproductives wing- near Osubeng), 23.x. 1992, secondary forest (burnt less queens found without workers, except in spe- 1983). - Forest 1 Kade. 1 6.x. 1992, forest (in cies where the queen is known to forage during primary Aiyeola Reserve); 2 - 12.x. 1992, secondary forest (farmland left nest foundation. in 1957); 3- 12.x. 1992, cocoa, m Esukawkaw Forest Reserve, 27.x. 1992, primary forest, — assess the sec- Analyses. In order to complete- n Nkawanda (near Nkawkaw), 12.xii.1991, roadside ness of our survey for the region sampled, i.e. the ondary forest, - near moist semi-deciduous zone disturbed habi- o Atewa Forest Reserve, primary forest. 1 2.iii.l992, plus - Kibi; 2 - 24.ih.1992, near Potrase; 3 26.h.l992, near tats we a accumulation within, plotted species - Sagymasi (logged in 1970's); 4 27. m. 1992, nearSagymasi curve. We first the sites in five random arranged (logged in 1970's). In each we calculated the sequences. sequence p Asiakwa, 1. v. 1992, cocoa. the first 1 Research Institute arboretum number of species found at the first site, q Bunso. -6.xi.1992, Crops forest; 2 - 17.iv.1992, for- two sites combined, the first three sites combined, (15 acres), primary secondary est forest cleared ca 20 years previ- on. the mean of the five (primary partially and so Finally, sequences - ously, left undisturbed for ca 12 years); 3 secondary was calculated. The extent to which the resulting forest (cocoa left in 1981);4 -24.ii.1992, cocoa;5-6.iii. 1992, curve flattens out indicates the proportion of the cocoa, actual fauna which has been recorded; a failure to r Old Tafo sacred grove (ca 3 acres), 31. i. 1992, primary forest. flatten out indicates that additional species would - s New Tafo (Cocoa Research Institute of Ghana). 1 have been found if the sampling had been contin- 23.xii.1991, secondary forest (farmland left for en 40-50 1991 and tests meth- ued. Palmer (1990, ) compares - years); 2 ll.xii.1991, cocoa, for the richness of a ods estimating species region t Nankasi, 17.ix.1992, cocoa. from samples taken within it. He concludes that the first-order jacknife is the most precise method, of a i.e. the one whose estimates are closest to the true Body lengths (= the outstretched length value, and we therefore also apply this analysis to point-mounted worker including mandibles) were our data. taken from the literature or from an average of five of worker We converted the body lengths each species specimens. In species with a dimorphic to biomass (= dry weight) using the following caste we did not count major and minor workers of 30:1 for equation, taken from Gowing and Recher (1984). separately. Instead we used an estimate the ratio of minor to major workers in all cases. = -4.0 + assess the effect of the distance between Log n weight(mg) 2.5(log n length(mm)) To Volume 3. 1 994

1 1 1 IVORY r -8° y s Journal of Hymenoptera Research

in our sur- comprised 9% of the species found, they com- clearly not been adequately sampled it is a much more prised only 2% of the total number of individuals. vey, and probably important In the leaf litter the average density of indi- component of African forests than our results 2 . 1990: viduals (excluding tourists) was 117m Using indicate (see Holldobler and Wilson 588). Berlese funnels, other studies have found similar We found one worker of Apomyrma stygia in leaf litter a densities of ants in leaf litter: in tropical moist Brown, Gotwald and Levieux damp forest on Barro Colorado Island, Panama, in the mile inside the Esukawkaw Forest Reserve. This is : wet season it was ca 200m (Levings 1983); in the sole described species in the Apomyrminae one local- temperate deciduous forest in Maryland, U.S.A. and was recorded previously only from in Cote D'lvoire etal. where four between May and September the mean monthly ity (Brown 1970), 2 in soil forest and density was 194m (Lynch and Johnson 1988). nests were found under gallery one under adjacent unburnt savannah. The spe- — cies is subterranean workers are blind Composition of the fauna. The relative impor- clearly (the ), tance of the different subfamilies in the combined and the remains of a geophilomorph centipede 2. in one of the Cote D'lvoire soil and leaf litter samples is shown in Fig. were found nests, The fauna is dominated by Myrmicinae. Where raising the possiblity that the species specialises their biology is known, the species we found are on this prey item. all for most generalist predators / scavengers except — Dacetonini, the workers of which forage singly for Habitat and geograpliic variation. In a separate Collembola and other soft-bodied (us- paper we have investigated in detail the differ- ant the inghighly specialised mandibles), and Decamorium ences in the assemblages between primary decern (Forel), which has been observed attacking forest, secondary forest and cocoa sites in this that termites (Bolton pers. obs.). Paedalgus distinctus study (Belshaw and Bolton 1993). We found Bolton and Belshaw has also been found in asso- they did not differ significantly either in species ciation with termites (Bolton and Belshaw 1993, composition or in species richness. Only two com- see below). mon species show an association with a particular The is the second most important forest type (see Table 2): Serrastruma hijae (Forel) subfamily in our survey. Many ponerines are large was with the exception of a single individual ants, so although the subfamily comprises only 8% found only in the Esukawkaw and Atewa Forest of the total number of individuals, it comprises Reserves, and one Oligomyrmex species (sp.indet.4) 27% of the total biomass. A good example is mostly occurred in cocoa. Given the large number Paltothyreus tarsatus (Fabricius): only 25 individu- of species present, one would predict that, even if als of this species were found during the survey, all species were randomly distributed among the but its contribution to the total ant biomass was three forest types, a handful of such apparent exceeded by only two other species. This subfam- associations would occur purely from chance. of in 4 is ily contains a high proportion of genera which are The slope the regression line Fig. very specialised predators. In our survey we recorded shallow, and the (logarithmically-transformed) Ambh/opone (preys on geophilomorph centipedes), distance between sites only accounts for 1.9% of Plectroctena macgeei Bolton (other members of the the variation in similarity. The distance between genus prey on millipedes), Paltothi/reus tarsatus sites therefore had at the most only a slight effect (preys on termites), and Discothyrea (preys on on species composition. eggs) (all sources in Holldobler and — Wilson 1990: 559). Comparison with other faunas. Several other Driver ants (Aenictinae and ) have studies of tropical forest ant faunas contain lists of - highly aggregated distributions their colonies leaf litter species: dry forest at Kimberley, N. Aus- are nomadic and very large (with between 60,000 tralia (Andersen and Majer 1991); moist forest on and 20,000,000 workers in other species (Holldobler Barro Colorado Island, Panama (Levings 1983); and Wilson 1990: 581)). They are represented in and dry forest in Madagascar (Olson and Ward, in our survey by single workers found at three sites press). Following elimination of obvious arboreal and by almost 1500 workers from one soil quadrat species, these studies all show the Myrmicinae to that hit part of a nest. This group has be the largest subfamily, followed by the Ponerinae Volume 3, 1994

[1 Biomass

Individuals 100 D Species

80

o 60

c 40 o

20

nl ^1 nL, ii11 J I L1 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 doryline and aenictine ants from the island has number of species respectively). They comprised possibly permitted diversification in this genus of 52% and 30% at Kimberley (33 species in total), specialised predators of other ants (Wilson 1971: and Island 65% 27% on Barro Colorado (93 species 68). Unfortunately, it is not possible to compare in total), and 55% and 20% in Madagascar (44 the species richness of the four areas owing to the species in total). The figures in the last study differences in sampling area, effort and method would have been but for the higher unusually (Berlese funnel, Winkler bag and pitfall trapping). large number of Cerapachys species. The absence of 10 Journal of Hymenoptera Research

200

o

o, 150

Cm O

,0 a 100

M

50 e u

10 20 30 40

Number of sites

of Fig. 3. Species accumulation curve for the leaf Utter samples. Line fitted by DWLS smoothing option computer program SYSTAT (Wilkinson 1990). — Efficiency of survey. The species accumula- (Wilson 1959; Bolton pers. obs.). In addition, some tion curve is shown in Fig. 3. If the fitted line is other species nest in the top l-2cm of the soil but extrapolated beyond the data it does not continue forage in the leaf litter. However, there are other to rise. Estimating the true regional species rich- smaller elements of the ground fauna which, be- ness using the first-order jacknife method, we find cause of their nesting or foraging habits, are likely our total of 176 species represents 81% of the actual to have been missed by our sampling method. species present. We infer from this that, within the 1) Completely subterranean species (= ones area of Ghana sampled, we recorded a large ma- which nest and forage only in the soil). In Ghana jority of the species foraging within leaf litter. only a few such species are known, e.g. Plectroctena In tropical forests the majority of ground ant anops Bolton and P. hastifera (Santschi) (the work- species nest in the leaf litter, either in small pieces ers of which are either blind {anops) or with very of rotting wood or between compressed leaves small eyes {hastifera)). However, this microhabitat Volume 3. 1994 11

1.0

X

s 75 12 Journal of Hymenoptera Research

In cases is of of Leptogem/s previously recorded from Ghana sole Baracidris. most nothing known of the not found in our (Bolton 1975), plus one undescribed species. The the ecology species survey genus Camponotus is poorly known taxonomically, except for earlier collection localities and dates. but there are more than 15 species from Ghana in However, their absence from our survey shows the collection of the Natural History Museum that they either do not forage in the leaf litter or are which we did not find. Non-arboreal members of rare. this genus, although nesting in the soil, are fast- soil —The of individuals moving surface foragers, which may also have The samples. density contributed to their poor representation in our was eight times higher in the soil samples than in = survey. the leaf litter (mean of the three soil samples 982 2 also found The three largest myrmicine tribes are suffi- m (S.E.= 62). Lynch and Johnson (1988) forest the of ants ciently well known taxonomically for us to com- that in a temperate density was in the soil than the leaf litter. data pare the list of species recorded in our survey with higher These as in both studies that of species already known from Ghana. In the may, however, be misleading Dacetonini, 25 species were previously known the soil samples probably included species which, (Bolton 1983), including two known to be arbo- although nesting in the top few cm of the soil, real. We found 18 of the remaining 23 species, plus forage primarily in the leaf litter. 14 previously unrecorded in Ghana. In the Although similar in overall species composi- Solenopsidrni, 15 species were known (Bolton tion to the leaf litter samples (also found by Lynch con- 1987), including three savannah species. We found and Johnson 1988), the three soil samples 10 of the remaining 12 species, plus six previously tained a small distinctive subterranean element. unrecorded . These results show that we have found We extracted 30 workers of Leptanilla boltoni Baroni many small cryptic species which had previously Urbani from a single soil quadrat in a small patch been overlooked. The world-wide household pest of primary forest at Bunso. The sole Afrotropical Monomorium pharaonis (L.) (Pharaoh's Ant) was representative of the genus, it is known previously recorded from nine sites (including primary for- only from 6 specimens extracted by Berlese funnel est). Its range is clearly not restricted to disturbed from leaf litter in a cocoa plantation at Mampong, habitats in Ghana. Ghana (Baroni Urbani 1977). The genus contains a In contrast, we found a smaller proportion of total of 33 described species, all of which appear to the Tetramoriini species known from Ghana. Pre- be very rare (Holldobler and Wilson 1990: 590). viously 53 species were known (Bolton 1976, 1980), The biology of only one species (L. japonica Baroni including 10 savannah and three arboreal species. Urbani) is known: it is strictly subterranean and We found only 21 of the remaining 40 species, plus appears to have true legionary behaviour, em- six previously unrecorded in Ghana. Members of ploying both group (of geophilomorph this tribe, however, often nest in exposed sites or in centipedes) and colony migration (Masuko, 1990). rotten logs, which may explain its relatively poor Of the 17 individuals of Acropyga sp., 10 were representation in our survey. found in soil samples. This species resembles oth- We cannot compare our results for the ers in the genus which tend coccids in subterra- Pheidologetonini, Pheidolini or Crematogastrini nean nests for their sugary secretions (Holldobler as no synthesising taxonomic studies have been and Wilson 1990: 527). It is the only species found undertaken. The remaining myrmicine tribes con- in our survey which is known not to be primarily tain a total of 40 species previously recorded from predacious, although the habit probably occurs in Ghana (Bolton 1981a, 1981b, 1982) including 20 species of several other formicine genera arboreal (chiefly Cataidacus), four savannah, and (Pseudolasius, Paratrechiiia and Lepisiota). One of two species which tunnel in living wood the species of the undescribed Ponerini genus also (Melissotarsus). We found 8 of the remaining 14 appeared to be soil-dwelling. species: one of the three Cardiocondyla, one of the The results of our few soil samples point three Leptothorax, two of the four Calyptomyrmex, tantalisingly to a rich and poorly known subterra- both Pristomyrmex, the sole Meranoplus and the nean fauna that would repay further collecting. Volume 3, 1994 13

1 for and 1 for their Table 2. Species recorded during the survey (see table details of sites, figure location). "Widespread" sites number is in denotes the species was recorded from more than five differently-lettered (the given parentheses). An asterisk (*) denotes species classed as tourists (see text).

Species 14 Journal of Hymenoptera Research

Species Volume 3, 1994 15

Species L6 Journal of Hymenoptera Research

Bolton, B. 1976. The ant tribe Tetramornni (Hymenoptera: Holldobler, B. and E.O. Wilson 1990 The ants Springer- Formicidae): constituent genera, review of smaller gen- Verlag, Berlin. 732 pp. 1982. of the era and revision of Triglyphothrix Forel. Bulletin of the Levieux, J. A comparison ground dwelling ant British Museum (Natural History) (Entomology) 34: 283- populations between a guinea savanna and an ever- 379. green rain forest of the Ivory Coast, pp. 48-53. In Breed, Bolton, B. 1980. The ant tribe Tetramornni (Hymenoptera: M.D., CD. Michener, and HE. Evans, eds. The biology of Formicidae): the genus Tetramorium Mayr in the Ethio- social . Westview Press, Boulder, Colorado. 419 pian zoogeographical region. Bulletin of the British Mu- PP- seum (Natural History) (Entomology) 40: 193-384 Levings, S. C. 1983. Seasonal, annual, and among-site varia- Bolton, B. 1981a. A revision of the ant genera Meranoplus F. tion in the ground ant community of a deciduous tropi- Smith, Dieroaspts Emery and Calyptomyrmex Emery (Hy- cal forest: some causes of patchy species distributions. menoptera: Formicidae) in the Ethiopian zoogeographi- Ecological Monographs 53: 435-455. Bulletin the British His- F. and A. K. 1988. and cal region. of Museum (Natural Lynch, J. Johnson. Spatial temporal tory) (Entomology) 42: 43-81. variation in the abundance and diversity of ants (Hy- Bolton, B. 1981b. A revision of six minor genera of My rmicmae menoptera: Formicidae) in the soil and litter layers of a (Hymenoptera: Formicidae) in the Ethiopian region. Maryland forest. The American Midland Naturalist 119: Bulletin of the British Museum (Natural History) (Entomol- 31-44. D. 1976. influence of ant ant ogy) 43: 245-307. Majer, J. The and manipulation Bolton, B. 1982. Afrotropical species of the myrmicine ant on the cocoa farm fauna. Journal of Applied Ecology 13: genera Cardiocondyla, Leptothorax, Melissotarsus, Messor 157-175. and Cataulaeus (Formicidae). Bulletin of the British Mu- Masuko, K. 1990, Behavior and ecology of the enigmatic ant seum (Natural History) (Entomology) 45: 307-370. Leptamlla japonica Baroni Urbani. hisectes Socwux 37: 31- Bolton, B. 1983. The Afrotropical dacetine ants (Formicidae) 57. Bulletin of the British Museum (Natural History) (Entomol- Olson, D. M. 1991. A comparison of the efficacy of litter ogy) 46: 267-416. sifting and pitfall traps tor sampling leaf litter ants Bolton, B. 1987. A review of the Solenopsis genus-group and (Hvmenoptera, Formicidae) in a tropical wet forest, revision of Afrotropical Monornorium Mayr (Hy- Costa Rica. Biotropica 23: 166-172. menoptera: Formicidae). Bulletin of the British Museum Olson, D. M. and P. S. Ward, in press. The ant fauna (Hy- (Natural History) (Entomology) 54: 263-452. menoptera: Formicidae) of Kinndy Forest (tropical dry B. and R. Belshaw. 1993. and of the In Bolton, Taxonomy biology forest) in Western Madagascar. Ganzhorn, J., ed. The lestobiotic ant supposedly genus Paedalgus (Hym.: ecology and economy ofa tropical dryforest. Springer Verlag, Formicidae). Systematic Entomology 18: 181-189. Berlin W. W. H. Gotwald and Levieux. 1970. W. 1990. Brown, L., Jr. J. A new Palmer, M. The estimation of species richness by genus of ponerine ants from West Africa (Hvmenoptera: extrapolation. Ecology 71: 1195-1198. Formicidae) with ecological notes. Psyche 77: 259-275. Palmer, M. W. 1991. Estimating species richness: the second Gowing, G. and H. F. Recher. 1984. Length-weight relation- order jacknife reconsidered. Ecology 72: 1512-1513. ships for invertebrates from forests in south-eastern Wilkinson, L. 1990. SYSTAT: the systemfor statistics- Evanston, New South Wales. Australian Journal of Ecology 9: 5-8 IL: SYSTAT, Inc. B. and M. D. Swaine. 1976. Classification E. Hall, J. and ecologv Wilson, O. 1959. Some ecological characteristics of ants in of closed-canopy forest in Ghana. Journal of Ecology h4 New Guinea rain forests. Ecology 40: 437-447. 913-951. Wilson, E. O. 1971. The insect societies Belknap Press, Cam- B. D. Hall, J. and M. Swaine. 1981. Distribution and ecology bridge, Massachusetts.

of vascular in a rain . plants tropical forest: forest vegeta- Wolda, H. 1981 Similarity indices, sample size and diversity. tion in Ghana. Geobotany 1: 1-383. Oecologia 50: 296-302. Halle, H and O. Pacal, eds. 1992. Biologie d'une canopee deforet equatoriale: volume 2. Unpublished report. 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. (, , n. n. ), carinata n. sp. (Caribbean), ciliosa n. sp. (, Ecuador), cornuta sp. (Belize, southern Mexico), cuneola sp. n. n. (Mexico, Central America), gibba n. sp. (Colombia, Ecuador), innotabilis sp. (Mexico, Central America), leptonana 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 belli blanda Fr. Smith bruclu n. sp. (Venezuela). Previously recognized species are Mann 1922, 1858, 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 subdivision was attractive, and for a brief period Irogera even enjoyed generic status (Brown This revision will redefine the genus, establish 1953), but with a better understanding of the varia- new synonymies, create informal species groups, tion in petiole shape within species, the splitting of redescribe a to the these lines became untenable species, provide key workers, Rogeria along ( Kempf and summarize ecological information. Keys to 1965). Meanwhile, Melanesian and Polynesian the reproducti ves are not included, because queens species were being added to the genus with no are known or proposed for less than half of the discussion of the disjunct distribution of the genus Rogeria species and males for only four species. or comparisons of New World and Old World The was erected 1 894 Santschi genus Rogeria by Emery ( ) species (Mann 1919, 1921, 1925; 1922, for a few ants then in and 194 1 that of these neotropical Myrmica ). Brown (1953) recognized many Tetramorium. He then extended the distribution of were actually Lordomyrma species, but still, the to the genus New Guinea with the addition of Pacific species including stigmatica, epinotalis, and stigmatica (Emery 1897). Discovery of other later, exsulans remained in Rogeria. Wilson and neotropical species led Emery (1915) to create the Taylor (1967) called this "...one of the most anoma- subgenus Irogera for those with a clavate petiole. lous discontinuous distributions found in ants." 18 Journal of Hymenoptera Research

lirata are sister The disjunct distribution of Rogeria was ques- that germaini and species. The un- tioned by Kusnezov (1958) on the basis of palpal usual shape of the spiracular plate ultimately con- as a formula differences between stigmatica and a South vinced me to propose innotabilis new species deliberation over the welter of variation American species, and by Kugler (1978b) on the after long basis of the sting apparatus. In contrast to all other within the creightoni-gioup. with myrmicine genera, there were two distinctly dif- This character system must be used caution is to reduction ferent sting apparatuses within the genus, and however. The apparatus prone and that difference coincided with the disjunct distri- convergence of form. The sting of Lordomyrma is like those of the bution of the genus. The sting of inermis from epinotalis more Rogeria stigmatica- Central America was strong, acute and with an group than those of Lordomyrma. In this case, I unusual low dorsal flange; the sting of stigmatica based my decision to transfer epinotalis to from the Pacific had a weak, narrowly spatulate Lordomyrma on external characters, since the com- well to sting shaft and no dorsal flange (among other mon sting features could be due reduction differences). I believed a revision of Rogeria using convergence. not been used in sting apparatus characters would separate the Hair has previously Rogeria Pacific and New World species into distinct gen- systematics, but in this revision I came to rely more era. and more on characteristics of pilosity for distin- The potential of the sting apparatus for defin- guishing species. Consequently it has become nec- ing ant genera and inferring phylogeny has been essary for species descriptions to identify pre- demonstrated by a number of comparative mor- cisely the several kinds of hairs and their distribu- phological studies (Kugler 1978b, 1980, 1986, 1992). tions on the ants, sometimes including the num- In addition, Bolton (1973, 1982, 1987) has often bers of pairs of erect hairs on the head, mesosoma, used externally visible parts of the sting to help and nodes. This means, of course, that care must define some myrmicine taxa. But this is the first be taken to not disturb the hair while cleaning and time that characters from the whole apparatus mounting the ants. have been used as an integral part of a taxonomic I have attempted to take a fairly conservative revision. This work, then, also tests the usefulness approach to synonymy and description of new of the sting apparatus in the alpha-taxonomy of species, choosing in ambiguous situations to err ants. on the side of not changing the number of species I examined sting apparatuses of 79 workers in the genus. Thus, for lack of sufficient evidence, and queens in 27 Rogeria species and seven some species may actually contain several sibling Lordomyrma species. In the five species in which species: alzatei, belli, creightoni, foreli, leptonana, and both queens and workers were dissected, there scandens. On the other hand, I have retained some were only the usual individual differences, such as names that may fall when collections improve: numbers of sensilla. Contrary to expectations, the brnchi,micromma. Treatment of the creightoni-group Western Samoan species exsulans is unmistakably has perhaps been somewhat less conservative (See related to Central American Rogeria species in the Species Group section). sting apparatus, pygidium, and external anatomy. All told, this study recognizes 39 species (in- Moreover, some South American species (ciliosa, cluding 19 new species), establishes nine new gibba, besucheti) had sting apparatuses and pygidia synonymies and transfers seven species to with the same distinctive features of stigmatica. It Lordomyrma. It fails to resurrect the subgenus the distribution is seems of Rogeria really disjunct. Irogera. It retains three Polynesian /Melanesian characters Sting and pygidial provided sup- species (stigmatica, megastigmatica, exsulans) with port for numerous other taxonomic decisions as the Neotropical bulk of the genus. well. For example, they provided strong support As presently constituted, members of the ge- for creating the stigmatica-group of species, and nus Rogeria are distributed from Buenos Aires to could conceivably be used to distinguish that group southern Texas and Arizona, and in the Pacific as a separate genus. Sting apparatus anatomy also between ION and 25°S from Tahiti to the western supported Brown's (1953) belief that levifrons, end of the island of New Guinea. So far it is striatella, and tortuosa are Lordomyrma species. unknown in Australia or southeastern United Autapomorphic shape of the sting shaft confirmed 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 South America Pacific America Northern Southern

alzatai 20 Journal of Hymenoptera Research more detail. In the North American region, most 55-60°C lactophenol for 5 minutes (or longer if species occur below the Isthmus of Tehuantepec, necessary), rinsing twice in 70% ethanol, and rins- but cuneola extend north- twice in 95% ethanol. After two species (creightoni, ) ing clearing, sting ward through the eastern lowlands of Mexico. apparatuses and genitalia were usually cut in half, Only foreli and creightoni have been collected in the except for the aedeagus and sting, which were United States. The North American region con- separated from the other sclerites. Stings, pygidia, tains 8-10 endemic species (depending on the un- and hypopvgia were mounted in glycerin jelly for certain distributions of innotabilis and leptonana); ease of repositioning. Mouthparts, genitalia, and the Caribbean, two endemic species; South America other sting apparatus sclerites were soaked in (including Trinidad), 19 endemic species and the xylene, then mounted in Canada balsam. Occa- Pacific, three endemic species. Only 5-7 species sionally, whole ants were similarly cleared and (depending on innotabilis and leptonana) are found mounted in balsam. Preparations were examined in both Central America and northern South with a Zeiss KF-2 phase contrast microscope. America. The "Materials Examined" sections of each

Little is known of the biology of these cryptic species description identifies which specimens ants. Collection records usually range from sea were dissected. Following the locality of the dis- level to 1000m, but five species extend higher and sected ant, brackets identify the structures slide- two (ungiiispina and merenbergiana) can be found mounted ("sting" means sting apparatus, py- at 2000m. Rogeria species are generally collected in gidium and hypopygium). Dissected specimens moist forests (primary or secondary forests, coffee are workers unless otherwise identified. or cacao plantations), but at higher elevations can Pinned vouchers are identified with the label be found in pastures {leptonana, merenbergiana). "Kugler 1991 Dissection Voucher." All are depos- Several species (creightoni, cuneola, foreli) have been ited in the collection of the MCZ, except for vouch- found in both moist and dry climates. Rogeria foreli ers of lacertosa, pellecta, subarmata, the Agudos, is the most unusual, with some members dwelling Brazil voucher of alzatei, and the Surinam vouch- at over 1800m in the temperate mountains of south- ers of curvipubens, which are all in the MZSP. ern Arizona. Most Rogeria species have only been collected Illustrations as strays or by Berlese or Winkler sampling, usu- ally in leaf litter and rotten wood, but occasionally Drawings were made using a grid eyepieces among epiphytes and moss (belti, creightoni, and grid-backed tracing paper. Drawings of sting exsulans). Nests of several species (belti, blanda, apparatus preparations were made at 400X with merenbergiana) have been found under loose bark estimated accuracy of <0. 001mm. Scanning elec- of rotten logs. Nests of blanda and tondnzi have tron micrographs were taken on an AMR-900 and been taken from the trunks of cacao trees. A nest of a JEOL JSM-35C. leptonana was found at 1750m under a rock in a Dorsal views of heads are in the same full-face pasture. view used to measure HL. Lateral views of Because nests are so rarely found, males are mesosomas are in the same view used to measure known for only four species (belti, blanda, leptonana WL, SpL, and MHI. and stigmatica), and queens associated through Drawings attempt to show all hairs on the left nest series for only nine species (See Synonymic side of the body that project well above the body List of Species). outline. Care was taken to present typical pilosity, rather than matted or damaged hairs. Hair is not METHODS AND TERMINOLOGY included in line drawings of: 1) dorsal views of heads, 2) ventral portions of head profiles, 3) most Dissections dorsal views of waists, and 3) queens and males. All line drawings of sting apparatuses are Mouthparts, sting apparatus, pygidium, from workers. and hypopygium, male genitalia preparations were Male genitalia drawings show the lateral view obtained ants in by rehydrating 70% ethanol, dis- of the aedeagus and medial view of the right these structures from the in secting ants, clearing paramere and volsella (with associated part of the Volume 3, 1994 21

PetW Petiole width: Maximum transversedistanceacross phallobase). the node. Nontype specimens are identified in figure PpetL Postpetiole length: In lateral view, the axial dis- legends by affixing their collection localities in tance from base of node in front to tip of posterior parentheses. peduncle. PpetW Postpetiole width: Maximum transverse distance Measurements and Indices across the postpetiole. PSI Petiolar spine index: SpL/WL. PW Pronotal width: Maximum width across pronotum Measurements and of whole ants, drawings in dorsal view. for the were made a except very largest, using SI Scape index: SL/HW. Zeiss SR stereomicroscope at 125X magnification SL Scape length: Maximum length excluding basal and neck. and a fiber optic ring lamp. Estimated accuracy: ± condyle In same view as for 0.01mm. SpL Propodeal spine length (Fig. 1): WL, from tip of propodeal spine to nearest edge of All specimens were measured for Weber's propodeal spiracle pentreme. then at least the and smallest Length (WL) largest MH Mesosoma height (Fig 1): In perfect lateral view, from each locality were measured completely. The the vertical distance from a line tangent to the maximum and minimum of each measurement ventralmost points of the of meso- and metasterna to the on the mesosoma a and index were double-checked, as were all mea- highest point profile (for broadlv arching pronotum, estimated as the mid- surements of holotypes. The number of specimens point of the arc). (N) that follows the list of metric characters at the MHI Mesosoma height index (Fig. 1): Mesosoma height beginning of a species description is the number of (MH) divided by the horizontal distance from the base of the line to the corner of ants that were measured completely. The numbers MH posteroventral at the ends of "Material Examined" sections are the propodeum. TL Total length: Sum of ML + HL + WL + PetL + total studied. PpetL specimens + GL to the nearest 0.1mm. In metric characters for new listing species, WL Weber's length (Fig. 1): In perfect lateral view of the ranges found in all type material are followed the mesosoma, diagonally from posteroventral corner of mesosoma to farthest anterior by the holotype measurements in parentheses. point on face of the neck. If no queens or males are described for a pronotum, excluding species, none were available for study. Additional Terminology

CL Cephalic index: HW/HL as EL Eye length: Maximum diameter of compound eye Terms such vertex, occiput, malar and genal viewed straight on. areas are of dubious accuracy when applied to FLW Frontal lobe width: In full face dorsal view, maxi- ants. Consequently, there is disagreement over mum transverse distance across frontal lobes. how to name the regions of the head (Trager 1989; GL Gaster length: In lateral view, from anterior edge of 1989; Wheeler 1989). Moreover, these first tergum to posteriormost point usually T2 or Snelling T3, but sometimes to end of pvgidium) terms are confusing jargon that make the use of GW Gasterwidth: Maximum transversedistanceacross keys and descriptions difficult for the uninitiated. gaster. Instead, I prefer to rely, as much as possible, on the HL Head length: Maximum longitudinal distance from generally understood directional terms, anterior, the antenormost portion of the clypeus (usually lateral, etc., that the mandibles are ante- the clypeal apron, but sometimes the more promi- assuming nent body of the clypeus) to the midpoint of a line rior. Five regions of the head usually contain dis- across the back of the head. tinctive sculpture in Rogeria (Fig. 1). 1) The HW Head width: Maximum width in same view as for middorsum of the head is the median portion of HL, excluding eves. For males, HW includes com- the dorsal surface between the maximum retrac- pound eves. tion of the and from the frontal ML Mandible length: In same view as for HL, from scapes extending antenormost portion of head to apex of closed area to a change in sculpture near the rear of the mandibles. head. 2) The posterior (region of the) head in- OI Ocular index: EL/HW. cludes the actual posterior surface of the head and PetL Petiole length: In lateral view and with petiole not the posterodorsum, the dorsal surface between strongly flexed up or down, the axial distance from the middorsum and the outline of the the dorsal corner of the posterior peduncle to the posterior nearest edge of the metapleural lobe. head. The "posterior head" seems equivalent to 22 Journal of Hymenoptera Research the "vertex" of some authors who consider the seen from dorsal view are subrectangular, which mandibles to be ventral and the back of the head to have evenly convex sides that are widest midlength and which have be the top, rather than posterior. 3) The laterodorsa (e.g. Figs. 53, 58), subtrapezoidal, sides that caudad are the portions of the head seen in dorsal view convex diverge (e.g. Figs. 49, 50, that lie beneath the sweeps of the scapes. 4) The 51). venter is the ventral surface of the head not seen in The terga and sterna of the gaster are referred lateral view. 5) The sides of the head lie between to by abbreviations Tl, SI, etc. Terminal seg- the venter and the laterodorsa (between the lateral ments of gaster rotated ventrad means that T2 outline of the head in dorsal view and the ventral and T3 are distinctly longer then their correspond- outline of the head in lateral view). ing sterna, causing the pygidium and hypopygium In triangular mandibles the basal angle is (T4 and S4) to shift anteroventrad and making T2 or T3 the distalmost sclerite of the distinct and nearly a right angle (e.g. Figs. 43, 49, gaster. Figs. 28, 63); in subtriangular mandibles the basal angle is 40, 43, and 63 show rotated terminal segments; and 8 unrotated reduced and more obtuse (e.g. Figs. 14, 58, 61). Figs. 2, 5, show segments. Sometimes tiny denticles are intercalated between Pygidial gland sculpture refers to a pair of these are not microareolate on the anterior of the or replace teeth (e.g. Figs. 35, 55); patches edge included in tooth counts. The palpal formula is pygidium. The presence of this characteristic sculp- the number of maxillary palpomeres followed by ture strongly suggests the presence of pygidial number of labial palpomeres. Palpal formula is glands, but its absence does not necessarily mean usually determined by dissection, thus the sample the glands are absent (Holldobler, et al. 1976; size is very small and may not show the variation Kugler 1978a; Holldobler and Engel 1978). In most within the species. The body of the clypeus is the species I was able to see the gland reservoirs medial portion of the clypeus anterior to the fron- attached to these patches, but failure to see the tal lobes and dorsal to the clypeal apron, which is reservoirs does not mean they are absent; they the anterior rim just above the mandibles (Fig. 1). could easily be lost during dissection. Clypeal apron shape is described as seen from Sting apparatus terms are identified in Fig- dorsal view. In describing eye shape, elliptical or ures 3 and 42. ellipsoid means similarly rounded at both ends, Sculpture descriptions follow illustrations and whereas oval or ovoid means the anterior end is definitions of Harris (1979). It seems to me that the more narrowly rounded. The shape of the poste- term "areolate" in Harris' nomenclature best de- rior outline of the head is as seen in the full dorsal scribes much of the sculpture in Rogeria, but I view used for measuring HL since shape some- suspect many would prefer the term "reticulate." times varies with the pitch of the head. The nuchal I have called the ridges on the head and mesosoma grooves are longitudinal impressions in the of many species "rugae," because they are irregu- posteroventral corners of the head, into which the lar, and wrinkled to varying degrees, but others lower corners of the pronotum fit when the head is may prefer to call them "carinae." Also following retracted. Harris, in compound descriptors, such as rugose- A rough index of the degree of inclination of areolate, the predominant type of sculpture is the propodeal spines is expressed by imagining a first. bisecting line that divides the lateral aspect of the The following are used to describe approxi- in spine half and extends forward over the mate hair inclination as seen in side view (Fig. 1): it intersects the erect mesosoma side, and noting where (> 70 ), suberect (70-50), subdecumbent of the or It is 5 edge mesopleuron pronotum (Fig. 1). (50-25), decumbent (25-5), appressed (< ) (in- important that the mesosoma be in perfect lateral ference from Wilson 1955). To arrive at the num- view. The diameter of a propodeal spiracle in- bers of erect hairs on the mesosoma dorsum, I cludes the peritreme and is as seen in full lateral used both dark and light backgrounds and a vari- view of the mesosoma. The spiracle's position is ety of positions to count all erect hairs on one half measured from the outer edge of the peritreme to of the dorsum from the anterior edge of the pronotal the nearest edge of the propodeum, which is the disc to the bases of the propodeal spines. Since the free edge of the infradental lamella. hairs are distributed symmetrically, that count most of The common shapes postpetioles as gives the number of pairs of erect hairs. Frequently, Volume 1994 3. 23

both sides were counted in order to check for the Macromedia (in part); Mann 1922:30. Type: Macromischa presence of broken hairs. Erect hair on head dor- purpurata Roger 1863, by designation of Wheeler, 1911. Brown 1953:4. sum means on frontal lobes, middorsum, and Irogera; Irogera; Kempf 1962a:435. posterior head, but not under the of the [Redescnbed]. sweep 1963:195. to Rogeria; Kempf [Key South American species], scapes. Erect hair on scapes means on the dorsal Irogera; Kempf 1964:66. [Key to species] or anterior surfaces. Usually no attempt is made to Rogeria; Kempf 1965:185. describe hair on ventral lateral Rogeria; 1972:227. of and aspects of Kempf [Catalog New World species]. body, nor on other than extensor surfaces of femora Worker — and tibiae. Diagnosis. Monomorphic myrmi- cines. Antenna 12-segmented; scape not reaching DEPOSITORIES posterior margin of head; distinct 3-segmented club than longer rest of funiculus; apical antennomere than combined of ANIC Australian National Insect Collection, CSIRO, longer lengths other two club Canberra, Australia. segments. No antennal scrobes or fossae. BMNH British Clypeus projects between frontal Museum (Natural History), London, U.K. narrowly lobes at least CKC Charles to posterior of antennal inser- Kugler Collection, Radford University, edge tions. of with Radford, VA, U.S.A. Body clypeus one or more pair of CUIC Cornell longitudinal carinulae. Lateral not raised University Insect Collection, Ithaca, NY, clypeus into a U.S.A. ridge in front of antennal insertions. Nuchal on DMOC David M. grooves present posteroventral corners of head . Olson Collection, University of Califor- Anteroventral corners of nia, Davis, CA, U.S.A. pronotum angular to dentate FML Fundacion and fitting into nuchal Miguel Lillo, Institute de Zoologia, grooves. Propodeal spiracle 3 diameters or less from the of the Miguel Lillo 251, Tucuman, Argentina edge below the JTLC John T. Longino Collection propodeum propodeal spines. LACM Los Metapleural lobes not Angeles County Museum, Los Angeles, CA, sharply pointed. Worker U.S.A. Description.—Mandibles usually trian- gular. in some ciliosa and MCSN Museo Civico di Storia Naturale "Giacomo Doha," Except foreli, mandibles with 5-7 teeth (3 teeth in size Genoa, Italy. apical decreasing MCZ Museum of basad, followed by 2-3 smaller basal Addi- Comparative Zoology, Cambridge, MA, teeth). tional U.S.A. denticles may occur among basal teeth or any basal tooth be a MHN Museum d'Histoire Naturelle, Geneva, Switzer- may replaced by pair of denticles. land. Palpal formula usually 3,2 or 2,2, but 3,3 in MIZA Museo some and 2,1 in the Instituto de Zoologia Agricola, Universidad stigmatica-group, very tiny minima. neither elbowed nor at the Central de Venezuela, Maracay, Venezuela. Scape ridged MZSP base, nor with an around the Museu de Zoologia da Universidade de Sao Paulo, apron peduncle. in Sao Paulo, Brazil. Clypeus profile usually with a very narrow anterior of rises near NMB Naturhistorisches Museum, Basel, Switzerland. apron. Body clypeus verti- in most PSWC S. cally species, but Philip Ward Collection, University of California, occasionally projecting the Davis, CA, U.S.A. beyond clypeal apron. Frontal lobes narrow, but USNM United States covering antennal insertions; at most National Museum, Washington, DC, feebly notched behind. U.S.A. Frontal triangle small, depressed slightly. Eyes with 1-100 facets; located on sides, in the anterior half of the head Genus ROGERIA Emery (excluding mandibles). Sides of head widest just behind the eyes, forming rounded corners with the Myrmica (in part); Fr. Smith 1858:131. Type: Formica rubra posterior head, which is Linneaus, 1758, designation of Wheeler, 1911 weakly concave to weakly convex in full face Tetramorium (in 1886:359. part); Mayr Type: Formica caespitum view. Linneaus, 1758. Mesosoma broad shoul- 1894:188. generally compact, Rogeria Emery Type: Rogeria curvipubens Emery, dered. Anterior face of rises 1894, by designation of Wheeler, 1911. pronotum nearly ver- Rogeria 1915:191. tically from the neck and forms a subgenus Irogera Emery Type: Rogeria procera usually distinct, Emery, by original designation. rounded angle with the dorsal surface. Mesosoma 24 Journal of Hymenoptera Research

mesonotal and antennae. Terminal of dorsum without sutures; no groove; procera ) segments gaster metanotal groove absent to distinct. Anterior edge with erect hair. No erect hair on laterodorsa of of propodeum often marked by a transverse ca- head (under sweep of scapes). Body of clypeus just rina. Propodeal spines absent to long. Metapleural above apron with pairs of erect hairs; members of lobes low carinae to rounded triangular. Legs not stigmatica group each with an additional median incrassate. No tibial spurs on middle or hind legs. hair. Tarsal claws simple. Color from brownish yellow to black, with man- Petiolar peduncle with or without a ventral dibles, antennae, and legs lighter. Most species keel; inferior process dentate except in stigmatica also with a lighter triangle on clypeus, cheeks, and group. Node unarmed; poorly to well differenti- frontal area. ated from peduncle. Postpetiole with short pe- Malpighian tubules in the one species exam- duncles and a low node that is broader than long. ined (belti) 5 in the worker (N=2); 5 to 6 in the male Terminal segments of gaster rotated ventrad in all (N=3) (Brown— 1988). but the stigmatica group. Wings. Notation (Fig. 30) follows Brown and All pygidia dissected have a pair of small Nutting (1949) and Holldobler and Wilson (1990:9). pygidial gland reservoirs and/or paired First and second radial crossveins of primitive microareolate patches present on anterior edge, formicid wings are lost, creating a single large except in the stigmatica group. Common features submarginal cell. The distal portion of the radial of the sting apparatus are: 1) medial connection of sector vein (Rsf4+5) vein is straight or curves spiracular plate incompletely sclerotized, 2) forward, but never reaches the edge of the wing. gonostylus single-segmented, 3) dorsoterminal The r-m crossvein present (Fig. 30) or absent (Fig. chaeta present, 4) at least one companion seta 37). Median and cubital veins of variable length, (except gibba), 5) each lancet with a single moder- but not reaching wing margins. Medio-cubital ate to large valve, 6) sting bulb large, with arched crossvein always present in females, but may be sting base. lost in males.— Mandibles usually predominantly smooth, Larvae. Similarities among the larvae of belti, with piligerous punctures and vestigial carinulae scandens, procera and blanda found by Wheeler and at insertions, but carinulae stronger and more Wheeler (1973, 1976, 1986, 1988) are summarized extensive in some members of the stigmatica-group. here. Form pheidoloid, i.e. "abdomen short, stout In all but ciliosa, the body of the clypeus is smooth and straight; head ventral near anterior end, with a pair of carinulae extending from the frontal mounted on short stout neck, which is the protho- lobes and stopping short of the clypeal apron; rax; ends rounded, one end more so than the these are sometimes accompanied laterally by 1-2 other" (Wheeler and Wheeler 1976:8). Thorax and shorter, weaker carinulae. Lateral extremities of abdomen with at least bifid and anchor-tipped clypeus and adjacent cheeks with longitudinal hairs; bifid hair on head (sometimes others). Cra- carinulae. Frontal triangle smooth, except in ciliosa. nium subtrapezoidal; clypeus bulging; antennae If macrosculpture present, frontal lobes and with 3 sensilla; labrum bilobed and narrowed middorsum with diverging longitudinal rugae; dorsally. Mandibles ectatommoid (stout, gradu- rest of body areolate, rugose, or occasionally cari- ally tapered and curved, with an apical tooth and nate. Microsculpture when present usually additional teeth and denticles in some species). microareolate, appearing granular or punctured Maxilla shape paraboloid or lobose, with cylindri- at lower magnifications. Posterior face of cal galea and palps of varying length. Labium propodeum smooth, except in gibba. Legs smooth narrow; palps only slight elevations. Maxillae and and shiny. First segment of gaster smooth and labial surfaces coarsely spinose. shiny; less so mprominula, and minima. Pygidium and sometimes other terminal terga with micro- COMPARISONS WITH SIMILAR GENERA scopic areolate sculpture on exposed posterior Some members of with surfaces; pygidium may also possess minute Leptotlwrax (placed in the have antennae like piligerous tubercles. Rogeria Leptothoracini) Rogeria and some have a narrow lobe of Short appressed or decumbent pilosity com- posterior the clypeus, but these have rounded anteroventral mon, most dependably on legs (except blanda and Volume 3, 1994 25

corners of the pronotum and no nuchal grooves. Synonymic List of Species Of 65 species of Leptothorax examined at the MCZ, only one had an angular inferior corner of the Castes that have been collected are recorded = pronotum, but in that species the scapes extend in brackets in the right column. W worker, Q = = beyond the head, the posterior lobe of the clypeus queen, M male. Upper case Q and M are used if is wider, and nuchal grooves are absent. at least some reproductives are associated through Some species of Lordomymm (Myrmecinini), a a nest series. Lower case q and m are used if no genus sometimes confused with Rogeria in pacific reproductives are linked to workers by a nest have similar antennal fea- series. is islands, and clypeal An asterisk (*) added if reproductives tures, but have a rounded anteroventral corner of were not even found in the same localities as the pronotum and lack nuchal grooves. The stings workers. of Rogeria species are dramatically different from alzatei n. those of five Lordomyrma species I have examined sp. [W, q] belti Mann, 1922 [W, Q, (caledonica Ern Andre, levifrons Mann, punctiventris Mj besucheti n. sp. [W] Wheeler, stnatella Mann, tortuosa Mann) in shape blanda Fr. Smith, 1858 [W, Q, M] of quadrate plate, oblong plate, gonostylus, trian- =foveata Kempf, 1964 (synonymy by Kempf 1965) gular plate and sting. Another species, L. epinotalis bruchi Santschi, 1922 [W] brunnea 1930 is very different from the other Lordomyrma I ex- Santschi, [W, q] =caraiba Santschi, 1936 n. amined and has a sting apparatus somewhat like syn. =cubensis Santschi, 1936 n. syn. that of Rogeria besucheti or R. ciliosa, but lacks the =habamca Santschi, 1936 n. syn. pronotal corner or nuchal of grooves Rogeria. Wings =scabra Weber, 1934 n. syn. of three I have been able to Lordomyrma species cannata n. sp. [W] examine (epinotalis, striatella, leae Wheeler) are dis- cilwsa n. sp. [W, Q] cornuta n. tinct from the known Rogeria wings in having the sp. [W] creightoni Snelling, 1973 [W, q*[ Rsf4+5 vein curve anteriorly to the wing margin. cuneola n. sp. [W, q] In L. caledonica this vein is like in to Rogeria failing curvipubens Emery, 1894 [W, Q] reach the the is margin, but m-cu vein absent. The exsulans Wilson and Taylor, 1967 [W] larvae of an unidentified Lordomyrma species dif- foreli Emery, 1894 [W, Q] Forel, 1914 n. fers from the known larvae of Rogeria as follows: =gaigei syn. form cranium =huachucana Snelling, 1973 n. syn. pogonomyrmecoid, subelliptical, 1894 germaini Emery, [W, q*] hairs fewer, none (Wheeler and anchor-tipped =minensis Santschi, 1923 n. syn. Wheeler 1977). n. gibba sp. [W, q] Stenamma (Pheidolini) workers are similar in inermis Mann, 1922 (W, Q] innotabilis n. form of clypeus, including narrow posterior por- sp. [W, q*] lacertosa 1963 tion Kempf, [W| between frontal lobes, and some have 3-seg- leptonana n. sp. [W, M) mented antennal but in that case the q", clubs, apical n. lirata sp. [W, q*| is shorter than the combined of the segment length megastigmatica n. sp. [W] other two segments. Also, Stenamma has no nuchal merenbergiaim n. sp. [W, Q] micromma grooves, the anteroventral corner of the pronotum Kempf, 1961 [W] minima Kusnezov, 1958 (Q] is rounded, and the metanotal is groove generally n. neilyensis sp. \\\ | more distinct than in Rogeria Larvae of species. nevadensis n. sp. [W] Stenamma differ from those of as follows: Rogeria pellecta Kempf, 1963 [wj form aphaenogastroid; cranium subhexagonal; procera Emery, 1896 [W] mandibles pogonomyrmecoid (similar to =brasiliensis Borgmeier, 1953 (Synonymy by Kempf 1962a) ectatommoid); body hairs bifid or denticulate, not pronnnula n. sp. [W] anchor-tipped (Wheeler and Wheeler 1953, 1976). 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 [W] terescandens n. sp. [W] stigmatica-group: besucheti, ciliosa, tonduzi Forel, 1899 [W] gibba, tribrocca n. sp. [W] megastigmatica, prominula, stigmatica. unguispma n. sp. [W] — Diagnosis. (prominula andmegastigmaticanot Names Transferred to Other Genera. dissected). WL 0.67-1. 30mm. Palpal formula 3,3 or 3,2. Eye with 4-41 facets. Metanotal groove distinct azumai Santschi 1941 (to Lordomyrma, Brown 1952) and extends down sides of mesosoma. Petiolar epinotalis Mann 1919 to Lordomyrma, n. comb. without rugosa Mann 1921 to Lordomyrma, n. comb. peduncle a keel and inferior process a tortuosa Mann 1921 to n. comb. small Lordomyrma, step or absent. Nodes and gaster wide. Gaster tortuosa Mann 1921 to Lordomyrma, n. comb. levifrons with terminal segments not rotated ventrad, or tortuosa pohta Mann 1921 to Lordomyrma, n. comb. only slightly rotated (besucheti). No tortuosa stonen Mann 1925 to Lordomyrma, n. comb. pygidial gland or evidence of reservoirs. striatella Mann 1921 to Lordomyrma, n. comb. sculpture any Sting ap- paratus: weakly sclerotized, spiracle occupying more than of SPECIES GROUPS 1/4 spiracular plate; no anal plate; valve chamber not arising abruptly from base of shaft; shaft and lancets Some species of Rogeria can be assembled into sting sting weak, sparulate for more or less distinct species groups. Others can (except perhaps prominula); sting shaft apex not be without furcula a placed easily in any group, or seem to link any flange; simple arch (no several dorsal arm). dense groups. These incertae sedis species are Predominantly areolate on described with the group to which thev may be macrosculpture head (except middorsum), most related. mesosoma (except some stigmatica), and nodes. Mandibles with piligerous punctures and stron- The more extensive carinulate stigmatica-group may be distinct enough ger, sculpture than in other to be treated as a separate genus, but for now Rogeria species. No microsculpture on most seems tied to of of with a the rest of Rogeria through blanda. body. Body clypeus median erect hair low near anterior In the creightom species group, extensive varia- margin. Similar incertae sedis tion and often limited and scattered collections species: blanda. This species made it difficult has and like ciliosa to ascertain which specimens sculpture pilosity and gibba, but has a were simple variants and which were distinct 2,2 palpal formula, weak metanotal groove, dentate inferior species. At one time or another, I considered rec- petiolar process, and gaster, py- between four and of other ognizing and 20 species in this group. gidium sting apparatus typical groups. Ultimately, I tried to steer a middle course be- tween these extremes; neither masking the varia- scandens-group: scandens, subarmata, terescandens tion by describing large conglomerate species, nor — encumbering the genus with a lot of dubious new Diagnosis. (terescandens not dissected). WL 0.78-1. 17mm. species names. As a result, I have been somewhat Palpal formula 3,2 Eyes large (30-80 more venturesome facets). faces about naming new species in Propodeal spiracle strongly caudad. this lobes group, yet there remain a number of rather Metapleural prominent, triangular. Petiolar node low, anterior face indistinct from heterogeneous species (alzatei, brunnea, creightom, peduncle I some leptonana). have described and illustrated some (except subarmata). Inferior petiolar process of the variants in dentate. sternum these heterogeneous species in Postpetiolar large. Pygidial gland order to alert other and reservoirs taxonomists to remaining spe- sculpture present, at least in cies and the subarmata. like that problems need for extensive collect- Sting apparatus of inermis (Fig. ing, especially in the Caribbean Islands and 42). Head, mesosoma, and nodes predominantly circumcaribbean countries. rugose; sides of head below eye smooth. Dorsa of In the head, mesosoma diagnoses below, if a species is listed as nodes, and gaster Tl with sparse, undissected, the palpal formula, pygidium, and stiff, untapered, erect hairs and sparse, incon- sting apparatus characters are unknown. 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. short, decumbent pilosity and 2) equally or less These both lack the stiff hairs and inflated abundant longer, erect-suberect hair (except for postpetiole that make the group so distinctive, but lack of erect hair on head of some scobinata.). otherwise are so similar that perhaps they should Usually more than 10 pairs on mesosoma dorsum be included. (9 in some leptonana, 8 in carinata) and 2 or more posterodorsally projecting hairs on each node. All germaini-group: germaini, lirata, lacertosa. hairs tapered. Similar incertae sedis species: inermis, belti, — 0.65-1. > Diagnosis. WL 05mm, usually cornuta, neilyensis, exsulans (cornnta and neilyensis 0.80mm. Mandibles triangular. Palpal formula 3,2. not dissected). These do not have two distinct Body of clypeus not projecting over apron. Eye types of hair on the mesosoma dorsum. small (6-20 facets; OI 0.09-0.14). Propodeal spines rather long (PSI 0.18-0.23); directed caudad. Peti- curvipiibens-group: cuneola, curvipubens ole with prominent keel and dentate inferior pro- cess. Petiolar node rises abruptly from anterior Diagnosis. —WL 0.50-0.63mm. Mandibles tri- peduncle and with more or less distinct anterior, angular. Palpal formula 2,2. Clypeal apron usu- dorsal and posterior surfaces. Posterior two- thirds ally convex (medially flattened in Haitian of postpetiolar node with parallel or slightly con- curvipubens). Eye small (4-11 facets). PSI 0.13-0.18. vergent sides. Postpetiolar sternum in side view Petiole with distinct node, weak keel, and dentate rather long, its anterior corner perpendicular. Py- inferior process. Postpetiole from above usually as gidial gland sculpture and reservoirs present. Sting in Fig. 74. Pygidial gland sculpture and reservoirs similar to that of but shaft apparatus inermis, sting present. Sting apparatus as in inermis (Fig. 42), and lancets lancet lacks weak; barbule. Laterodorsa, except for sting shaft and lancets of curvipubens. sides, and posterior head rugose-areolate (some- Median head and cheeks with weak longitudinal times effaced on sides). Mesosoma predominantly rugose-areolate macrosculpture. Mesosoma with rugose or vermiculate-rugose. Microsculpture predominantly rugose macrosculpture and ar- weak; intervals on head and mesosoma shiny and eolate microsculpture, both often weak. Body cov- smooth. nearly Mesosoma dorsum and gaster Tl ered with appressed-decumbent pilosity. Erect with abundant fine, tapered, erect-suberect hairs, hairs sparse: none on scapes or extensor surfaces but if or of very little, any, decumbent appressed legs, 0-16 on head (if present, short and limited pilosity. to posterior head), usually 2-7 pairs on mesosoma Similar incertae sedis species: pellecta, sicaria dorsum (rarely 8), 0-1 pair posterodorsally pro- not (sicaria dissected) These have different petiole jecting and 0-1 laterally projecting hairs on each shapes and more decumbent hair on the mesosoma. node. Anterior portions of gaster Tl often lack sicaria also Rogeria differs in clypeus, propodeal erect hair. All hairs tapered. and erect spines, pilosity. Similar incertae sedis species: micromma, minima (neither dissected). Rogeria micromma has a creightoni-group:alzatei,brunnea,carinata,creightoni, subrectangular postpetiole in dorsal view and innotabilis, leptonana, merenbergiana, nevadensis, abundant erect hair on head. Rogeria minima has scobinata, tribrocca, unguispina stiff, cuneate hairs and a 2,1 palpal formula. — Diagnosis. (brunnea, nevadensis, and tribrocca /ore//-group: bruchi,foreli not WL 0.51-0.93mm. trian- dissected). Mandibles — gular, except as noted. Palpal formula 3,2 or 2,2. Diagnosis. WL 0.50-0. 80mm, usually < small Eyes (6-19 facets). Petiole with a distinct 0.75mm. Mandibles triangular. Palpal formula 2,2. node and dentate inferior process. Pygidial gland Eye 5-20 facets. Propodeal spines wide at base. sculpture and reservoirs present. Sting apparatus Petiole with distinct node and dentate inferior like that of inermis unless otherwise noted. process. Pygidial gland sculpture and reservoirs Mesosoma predominantly rugose or carinate. present. Sting apparatus almost identical to that of Head and dorsum, mesosoma, nodes, gaster cov- inermis (Fig. 42). Sculpture predominantly ered with two distinct kinds of hairs: 1) abundant 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 2 may also be present ,

' 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. (, 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' with Mesosoma dorsum 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' < 0.90. If then EL/SpL usually greater (some belti), 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 small < 19(17') Eyes 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 with erect to suberect hair in 21(15') Scapes longer 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' If Propodeal spines usually much shorter. similar in size, then not strongly inclined. EL/SpL > 0.20 .. 24 > 0.71mm. 24(23') WL 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 with tooth. 25(24') Clypeal apron median 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 Part of Tl 28(26') gaster microareolate. Most erect hairs stiff, cuneate-fimbriate. (Argentina) (Figs. 72-73) minima 28' Gaster Tl smooth. Stiff hairs absent (except on some micromma) 29 on sides 29(28') Rugae pronotal dorsum and 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' on Rugae 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 with 2-5 facets. 30(29') Eye Mesosoma sides opaque with dense areolate microsculpture, but macrosculpture absent. Mesosoma with of dorsum 8-10 pairs erect hairs. (Surinam, Para State of Brazil) (Fig. 71) .... micromma 30' with 4-21 < Eye 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 convex. Nuchal visible in lateral view. wider. 32(31 ) Clypeal apron groove clearly Propodeal spines generally 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 alzatei Republic) (Figs. 58-60, 99) 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 < If belt then decumbent hair abundant on 33' WL usually 0.90mm. larger (some i, brunnea, creightoni), 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. > 0.65. EL/SpL (Colombian and Ecuadorian Andes above 1000m) (Figs. 46-48) meretibergiana 41' little or effect Metanotum with no 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 smooth intervals or largely areolate with a few carinulae between the spines. Most of petiole and Stigmatica-Group and Related Species postpetiole strongly areolate as well. Sculpture on anterior petiolar node effaced; dorsum of Rogeria ciliosa new species peduncle smooth. First tergum and sternum of gaster rather Figs. 2-4 coarsely punctured in front and more finely punc- Additions to stigmatica-group diagnosis. WL tured caudad; smooth and shiny between punc- 1 .06-1 .20mm. Mandibles triangular; coarsely punc- tures. tured, weakly carinulate. Anterior clypeus evenly Fine, long, flexuous, erect to suberect hair convex. Propodeal spines long (> 0.20mm), covers middorsum of head, dorsum of mesosoma, strongly inclined and diverging. Metapleural lobes dorsum and sides of waist, and first segment of well developed; angular. Inferior petiolar process gaster. On terminal segments of gaster these be- reduced to a small step. Dorsal head, mesosoma come denser, stif fer and more erect to form brush- and gaster densely covered with long flexuous like rings. Shorter, subdecumbent hairs occur on hairs; terminal segments of gaster with dense, stiff lateral and ventral surfaces of head, dorsal sur- erect hair. — faces of scapes, sides of mesosoma, and sometimes Holotype and Paratype Workers. TL 4. 1-4. 8 (4.1), dorsal face of propodeum. Very short, decumbent 0.90-1.02 HL (0.90), HW 0.78-0.89 (0.785), SL 0.69- to appressed pilosity on extensor surfaces of legs 0.77 0.14-0.16 (0.69), EL (0.15) (32-42 facets), PW and ventral petiole. Median hair on clypeus fine 0.60-0.66 (0.60), WL 1.06-1.20 (1.06), SpL 0.24-0.28 and often obscured by surrounding paired hairs. (0.245), PetL 0.44-0.53 (0.44), PpetL 0.23-0.29 Body uniformly rusty-brown; appendages (0.235)mm, CI 0.84-0.87 (0.87), OI 0.17-0.19 (0.19), lighter, more yellowish-brown. SI 0.85-0.90 (0.88), PSI 0.21-0.23 (0.23). N=9 Paratype Queen.— TL5.3,HL 1.04, HW 0.91, SL Mandibles with 7-9 teeth (3 apical teeth de- 0.80, EL0~27, PW 0.97, WL 1.50, SpL 0.30, PetL 0.58, creasing in size basad, followed by 4-6 small, PpetL 0.30mm, CI 0.88, SI 0.88, PSI 0.20. N=l subequal teeth). Palpal formula 3,2. No clypeal Differing from paratypes only in the normal of apron. Body clypeus not projecting over ante- queen characters (Fig. 4). Mandible8-toothed. Short rior edge. Nuchal grooves weak. Pronotal shoul- parapsidal furrows present on mesoscutum. Both ders rounded. Ventral petiole with a low median mesoscutum and mesoscutellum with same dense carina rather than a distinct keel. Node distinct, areolate sculpture as in worker. Metanotum wider than long. Postpetiole dorsal view shape as vaguely microareolate. Dorsal face of propodeum in Fig. 66. Postpetiolar sternum short, with a dis- with transverse carinulae mesad; areolate laterad. tinct anterior followed a — lip by narrow sulcus. Discussion. Rogeria gibba (Figs. 5-6) from Gaster large (GW/WL 0.94-0.97). Quadrate plate western Ecuador and Colombia resembles ciliosa of sting apparatus with somewhat reduced in size, sculpture and pilosity, but has different that lacks apodeme lobes on anterodorsal corner; mandibles, clypeal margin, propodeal spines, and oblong plate ventral arm very reduced (see also metapleural lobes. Rogeria stigmatica and 3 and Fig. stigmatica-group diagnosis). megastigmatica (Figs. 7-13) from the Pacific have of Body clypeus with fragmented longitudi- much shorter propodeal spines, less punctured nal surrounded effaced rugulae by areolate sculp- gaster, and different pilosity. ture. Longitudinal rugae on head dorsum mostly The name ciliosa refers to its dense covering of confined to the frontal lobes, frontal area, and flexuous hairs. midline. Frontal lobes — rugose-areolate in some Distribution. This species is known only from paratypes. Rest of dorsum, cheeks, venter, and lowland rain forests in the amazonian basin of posterior head densely areolate; intervals bearing Ecuador and the Orinocan basin of Venezuela. shallow — piligerous punctures. Promesonorum with Material Examined. Holotype locality. EC- the same areolate sculpture. Mesopleura, UADOR: Napo Province, Limoncocha, 250m, 18- and sides of metapleura, propodeum with more VI-1976, #B-348 (S. and J. Peck) [MCZ]. and confused areolate irregular sculpture. Met- Paratype locality. 9 workers, 1 queen, VEN- anotal groove scrobiculate. Dorsal face of EZUELA: Bolivar State, Campamento Rio Grande, propodeum transversely rugose with undulating, 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 of posterior face transversely rugulose. Rest of Petiolar dissected: mouthparts, 2 stings] [BMNH, LACM, posterior face shagreened. peduncle MCZ, MIZA, MZSP, USNM]. colliculate; anterior face of node smooth, except for piligerous punctures. Rest of petiolar node and Rogeria gibba new species all of postpetiolar node transversely areolate-rug- to Figs. 5-6 ose, with granulate ridges and smooth weakly punctured intervals, as on head and prome- Additions to stigmatica-group diagnosis. WL sonotum. Gaster smooth and shiny, dotted by 0.85-0.93mm. Mandibles subtriangular. Clypeal numerous small, shallow piligerous punctures. of and sides of apron slightly convex medially, with sharp cor- Dorsum head, mesosoma, top ners on either side. Eyes with 16-20 facets. waist, and all sides of gaster densely covered with erect to suberect hairs Mesosoma profile humpbacked. Propodeal spines long, flexuous, and without short (< 0.15mm), not strongly inclined. Propodeal interspersed shorter pilosity. darker. An- spiracle rather large, within one diameter of pos- Body rusty-brown, gaster slightly terior edge of mesosoma. Metapleural lobes re- tennae, legs and mandibles yellowish-brown to duced to short carinae, sometimes nearly absent. yellow. No inferior petiolar process. Dorsal head, Queens.—TL 4.2-4.3, HL 0.83-0.85, HW 0.80, mesosoma and gaster densely covered with long SL 0.58-0.59, EL 0.19-0.20, PW 0.72-0.73, WL 1.14- flexuous hairs. — 1.15, SpL 0.15-0.17, PetL 0.49-0.53, PpetL 0.24- Holotypeand Paratype Workers. TL 3.5-3.8 (3.7), 0.25mm, CI 0.94-0.96, SI 0.73-0.74, PSI 0.13-0.15. HL 0.78-0.85 (0.83), HW 0.73-0.81 (0.78), SL 0.54- N=4 0.60 (0.59), EL 0.09-0.10 (0.10) (15-22 facets), PW Queens differ from the workers in the usual 0.51-0.57 (0.55), WL 0.85-0.96 (0.93), SpL 0.11-0.14 and the following ways. Short parapsidal furrows (0.11), PetL 0.40-0.49 (0.46), PpetL 0.21-0.23 present. Longitudinal areolate-rugose sculpture (0.22)mm, CI 0.91-0.96 (0.94), OI 0.12-0.13 (0.13), SI on median head extends to posterior of head. 0.72-0.77 (0.76), PSI 0.12-0.15 (0.12). N=10 Median pronotum transversely rugose. Mandibles with 5 teeth diminishing in size Mesoscutum with longitudinally rugose sculp- of ture that areolate on mesoscutellum. basad. Palpal formula 3,3. Body clypeus does becomes—more not project over apron. Head capsule about as Discussion. See the ciliosa discussion for com- wide as long. Nuchal grooves visible from below parison. Like stigmatica also in many features, but or behind. Promesonotal dorsum almost flat and differs in pilosity and shape of promesonotum. to metanotum Petiolar The name is from Latin dropping abruptly (Fig. 5). gibba — meaning humpback. peduncle lacks inferior petiolar process. Petiolar Distribution. One worker is from the eastern node distinct, bulbous, wider than long. Postpetiole side of the central cordillera in northern Colombia, node subrectangular from above; sternum short. the rest are from about 1000km away on the west- Gaster large (GW/ WL 0.90-0.97). Sting apparatus ern slope of the Andes and the coastal range of like that of for: 1 a tulcral arm northern Ecuador. In all three areas were ciliosa, except ) longer they on the oblong plate (like inermis Fig. 42), 2) no collected at 300-800m elevation in natural rain companion seta on gonostylus, 3) more enlarged forest, probably by Berlese sampling. lancet apex, and 4) no anterolateral processes on Material Examined.—Holotype locality. EC- sting base (Fig. 6). UADOR: Pichincha Province, 4 km E. Santo Middorsum of head longitudinally rugose Domingo de los Colorados, 22-VI-1975, #B-304 (S. level of becoming rugose-areolate behind eyes. and J. Peck) [MCZ]. Laterodorsa, posterior, sides, and ventral surfaces Paratype localities. COLOMBIA: 1 worker, of head areolate with minutely granulate ridges; Antioquia Department, near El Bagre, Providencia, intervals smooth except for some piligerous punc- Estacion Biologica, Zona Buenos Aires, 30-31-XII- tures. Promesonotum with the same sculpture as 1977 (C. Kugler) [MCZ]; 1 worker, Choco Depart- sides and back of head. Meso- and metapleura ment, Rio Napipi, 1968 (P. A. Silverstone) [LACM]. with more confused areolate sculpture, but simi- ECUADOR: 5 workers, holotype locality, 22- VI- lar and intervals. Metanotal 1975 8-VII-1976 ridges groove and (S. and J. Peck) [2 mouthparts, scrobiculate. Dorsal face of propodeum and part stings] [BMNH, CKC, MCZ, MZSP]; 2 workers, Volume 3, 1994 33

Pichincha Province, 47km S. Santo Domingo, Rio Node large, wider than long, more or less sym- metrical in side Palenque Station, 23-V-1976 (S. and J. Peck [MCZ]; view (Fig. 8). Postpetiolar node in 1 worker, Pichincha Province, Tinalandia, 16km side view rounded front to back; usually widest in SE. Santo Domingo de los Colorados, 4-VI-1976 (S. anterior half, much as in Fig. 66, but sometimes

and J. Peck) [MCZ]. widest midlength. Postpetiolar sternum short. localities. 3 like that Nontype ECUADOR: queens, Sting apparatus of ciliosa (Fig. 3), except Pichincha Province, 47km S Santo Domingo, Rio for: 1) spiniform medial and and lateral projec- 1 tions Palenque Station, 1975 (S. and J. Peck); queen, from anterodorsal corner of quadrate plates, Manabi Province, 73km NE Chone, 12-VI-1976 (S. 2) smaller valve chamber, and 3) lack of

and J. Peck) [MCZ]. anterolateral processes on sting base (Fig. 9). Gonostylus sometimes with no clear sensillar gap; Rogeria stigmatica Emery sometimes lacking a companion seta. The "Rogeria Figs. 7-12 (stigmatica group) spp. 1 and 2" in Kugler (1978b) are both stigmatica. The sting shown here is more 1897:589. workers, NEW Rogeria stigmatica Emery Syntype accurate than the previous one, which was not in GUINEA: Friednch-Wtlhelmshafen [=Madang] (Biro) full lateral orientation when drawn. [MHN] (Both syntypes examined]. Middorsum of head Rogeria stigmatica subsp. sublevinodis Emerv 1914:415. Svntvpe longitudinally rugose; workers, LOYALTY ISLANDS: Mare, Raoua [MHN] rest of head, including venter, coarsely areolate. N. [Both syntypes examined). syn. Dorsum of promesonotum coarsely areolate to sublevinodis; Wilson and 1967:76, 61. Rogeria Taylor Fig. rugose (intermediate specimens predominantly Rogeria m < EL usually 0.10mm. Propodeal spiracle half its along very anterior margin, followed by either diameter of from edge infradental lamella. transversely rugulose or densely punctate sculp- Propodeal spines short (PSI 0.07-0.13), strongly ture, or both in varying degrees of density and inclined dorsad. Metapleural lobes much reduced. definition. Petiolar peduncle finely colliculate or Inferior petiolar process absent. Abundant de- smooth. Anterior and apex of node smooth or cumbent hair on head, mesosoma, waist and gaster; areolate; posterior face and sides areolate, some- erect hairs rather sparse, not flexuous. times with a few rugae. Postpetiole transversely Workers.— TL 3.0-3.7, HL 0.66-0.83, HW 0.58- rugose-areolate behind, becoming more effaced 0.71, SL 0.46-0.61, EL 0.05-0.15 (7-34 facets), PW anteriorly, often leaving anterior face smooth and 0.44-0.52, WL 0.72-0.92, SpL 0.05-0.10, PetL 0.32- shining. CI 0.42, PpetL 0.19-0.23mm, 0.84-0.92, OI 0.08- Decumbent to subdecumbent pilosity covers 0.22, SI 0.79-0.87, PSI 0.07-0.13. N=28 most of body. Sparser erect to suberect hair also on Mandibles subtnangular (usually) to triangu- dorsa of scapes, head, mesosoma, nodes, and with 5 teeth or 3 lar; subequal apical teeth followed gaster. Erect hair moderately abundant on gaster 3-4 smaller teeth or denticles. Tl of by (rarely 5) Palpal most specimens (Fig. 8), but sparse on speci- formula Little 3,3. or no clypeal apron; median mens from Papua New Guinea. Body of clypeus clypeal margin truncate, weakly convex, or weakly with strong median seta. of not angular. Body clypeus projecting over clypeal Color of mandibles, frontoclypeal region, an- margin. Eyes oval, large (EL 0.10-0. 15mm and 17- tennae, and legs light brownish-yellow to brown. 34 in of the facets), except some Papua New Guinea Rest of body light brown to blackish-brown. workers (EL 0.05-0.07mm; 7-9 indistinct facets). Queens.—TL 3.6-4.5, HL 0.73-0.85, HW 0.66- Nuchal groove distinct from behind, but not clear 0.75, SL 0.53-0.62, EL 0.19-0.24, PW 0.60-0.71, WL in side view. Promesonotum with evenly convex 1.00-1.17, SpL 0.10, PetL 0.41-0.48, PpetL 0.21- profile. Metanotal suture narrow, emphasized by 0.26mm, CI 0.87-0.90, SI 0.80-0.84, PSI 0.09-0.10. a sharp ridge at anterior edge of propodeum. N=6 34 Journal of Hymenoptera Research

As in worker except for the usual caste differ- punctation occur in the Solomon Islands, Irian ences. Mesosoma habitus as in Fig. 10. Queen from Jaya, and Pohnpei. McAdam Park, Papua New Guinea with median Santschi (1922) claimed that manni differed bulge on pronotum. Pronotum areolate on sides; from stigmatica in a variety of ways. After examin- finer and transversely rugose-areolate mesad. ing manni types, I could confirm only one clear Mesoscutum longitudinally rugose; mesoscu- way they differ from the types of stigmatica and tellum areolate-rugose. Metanotum smooth. sublevinodis: the presence of rugose sculpture in- Mesosoma sides confused areolate, except for stead of areolate sculpture on the promesonotum. smooth area on mesokatepisterna and costulate However, in non-type material, I found all inter- metapleural gland bullae. Wing venation as in mediate states, sometimes within the same local- in be/f; different characters also Hrata (Fig. 30), except for Rs vein as (Fig. 37). ity. Other supposedly Males.—TL 2.6-3.1, HL 0.45-0.54, HW 0.56- intergrade or are due to the manni types being at 0.66 SL 0.27-0.35, EL 0.20-0.26, PW 0.50-0.62, WL the small end of the size distribution. 0.84-1.04, PetL 0.22-0.30, PpetL 0.14-0. 19mm, CI I have too few collections from Papua New 1.22-1.27, SI 0.48-0.53. N=3 Guinea to know if those specimens with unusually Mandibles with a large apical tooth and 4 small eyes and few erect hairs on gaster Tl might others decreasing in size basad. Posterior lobe of be a distinct species. clypeus projects more broadly between antennae See sister species megastigmatica description. than in worker; anterior clypeal margin weakly See also ciliosa and gibba for discussions of related convex. Frontal lobes absent. No distinctly im- species in South America and exsulans for discus- pressed frontal area. Funicular segment 6 curved sion of a not so related species—from the Pacific. and longer than 4 and 7; more extremely curved Distribution and Behavior. Rogeria stigmatica and elongate on one side of the head than the is known only from the Central and West Pacific, other. Posterior outline of head medially concave; from as far east as Tahiti to the western tip of the sharp crests run from ocelli to posteroventral cor- Island of New Guinea and from about 22 S to 7°N. ners of head, which project slightly and fit around Most collections come from berlesate or sift- prothoracic sternum when head is retracted. ing of leaf mould, rotten wood, soil, moss, or bases

Mesosoma and waist as shown in Fig. 1 1 . Genitalia of fern epiphytes in rain forest. The one nest series as shown in Fig. 12. Head integument vaguely with ecological data (Sorong, Irian Jaya) is from roughened. Mesosoma and waist smooth, except rotten wood. Mann (1921:451) found colonies be- along furrows and on sides of propodeum, neath stones and logs. Twelve specimens, were metepimera, and petiolar peduncle. Gaster smooth collected on imported coconuts in Honolulu. If and shining. Pilosity all erect to suberect, except nesting occurs in coconuts, colonization of around eyes. Propodeum nude. Color variation as Polynesia and Melanesia from South America may in worker. have occurred by rafting on the South Equatorial Discussion. —According to Emery (1914) and Current. Wilson and Taylor (1967), sublevinodis differs from Mann (1921:451) observed workers produc- stigmatica in having larger size, coarser sculpture ing long, worm-like stands from the anal area on head and mesosoma, and smooth nodes with when the nest was disturbed. coarser punctures on other parts of the waist. Material Examined.—SOCIETY ISLANDS: Ta- When two of District Emery's stigmatica syntypes and two hiti, Punaauia (J. Dixon). AMERICAN of his sublevinodis in syntypes the MHN were SAMOA [=E. Samoa]: Tutuila [sting; whole speci- compared side by side, the TL and WL of the men]; Tafuna; Alega (T. E. Woodward). WEST- fell within the of the stigmatica syntypes range ERN SAMOA: Falepuna [sting]; Matautu; Vaipoto; sublevinodis syntypes. The sculptural characteris- Poutasi (T. E. Woodward); Le Mafa; Gagaifomauga tics were not distinct either, except on the dorsal (G Ettershank); Apia (H. Swale). WALLIS IS- face of the is propodeum, which transversely ru- LANDS: Nukuione (G. Hunt) [mouthparts, sting, gulose and very weakly punctate and shiny in whole specimen]. ILES DE HORNE [=Hoorn Is- stigmatica syntypes, but densely punctate and lack- lands]: Futuna (G. Hunt). FIJI: Vanua Ava; ing rugulae in the sublevinodis syntypes. However, Waiyanitu; Ovalau; Lasema; Somosomo; Nagasau; intergrades with rugulae and various degrees of Saiaro; Munia; Nadarivatu; Labasa (W. M. Mann); Volume 3, 1 994 35

Sigatoka (W. L. and D. E. Brown) [2 stings; male cate clypeal apron. Frontal region elevated and genitalia]. VANUATU [=New Hebrides]: Espiritu laterodorsa slightly concave. Eyes very small. Santo (E. O. Wilson). LOYALTY ISLANDS: Mare. Metapleural lobes and metapleural gland bulla SOLOMON ISLANDS: Ysabel, Fulakora [sting]; reduced. Body almost devoid of erect hair. Ugi [=Uki] (W. M. Mann) [sting]. PAPUA NEW Holotype Worker.— TL 2.88, HL 0.66, HW 0.56, GUINEA: New Guinea, Huon Peninsula (E. O. SL 0.51, EL 0.05 (4 facets), PW 0.435, WL 0.78, SpL Wilson); Wau, McAdam Park (S. Peck); Friedrich- 0.15, PetL 0.30, PpetL 0.19mm, CI 0.85, OI 0.09, SI Wilhelmshafen [=Madang] (Biro). INDONESIA: 0.91, PSI 0.19. Irian Jaya, Sorong (W. L. Brown). CAROLINE Mandibles with 5 teeth, basal angle greatly ISLANDS: Ponape [=Pohnpei] (Y. Kondo). 170 reduced (Fig. 14). Frontal lobes wide, closely ap- workers, 9 queens, 3 males [CUIC, MCZ, MHN, proximated. Eyes small, round, sunken; facets USNM], indistinct. Nuchal groove visible laterally as a distinct notch. Pronotal shoulders well rounded. Rogeria megastigmatica new species Dorsal face of propodeum wide, weakly concave. Fig. 13 Propodeal spiracle about 2 diameters from edge of infradental lamella. Metapleural lobes reduced to Holotype and Paratype Workers. — TL 4.5-4.9 low carinae. Metapleural gland bulla small. Peti- (4.5), HL ~1.01-1.09 (1.01), HW 0.90-1.00 (0.91), SL olar node large, wider than long. Petiolar pe- 0.77-0.83 (0.77), EL 0.16-0.19 (0.16) (41-52 facets), duncle short, with weak keel and small, dentate PW 0.62-0.70 (0.62), WL 1.18-1.30 (1.18), SpL 0.10- inferior process. Posfpetiole subrectangular in 0.11 (0.11), PetL 0.51-0.52 (0.52), PpetL 0.28-0.31 dorsal view. Postpetiolar sternum projects shelf- (0.28)mm, CI 0.88-0.92 (0.90), OI 0.18-0.19 (0.18), SI like under articulation with petiole. GW/WL 0.84. 0.83-0.87 (0.85), PSI 0.08-0.09 (0.09). N=5 Terminal segments of gaster not rotated ventrad. Like stigmatica in most respects, but markedly Shaft of undissected sting slender, with slightly larger (at least 35% larger than stigmatica on other enlarged apex; lancets acute and appear strong. Solomon Islands). Metanotal groove not as nar- Mandibles strongly carinulate for most of row or sharply defined. Petiolar node strongly length. Frontal area impressed, smooth. Median asymmetrical in side view (Fig. 13). head longitudinally rugose-areolate. Cheeks near Mandibles triangular, with at least 6 teeth, antennal insertions microareolate. Laterodorsa, some basal denticles may have been abraded. sides and posterior head densely macroareolate; Median clypeal apron convex. Mandibular cells small, usually obscuring the effaced carinulae effaced. Pronotum areolate with a rug- microareola te background. Promesonotal dorsum ose-areolate patch in center of disc. Dorsal face of with same areolate sculpture as on back and sides propodeum transversely rugulose. Petiolar pe- of head. Mesosoma sides strongly microareolate duncle smooth; sides of node areolate; posterior and macroareolate; in different places one or the face transversely rugulose-areolate or areolate- other predominates. Scrobiculate sculpture in rugose. — mesosoma sutures. Dorsal face of propodeum dis- Material Examined. Holotype locality. tinctly microareolate, with branching rugulae SOLOMON ISLANDS: Guadalcanal, Ilu Bush, 16- crossing between propodeal spines. Peduncle and III-1962, #1181 (P.J. M. Greenslade) (MCZ). anterior face of petiole as well as sterna of petiole Paratype localities. SOLOMON ISLANDS: 1 and postpetiole microareolate; rest of petiole and worker, holotype locality [MCZ]; 3 workers, San postpetiole macroareolate, somewhat effaced Kira #1579 Cristobal, Kira, 24-IV-1962, (P. J. M. mesad. Gaster Tl densely covered with piligerous Greenslade) [MCZ]. punctures; punctures weaker caudad. Other terga smooth except for vaguely roughened posterior Rogeria prominula new species margins. Fig. 14 Most of body covered with very fine, ap- pressed pilosity. Sparse erect hairs on clypeus, Additions to stigmatica-group diagnosis. WL frontal lobes and nearby head middorsum. A few 0.78mm. Basal angle of mandibles greatly reduced. short, decumbent hairs on mesosoma dorsum; of Body clypeus projects strongly over the trun- dense erect hairs on terminal segments of gaster. 36 Journal of Hymenoptera Research

like that of ciliosa No pilosity on ventral petiole. gaster. Sting apparatus much — of Discussion. This is perhaps the most aber- (Fig. 3), but: 1) anterior apodeme spiracular rant Rogeria species, with its unusual head shape plate widest midlength, 2) anterodorsal corner of anterior and pilosity. But it has the characteristic Rogeria quadrate plate longer, narrower, 3) antennae, nuchal grooves and square anteroventral apodeme of oblong plate longer, 4) gonostylus a corners of the pronotum. It has affinities with little longer, with two companion setae and less distinct in setation base lower ciliosa, gibba, and stigmatica, as described in the gap and 5) sting and stigmatica-group discussion. without anterolateral processes (Fig. 16). The name prominula, meaning little promi- Longitudinally rugose macrosculpture on nence, describes the body of the clypeus. frontal lobes becomes rugose-areolate on Material Examined.—Holotype locality. BRA- middorsum. Laterodorsa, sides, and posterior head areolate with rather small intervals smooth ZIL: Amazonas, Ig. Mananil, Rio Branco Road, areolae; 24km NE Manaus, 22-VIII-1962, #M-2 (W. L. and shining, except for piligerous punctures. Brown) [MCZ]. Pronotal disc varies from all rugose-areolate to all areolate. Rest of promesonotum slightly less Rogeria besucheti new species coarsely areolate. Intervals smooth, except for Figs. 15-16 piligerous punctures. Dorsal face of propodeum densely microareolate, with or without overlying Additions to stigmatica-group diagnosis. WL transverse rugulae. 0.67-0.75mm. Eye small (8-10 facets). Propodeal Dorsum and anterior face of petiolar node spiracles small, more than 3/4 diameter from smooth; rest of petiolar and postpetiolar nodes infradental lamella. PSI 0.17-0.20. Metapleural effaced areolate. Gaster predominantly smooth lobes well developed. Inferior petiolar process a and shiny; Tl and SI with piligerous punctures small step. Head, mesosoma, nodes and gaster that in some specimens become weaker caudad. with abundant decumbent pilosity and more sparse Remaining terga and sterna very weakly rough- erect hairs. — ened and shiny. Holotyi>e and Paratype Workers. TL 2.7-3.1 Paraguayan specimens have suberect hair on (2.85), HL 0.65-0.71 (0.66), HW 0.59-0.63 (0.60), SL scapes; others do not. Head dorsum with suberect 0.43-0.49 (0.44), EL 0.05-0.08 (0.06) (8-10 facets), hairs. Pilosity on mesosoma dorsum and nodes PW 0.42-0.46 (0.43), WL 0.67-0.75 (0.69), SpL 0.12- ranges from short and decumbent to long and 0.15 (0.13), PetL 0.30-0.34 (0.32), PpetL 0.16-0.18 erect; all curving quite strongly toward midline. (0.18)mm, CI 0.87-0.90 (0.90), OI 0.08-0. 13 (0.10), SI Gaster Tl similar, but with no erect hairs. Termi- 0.72-0.78 (0.72), PSI 0.17-0.20 (0.18). N=7 nal segments of gaster with rather dense erect hair Holotype mandible with 5 visible teeth de- and decumbent pilosity. No hair on ventral peti- creasing in size basad. In paratypes, mandibles ole. always with 3 apical teeth, but basal teeth may Extremities and mandibles light brownish have additional denticles or be replaced by pairs yellow. Rest of body brown with more yellowish of denticles. Palpal formula 3,2. Clypeal apron than reddish accents; frontoclypeal area and ter- truncate; body of clypeus projecting enough to minal segments of gaster lighter. block view of apron in full dorsal view. Pronotal Discussion. —Rogeria besucheti differs from shoulders rounded. Shallow metanotal groove ciliosa, gibba, prominula, and blanda in pilosity. It shallow present on dorsum and sides of mesosoma. differs from stigmatica and megastigmatica in Propodeum lacking a distinct transverse carina at mesosoma shape, propodeal spine length, and anterior border. Peduncle of petiole with weak generally smaller eye size. ventral keel; inferior process reduced to a small This species is named for Claude Besuchet, step. Petiolar node bulbous, wider than long. who as director of the MHN in Geneva was most Postpetiolar node widest in anterior half (as in Fig. helpful and patient in loaning material valuable sternum anterior not for this work. 74). Postpetiolar short, lip — greatly prominent. GW/WL 0.91-0.98. Terminal Distribution. Paraguayan specimens come segments of gaster slightly rotated ventrad, but from gallery forest with some bamboo. Peruvian not enough to make T3 the distalmost point of the specimens are from mixed broadleaf primary for- Volume 3, 1 994 37

est on a steep hillside at 1000m. In both localities curved, with weak keel and dentate inferior peti- collections resulted from Berlese and Winkler sam- olar process. Petiolar node widest in posterior pling of leaf litter and—rotten wood. half. Postpetiolar node highest in posterior half; Material Examined. Holotype locality. PARA- shape from above as in Fig. 32 or Fig. 49. Gaster not GUAY: Alto Parana Province, Puerto Santa Teresa, enlarged (GW/ / WL 0.63-0.70); terminal segments 3-XI-1979 (F. Baud, et al.) [MHN]. rotated ventrad. Pygidial gland sculpture present. Paratype localities. COLOMBIA: 3 workers, Sting apparatus nearly identical to that of inermis Putumayo Department, Villa Garzon, 23-VII-1977 (Fig. 42). (D. Jackson) [BMNH], PERU: 2 workers, Pasco Longitudinal rugae on frontal lobes rapidly Department, near Pozuzo, 1000m (C. Kugler and give way to areolate sculpture usually by mid-eye R. R. Lambert) [mouthparts, sting] [MCZ, MZSP]. level and continuing onto posterior head. Sides of PARAGUAY: 1 worker, San Benito Province, head strongly sculptured: rugose-areolate in front Itapua, 29-X-1982 (F. Baud, et al.) [MHN]. of eye to areolate behind. Mesosoma and nodes (Figs. 83-84) also densely areolate, except for trans- Rogeria blanda Fr. Smith verse carinulae between spines and smooth poste- Figs. 17, 83-84 rior face of propodeum. Some elongate cells on pronotal disc. Very apex of postpetiolar node some- blanda Fr. Smith 1858:131. BRA- Myrmica Syntvpe workers, times smooth. No microsculpture. ZIL: Amazonas, [=Tefe] [BMNH] [4 exam- Ega syntypes Scapes with very long erect hairs and shorter ined]. hairs ranging from suberect to decumbent. Erect Irogera foveaia Kempf 1964:64, Figs. 19-20. Holotype and hair on terminal of not brush-like. paratype workers, BRAZIL: Amazonas, Manaus (K. segments gaster Lenko) [MZSP] [Paratype examined] Legs with little decumbent or appressed pilosity. Rogeria blanda; Kempf 1965:185. Head, mesosoma, and waist black to yellow- ish-brown; appendages lighter. Gaster darker than — WL0.87-l.15mm. formula Diagnosis. Palpal rest of body. Head sometimes with black patches rather oval. 2,2. Eye large, Propodeal spines long around and/or between eyes. Petiolar node and low. Gaster (PSI 0.20-0.29). long Queens.—TL 3.6-4.1, HL 0.79-0.87, HW 0.70- with terminal Inferior segments rotated ventrad. 0.78, SL 0.52-0.58, EL 0.19-0.20, PW 0.61-0.70, WL dentate. and petiolar process Head, mesosoma, 1 .02-1 .17, SpL 0.24-0.30, PetL 0.40-0.52, PpetL 0.22- nodes areolate. rather densely Body densely cov- 0.25mm, CI 0.87-0.90, SI 0.73-0.81, PSI 0.22-0.26. ered with long, fine, soft, erect hair; no decumbent N=8 hair. Median pronotum sometimes transversely Workers.— TL 3.0-4.0, HL 0.73-0.92, HW 0.65- rugose-areolate; mesonotum longitudinally rug- 0.82, SL 0.49-0.62, EL 0.12-0.16 PW (27-59 facets), ose-areolate. Otherwise, differing from the worker WL PetL 0.35- 0.48-0.63, 0.87-1.15, SpL 0.20-0.30, only in the normal queen characters. Wing vena- 0.49, 0.18-0.25mm, CI 0.85-0.91, OI 0.18- PpetL tion like that of belti (Fig. 37). SI PSI 0.20-0.29. 0.21, 0.73-0.80, N=20 Males.—TL 3.2-3.4, HL 0.60-0.62, HW 0.70- Additions to of description and figures Kempf 0.75, SL 0.23-0.26, EL 0.32-0.35, PW 0.79-0.86, WL Mandibles with 6 teeth (1964). triangular, large 1.10-1.24, PetL 0.34-0.35, PpetL 0.18-0.20mm, CI decreasing in size basad, then basal tooth larger 1.16. SI 0.32-0.35. N=4 than neighbor. One or two denticles sometimes All males came from Rio Akaban, Venezuela. added between basal teeth. with a Clypeal apron Mandibles with 4 subequal teeth. Anterior edge of shallow median notch. Nuchal not visible grooves clypeus with weak median notch. Frontal area a in lateral view. Metanotal weak to groove (Fig. 83) distinctly impressed triangle. No frontal lobes. absent. small, directed Propodeal spiracle strongly Flagellomeres 2-11 straight, subequal in length caudad. or Propodeal spines long, straight weakly and width. Habitus much like male of belti (Fig. at in the two upturned apex; distinctly longer 38), but junction of dorsal and posterior faces of from Ecuador than in specimens (PSI 0.29) the propodeum has blunt lateral corners, and the others (PSI 0.20-0.25). lobes Metapleural triangu- propodeal spiracle faces more caudad. lar; blunt to subacute. Petiolar apex peduncle Back of head microareolate with piligerous 38 Journal of Hymenoptera Research

median vertex with addi- and Related Species punctures in the pits; Scandens-Group tional longitudinal rugulae. Pronotum and scandens Mann mesopleura largely smooth. Mesonotum densely Rogeria in 19-20 and finely rugulose with scattered punctures Figs. intervals. Posterior face of propodeum, nodes, work- in 17. Macrormscha scandens Mann 1922:30, Fig. 14. Syntype and gaster smooth. Genitalia as shown Fig. ers, HONDURAS: Lombardia (Mann) (USNM) [5 of b Hairs shorter, less flexuous than in worker; erect syntypes examined]. of to suberect and dense over much 438. moderately Irogera scandens; Kempf 1962a:436, with brown body. Color dark brown lighter ap- Rogeria scandens; Kempf 1965:185. pendages. Discussion. —Rogeria ciliosa and gibba from low- Additions to scandens-group diagnosis. WL land South America resemble blanda in having 0.93-1. 17mm. Eye very large (about 60-80 facets). areolate Petiole clav- long, soft, dense pilosity and sculpture, Propodeal spines long (PSI 0.25-0.28). but see the stigmatica group diagnosis. ate, with rather large node (PetW/PetL 0.49-0.58). Distribution. —Rogeria blanda is found in south- Postpetiole wide (PpetW/PpetL 1.38-1.61). of ern Central America and in South America east Mesosoma and petiolar node strongly macro- the Andes to southern Brazil. Elevations range sculptured. Erect hairs with dentate ends. from 50m (Costa Rica) to 1000m (Venezuela). Nests Workers.— TL 3.5-4.2, HL 0.83-1.02, HW 0.70- have been found in trunks of cacao trees in Costa 0.85, SL 0.59-0.73, EL 0.17-0.20 (60-80 facets), PW Rica and in a small rotten log suspended about 0.53-0.62, WL 0.93-1.17, SpL 0.23-0.30, PetL 0.42- 50cm above the ground in Peru. 0.55, PpetL 0.23-0.27mm, CI 0.82-0.86, OI 0.21- Material Examined.—COSTA RICA: Heredia 0.24, SI 0.84-0.86, PSI 0.25-0.28. N=6 Province, Puerto Viejo de Sarapiqui, La Selva Sta- Mandible in most specimens with 5 teeth de- tooth. denticle tion (L. Garling); Santa Clara Province, Hamburg creasing in size until large basal A Osa tooth. Farm (F. Nevermann); Puntarenas Province, may appear between basal and penultimate PANAMA: not visible in lateral view. Peninsula, Corcovado (J. Longino). Nuchal grooves shallow, Barro Colorado Island (Brown and McCluskey; D. Metanotal groove in lateral view broad and shal- TRINIDAD: Basin Hill Re- accentuated a E. Wheeler; J. Zetek). low to absent; groove sometimes by serve (N. A. Weber) [mouthparts, sting, whole low transverse ridge at anterior of propodeum. near in Hon- specimen]. VENEZUELA: Falcon State, Propodeal spines more strongly inclined Curimagua, Haitoncito; Monagas State, Caripe; duran (Fig. 19) than Panamanian specimens. Bolivar State, Talud, south of Amarawai Tepuii Metapleural lobes somewhat longer and narrower Rio male Petiolar node and Akaban (J. Lattke) [2 genitalia]. in Honduran specimens. slightly River A. GUYANA [=British Guiana]: Oko (N. more distinct in Honduran specimens (Fig. 19). Manaus Weber). BRAZIL: Amazonas State, Tefe; Sting apparatus like that of biennis (Fig. 42), except W. L. K. Para State, Icoarci anterior and a more rounded vicinity ( Brown; Lenko); for wider apodemes and (W. L. Brown), Belem (N. Rosa), Jacareacanga (M. posterodorsal corners on spiracular plates Alvarenga); Mato Grosso State, Municipal larger anterolateral processes on the sting base Santo Diamantino (W. L. Brown); Espiritu State, (pygidium and anal plate lost in preparation). Linhares (M. Alvarenga). ECUADOR: Napo Prov- Median clypeus with additional 1-2 pairs of PERU: Madre de the usual for ince, Limoncocha (R. Chadab). fairly distinct carinulae lateral to pair Lake Dios Department, Puerto Maldonado, the genus. Posterior head longitudinally rugose 93 work- Sandoval (C. Kugler) [mouthparts, sting]. (continuing from middorsum), transversely arch- ers, 8 queens, 4 males [BMNH, CKC, CUIC, MCZ, ing rugose, or transversely rugose-areolate. MIZA, MZSP, USNM]. Interrugal spaces on head distinctly to weakly granular; microsculpture weaker and surface half of shinier on posterior. Oval area on ventral 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. on Panamanian Rugae specimens Rogeria terescandens new species not so vermiculate on and with lat- mesonotum, Fig. 21 eral spurs. Sides of pronotum with parallel, —Like that of upcurved carinae, which are more numerous and Diagnosis. scandens, except: 1) more distinct on Honduran specimens. Dorsal Propodeal spines shorter, 2) petiolar node lower, more slender face of propodeum confused areolate or areolate- (PetW/PetL0.40-0.41) / 3)postpetiole narrower and rugose. Intervals in mesosoma macrosculpture (PpetW/PpetL 1.08-1.16), 4) smooth and shiny. Petiolar node heavily areolate- macrosculpture very weak, especially on on mesosoma and waist, and smaller. rugose sides and posterior face; sculpture 5) eye slightly and Workers.— weaker dorsad; no clear microsculpture. Holotype Paratype TL 3.6, HL (0.88)-0.89, HW SL EL 0.16 Postpetiolar node weakly rugose or rugose-ar- (0.68)-0.70, (0.70)-0.71, eolate on (49-54 facets), PW WL sides and posterior surface, becoming (0.50)-0.52, 1.00, SpL (0.18)- PetL weaker, sometimes absent toward midline; 0.21, (0.45)-0.47, PpetL 0.25mm, CI (0.77)- 0.79, OI SI PSI microsculpture vaguely microareolate to nearly 0.23-(0.24), 1.01-(1.03), (0.18)-0.21. smooth. N=2. Also from scandens in the Scapes and extensor surfaces of legs without differing following Mandible with erect hair. Dorsa of head, mesosoma, nodes, and ways. 6 teeth. Profile of mesosoma Tl dorsum almost no gaster sparsely covered with fine appressed evenly convex; metanotal or at front of hair and longer, stiff, thick, erect-suberect hair groove ridge propodeum. Median with toothed carinulae on weak; no lateral carinulae. apex (Fig. 20). Mandibles, clypeus clypeus and terminal Head dorsum overlain on segments of gaster with long erect densely microareolate; middorsum of of hair that is tapered and less stiff. by wisps longitudinal rugulae, on laterodorsa Color of Honduran specimens reddish-brown. by faint reticulations, and on poste- rior of head Panamanian specimens blackish-brown. Append- by fragmented, transversely arching on sides ages often lighter. rugulae. Microsculpture and posterior head effaced. Mesosoma Queen.—TL 4.2, HL 0.95, HW 0.82, SL 0.63, EL dorsum, including dor- sal face of 0.22, PW 0.69, WL 1.20, SpL 0.28, PetL 0.50, PpetL propodeum, densely microareolate, with 0.30mm, CI 0.86, SI 0.76, PSI 0.23. N=l superimposed patches of fine rugulose-ar- eolate Differing from the Panamanian workers in the macrosculpture. Pronotal sides shiny and usual coriarious with some effaced queen characteristics. Pronotum transversely longitudinal rugulae. and also near rugose. Mesonotum longitudinally rugose with- Mesopleura metapleura shiny coxae, out but more dorsad with microareolate and cross-connections or vermiculate appearance. opaque confused Discussion. —The Honduran and Panamanian rugulose sculpture. Sides of petiolar node be with effaced microareolate and specimens may separate species. They differ shiny, background in of vestigial smooth slightly shape mesosoma and petiole, sculp- longitudinal rugulae. Postpetiole and ture, and color, and the Honduran specimens are shiny. Discussion. —The a little larger (WL 1.13-1. 17mm) than the Panama- name of this species refers to nian like that of R. but specimens (WL 0.93-1. 08mm). I prefer to call sculpture scandens, smoother, as if rubbed L., rubbed these geographic variants, however, until we have (teres — off). more specimens from more localities. Distribution. Both specimens of terescandens See were taken from trees in lowland the terescandens description for compari- forest on the Pacific side of Costa Rica. sons with its sister species. The holotype was col- — lected in a two week old treefall Ecology. Some specimens from the Canal by general collect- Zone were collected from Heliconia. ing on trunks. The paratype was on or beneath a Material thick mat on the base of a fallen Examined.—HONDURAS: Lombardia epiphyte branch (J. (W. M. Mann). PANAMA: Canal Zone, Barro Colo- Longino unpublished field notes). rado Material Examined. — Island (W. L. Brown and E. S. McCluskey; D. Holotype locality. COSTA RICA: Osa 8.28N Wheeler; J. Zetek) [2 workers: mouthparts, whole Peninsula, Sirena, 83.35W, 50m, #0950 specimen; sting]. 31 workers, 1 queen [CKC, 31-111-1982, (J. Longino) [MCZ]. 1 LACM, MCZ, USNM]. Paratype locality. worker, holotype locality, 28- #1100 V-1981, (J. Longino) [LACM]. 40 Journal of Hymenoptera Research

Rogeria subarmata Kempf miculate-rugose to rugose-areolate. Postpetiolar effaced. Figs. 22-23 node similar, but rugae straighter, more Microsculpture on nodes slightly weaker than on Irogera subarmata Kempf 1962a:438, Figs. 1-4. Holotype and head and mesosoma. workers, BRAZIL: Guanabara, Rio de Janeiro, paratype Scapes and extensor surfaces of legs lack erect Deodoro (A. Ronna) [MZSP] [12 of 38 paratypes exam- hair. Rest of body with both short, appressed- ined; holotype not examined]. decumbent and longer, erect-suberect hairs. Erect Rogeria subarmata; Kempf 1965:185. hairs are as stiff as those of scandens and Rogeria subarmata; Kempf 1975:367 [new records]. nearly terescandens (Fig. 20), but seem to have acute tips. Additions to WL scandens-group diagnosis. Color dark brown to yellowish-brown with a 0.78-1. 00mm. with 30-53 facets. Eye Propodeal reddish tint on mesosoma, waist and middle of short (PSI 0.09-0.12). Pygidium with a pair spines gaster; appendages and ends of gaster lighter. of median tubercles near caudal mar- piligerous Distribution. —All available specimens are from on mesosoma and gin. Strong macrosculpture localities along the coast of Brazil. The type series node. Erect hairs not as as in petiolar rigid scandens; was collected from the stomach of an anteater acute. tips (Tamandua tetradactyla). Workers.— TL HL 0.69-0.87, HW 0.60- 2.9-3.7, Material Examined. —BRAZIL: Para State, 0.75, SL 0.46-0.57, EL 0.12-0.16 (30-53 facets), PW Belem (N. Rosa) [mouthparts, sting]; Bahia State, 0.45-0.57, WL 0.78-1.00, SpL 0.08-0.10, PetL 0.37- Itabuna (J. A. Winder); Espirito Santo State, Pedro 0.47, 0.20-0.26mm, CI 0.86-0.89, OI 0.20- PpetL Canario near Conceicao da Barra (M. Alvarenga); 0.23, SI 0.73-0.77, PSI 0.09-0.12. N=6 Guanabara State [=Rio de Janeiro State], Rio de Additions to (1962a) Kempf's description. Janeiro, Deodoro (A. Ronna). 20 workers [MCZ, Mandibles with 5 teeth that decrease in usually MZSP]. size basad. Sometimes basal tooth replaced by two small teeth, or 1-2 denticles are found be- very Rogeria procera Emery tween the basal and penultimate tooth. Clypeal Fig. 18 apron weakly notched medially to evenly convex. 19. worker, Frontal lobes narrow as in scandens (Fig. 19). Nuchal Rogeria procera Emery 1896:92, Fig. Holotype BRAZIL: Para, Belem [MCSN] examined]. grooves shallow, forming only a weak notch in [Holotype Rogeria {Irogera) procera; Emery 1915:191. lateral view. Figs. 22-23 show the range of Macromischa brasiliensis Borgmeier 1953(1951 ):107, Figs. 1-4. size and but some- propodeal spine angle, tips Holotype worker, BRAZIL: Para, RioCumina, Cachoeira times more rounded. Petiole clavate to rather dis- do Breu [MZSP] [Holotype examined). [Synonymy by tinctly set off from peduncles (Figs. 22-23). Kempf 1962a:437], Irogera procera; Kempf 1962a:436 [partial description]. Postpetiole from above much as in Fig. 21. Poste- Irogera procera; Kempf 1964:66 [partial key]. rior surface of pygidium with a caudal pair of Rogeria procera; Kempf 1965:185. long, columnar, piligerous tubercles that are vis- — ible at 50X with a dissection microscope. Sting Diagnosis. WL 1.28-1. 53mm. Eye very large. apparatus nearly identical to that of inerniis (Fig. Propodeal spines long, not inclined. Postpetiolar 42); sting as in pellecta (Fig. 33). node small, subconical; sternum inconspicuous Median clypeus with 1-2 pair of fairly distinct and without a differentiated peduncle. Middorsum extra carinulae lateral to the usual pair. Posterior and posterior head with fine, nearly straight longi- head with transversely arching rugose-areolate tudinal rugae. Mesosoma and petiolar node with macrosculpture. Head covered with dense, indis- thick, vermiculate rugae. Erect hairs tapered; rarely tinctly microareolate roughening that appears any appressed or decumbent hairs on mesosoma punctate or granular at lower magnifications. dorsum, gaster, or legs. Scapes and extensor sur- Mesosoma dorsum longitudinally rugose; rugae faces of legs with erect hairs and appressed to with numerous lateral spurs that occasionally con- decumbent pilosity. Palpal formula, propodeal nect rugae on shoulders. Macrosculpture on sides spiracle, metapleural lobes, petiole, pygidium, and of mesosoma and dorsal face of propodeum con- sting apparatus as in scandens-group diagnosis. fusedly rugose to rugose-areolate. Mesosoma Workers.— TL 4.5-5.4, HL 1.07-1.19, HW 0.91- microsculpture as on head. Petiolar node ver- 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, on extensor surfaces of legs. The following as in scandens- PetL 0.60-0.73, PpetL 0.28-0.33mm, CI 0.85-0.91, group diagnosis: metapleural lobes, peti- OI 0.20-0.22, SI 0.70-0.75, PSI 0.18-0.23. N=7 ole, pygidial gland sculpture, sting apparatus, and Mandibles triangular with 6 teeth; basal larger sculpture. than neighbor. Clypeal apron with shallow me- Workers.— TL 3.0-3.2, HL 0.72-0.78, HW 0.61- dian notch. Nuchal grooves shallow, not visible in 0.68, SL 0.48-0.53, EL 0.12-0.15 (39-48 facets), PW lateral view. Sting apparatus like that of inermis 0.46-0.50, WL 0.81-0.90, SpL 0.18-0.21, PetL 0.37- (Fig. 42), but spiracular plate with more rounded 0.40, PpetL 0.17-0.19mm, CI 0.85-0.88, OI 0.19- anteroventral corner and gonostylus with sepa- 0.24, SI 0.78-0.81, PSI 0.20-0.25. N=7 rate proximal and distal patches of sensilla. Mandibles triangular; most specimens with 6 Laterodorsa of head longitudinally rugose- teeth, the first 5 decreasing in size basad then a areolate; sides below eye smooth and shiny. large basal tooth. In others, the penultimate basal Microsculpture vaguely microareolate on anterior is replaced by 2-3 denticles. Clypeal apron convex laterodorsa; more effaced on rest of head, impart- with median angle or small tooth. Body of clypeus ing a vaguely granular, shiny appearance between rises perpendicularly. Eyes oval with narrow an- rugae. Promesonotum longitudinally vermiculate- terior point. Nuchal groove inconspicuous. rugose. Meso- and metapleura rugose. Interrugal Pronotum in lateral view lacks a distinct angle spaces on mesosoma and petiole almost smooth. between anterior and dorsal faces. Metanotal Postpetiolar node shiny with vague rugae and groove very weak to absent. Propodeal spines weak microsculpture. long, weakly inclined. Postpetiole subtrapezoidal Scapes and head dorsum with erect hair and from above; sternum flat, with a distinct peduncle. much shorter to appressed decumbent pilosity. Sting apparatus like that of inermis (Fig. 42) except Mesosoma, legs, waist, and gaster with erect hairs, for more elongate anterolateral processes on sting but rarely any decumbent or appressed hairs. bulb. Mesosoma and waist black to dark brown; Middorsum of head longitudinally rugose; appendages and sometimes gaster lighter brown. laterodorsa and dorsal part of sides rugose-ar- Material Examined.— eolate. GUYANA (British Mesosoma (Figs. 85-86) with thicker rugae Guiana): Oronoque River (N. A. Weber). BRAZIL: and narrower interrugal spaces than on head. Para State, Rio Cumina (A. Sampaio), Ourem (A. Rugae transverse on anterior pronotum, trans- Schulz); Amazonas State, Manaus (K. Lenko); verse to confused on metanotum and dorsal face Manaus to Itacoatiara (W. L. Brown) [mouthparts, of propodeum, predominantly longitudinal on 23 workers sting]. [CUIC, MCZ, MZSP]. sides and pronotal disc. Petiolar node longitudi- nally rugose on sides; smooth along midline. Rogeria tonduzi Forel Postpetiole smooth. Microsculpture weak or ab- Figs. 85-86 sent throughout, integument very shiny. Scapes and head with erect to suberect hair Rogerm tonduzi Forel 1899:53. worker, COSTA RICA Holotype along with the typical short decumbent pilosity. (Tonduz) (MHN) [Holotype examined). Mesosoma dorsum and waist generally with erect Irogera tonduzi, Emery 1915:191. to suberect hair of a variety of lengths; no decum- Irogera tonduzi; Kempf 1962a:436. Gaster with erect Rogeria tonduzi; Kempf 1965:185. bent-appressed pilosity. long, hair and a few short, decumbent hairs. Color black with dark brown Diagnosis.—WL 0.81-0.90mm. Eye large. Pal- shiny mandibles, and to reddish-brown with pal formula 2,2. Propodeal spiracle faces laterally. scapes legs vellowish- Propodeal spines long. Postpetiolar sternum not brown appendages. Discussion. — belti occurs in enlarged. Posterior head with transversely arch- Rogeria the same localities and could be confused ing rugae. Rugae on mesosoma and petiolar node with tonduzi, but thick and rounded. Decumbent hair abundant on belti has a distinct petiolar node and predomi- areolate head dorsum and legs; little on gaster; no decum- nantly pronotal sculpture. Rogeria lirata bent or appressed hair on mesosoma or nodes. (Figs. 28-30, 89-90) from northern South America is similar to tonduzi in Scapes, head dorsum, mesosoma, nodes and gaster size, clypeus, and sculpture, but lirata with abundant flexible, tapered, erect hair; none has smaller eyes (6-12 facets), a distinct 42 Journal of Hymenoptera Research

petiolar node, and scapes with suberect hair only. 0.44, PpetL 0.19-0.23mm, CI 0.86-0.89, OI 0.09- Distribution. —With the exception of one speci- 0.14, SI 0.75-0.80, PSI 0.20-0.23. N=9 men from Guatemala, all tonduzi specimens come Additions to Kempf (1962b). Dentition vari- from Costa Rica at elevations of 0-100m on both able; simplest pattern is 6 teeth of decreasing size, sides of the cordillera. Most specimens were col- however, it seems that any or all of the last 3 teeth lected by Jack Longino as strays on ground and may be replaced by a pair of denticles. FLW/HW vegetation. He found one worker in a Cyphomyrmex 0.32-0.36. Posterior outline of head concave to nest and another "...on the base of a small tree, convex. Anterior propodeum marked by a trans- amongst some Pheidole workers" (unpublished verse carina. Petiolar node distinct; Figs. 24 and 26 field notes). Lyn Garling found a nest with a show extremes of shape. Postpetiolar node with a "tubular entrance with white 'fuzz'" in a cacao distinct posterior face; shape from above as in tree (field note on label). Figs. 24 or 32. Sting apparatus differing dramati- Material examined. cally from that of inermis (Fig. 42) in some features: GUATEMALA (no locality or collector). 1) sting shaft and lancets very weak, 2) lancets COSTA RICA: Heredia Province, Puerto Viejo de spatulate as in Fig. 29, 3) sting shaft apex with Sarapiqui, La Selva Station (L. Garling); Puntarenas eroded sides, and no dorsal flange (Fig. 27), and 4) Province Carara Biological Reserve (P. S. Ward), furcula with a shorter dorsal arm that broadly lateral Osa Peninsula, Corcovado National Park (J. merges with the arms, thus appearing Longino) [mouthparts, sting]. 24 workers [BMNH, broadly V-shaped in anterior view. CKC, JTLC, LACM, MCZ, MZSP, USNM]. Posterior head transversely arching rugose to Gerwaim-Group and Related Species rugose-areolate. Promesonotal dorsum varies from predominantly areolate (with occasional elongate Rogeria germaini Emery cells) (Fig. 88) to predominantly vermiculate-rug- Figs. 24-27, 87-88 ose. Pronotal sides rugose; rugae subparallel with ventral edge of pronotum to diagonal. BRAZIL: Rogeria germaini Emery 1894:189. Syntype workers, Microsculpture vestigial, leaving irregular, but MatoGrosso (Germain) [MCSN] [1 of 2 exam- syntypes shiny spaces in macrosculpture; sides of mesosoma ined] especially smooth. Rogeria germaini minensis Santschi 1923:1262. Lectotype and Color to brown with darker paratype workers, BRAZIL: MinasGerais, PassaQuatro brownish-yellow at (Reichensperger) [NMB] [Both lectotype and paratype gaster; appendages times slightly lighter. examined]. N. syn. Queen.—TL 4.1, HL 0.90, HW 0.79, SL 0.73 EL 1962b:20, 18, 19 Rogeria germaini; Kempf Figs. [Redescnbed] 0.17, PW 0.67, WL 1.12, SpL 0.15, PetL 0.48, PpetL Rogeria minensis; Kempf 1963:189, Figs. 1, 2 [Redescnbed, 0.28mm, CI 0.88, SI 0.92, PSI 0.13. N=l raised to species]. Differing from the worker in the normal queen Additions to gemiaini-group diagnosis. WL characteristics and the following. Notch formed 0.74-0. 90mm. Clypeal apron with median tooth. by nuchal groove not so distinct in lateral view. Metanotal groove very weak or absent. Metapleural Anteroventral corner of pronotum not as clearly lobes reduced, not angular. Petiolar keel not lamel- dentate. Parapsidal furrows not distinguishable late. Sting shaft apex weak, lacks dorsal flange; from furrows in sculpture. Diverging rugae on lancets spatulate. Macrosculpture effaced on side middorsum of head continue onto posterior head, of head below eye; sometimes nearly smooth. with few or no lateral spurs. Laterodorsa and sides Promesonotal rugae sharp and narrow like those of head similarly rugose. Anterior face of prono- on head; rugae low on sides of pronotum do not tum transversely areolate, mesonotum with longi- continue onto me tanotum. Macrosculpture on both tudinal rugae diverging from an anterior point, nodes and petiolar peduncle vestigial. Scapes with- then parallel in posterior half; some branching, out erect hair. Mesosoma with erect-suberect only. but no cross-ridges; less vermiculate than in Head and gaster with both erect and decumbent worker. — pilosity. Discussion. Kempf (1963) examined the types Workers.— TL 2.7-3.5, HL 0.66-0.85, HW 0.59- of germaini and minensis and noted that they were 0.74, SL 0.45-0.59, EL 0.06-0.10 (7-13 facets), PW very similar, but chose to retain both names on the 0.44-0.51, WL 0.74-0.90, SpL 0.15-0.21, PetL 0.30- 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 PW 0.39-0.51 (0.51), WL 0.65-0.87 (0.86), SpL 0.12- specimens collected in Paraguay in 1979 and 1982. 0.20 (0.175), PetL 0.31-0.47 (0.44), PpetL 0.17-0.23 These new collections bridge the gap between the (0.21)mm, CI 0.83-0.88 (0.86), OI 0.09-0.12 (0.12), SI types oigermaini and minensis. They are all inter- 0.79-0.85 (0.82), PSI 0.19-0.23 (0.20). N=10 mediate in size between the smaller germaini type Dentition as in germaini. Nuchal groove forms and the larger minensis types, have convex poste- a weak notch in lateral view. No clear angle be- rior heads like ihegermain i type, have promesonotal tween anterior and dorsal faces of pronotum. Spi- sculpture varying from nearly as extensively rug- racle faces slightly caudad, posterior edge within ose as thege rmaini type, to areolate like the minensis one diameter of nearest edge of propodeum. Peti- types; and some have petiolar nodes intermediate olar peduncle with a sharp, but not lamellate keel. between the more abruptly arising tninensis-hke Petiolar node profile angular. Postpetiolar node in and the more obliquely arisingger??!fl;>n-like nodes. dorsal view shaped as in Figs. 24 or 49; sternum Rogeria lirata (Figs. 28-30, 89-90) from more flat. Sting apparatus as described for germaini. northern parts of South America is germaini' s clos- Divergent longitudinal rugae on frontal lobes est relative (See lirata description for compari- grade into areolate sculpture at level of eyes. Sculp- sons). Rogeria lacertosa (Fig. 31), also from southern ture of posterior head transversely arching rugose Brazil, differs in size, clypeal shape, sculpture on to rugose-areolate. Laterodorsa and sides of head side of . head, and pilosity Rogeria pellecta (Figs. 32- longitudinally rugose-areolate; sides may be ef- 33), collected further south in Brazil, differs in faced to nearly smooth. Head microsculpture ves- clypeal shape, metapleural lobes, sting, and pilos- tigial; intervals nearly smooth, quite shiny. Ante- ity. rior face of pronotum areolate, becoming ver- Distribution. — far So germaini is known only miculate-rugose on disc (Fig. 90). Thick, rounded from southern Brazil and Paraguay. Most speci- rugae of promesonotum may merge, but are rarely mens have come from Winkler apparatus collect- joined by cross-ridges. Rugae on meso- and ing by expeditions from the MHN in Geneva. metapleura sharper and more separated. Anterior Specimens were extracted from rotting leaf litter and dorsal faces of petiolar node smooth. Sides and wood in forests. of — and posterior faces node in holotype longitudi- Material Examined. BRAZIL: Minas Gerais nally rugose-areolate, but weaker and more rug- State, Passa Quatro (Reichensperger); Mato Grosso ose in Trinidad specimen and nearly smooth in State; Sao Paulo State, Anhembi [=Piramboia, 29km specimens from Peru. Microsculpture weak or E Botucatu] (Kempf et al.). PARAGUAY: Alto absent; integument shiny. Parana Province, Puerto Santa Teresa; Misiones Scapes with decumbent to suberect hair (holo- Province, Panchito Lopez; Itapua Province, San type) or with short, uniformly subdecumbent hair. Benito Island [2 workers: whole specimen; mouth- Head dorsum with subdecumbent to erect hair. parts, sting]; Itapua Province, Santa Maria; Cen- In most, mandibles, frontoclypeal region, an- tral Province, Asuncion (F. Baud et al.). 15 work- tennae and legs brownish-yellow; other parts rusty- ers, 1 queen [BMNH, CKC, MCSN, MCZ, MHN, brown, becoming darker, almost black, on dorsa NMB], of mesosoma and petiole. Trinidad and Guyana specimens lighter on all parts. Rogeria lirata new species Queens.—TL 3.6-3.7, HL 0.76-0.80, HW 0.69- Figs. 28-30, 89-90 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- Diagnosis. As in germainibut: 1) metapleural 0.24mm, CI 0.88-0.90, SI 0.76-0.81, PSI 0.18-0.21. lobes larger, triangular, and 2) rugae on N=4. promesonorum more rounded and thicker than Besides the usual features of an alate female, on head; one ruga begins near anteroventral cor- differing from the worker in the following ways. ner of pronotum and continues unbroken to the Mandible with 7 teeth or 3 teeth and 5 denticles. metanotum. Parapsidal furrows not distinct from furrows in and — Holotype Paratype Workers. TL 2.6-3.4 sculpture. Propodeal spines project caudad or (3.3), HL 0.63-0.80 (0.78), HW 0.53-0.70 (0.67), SL posteroventrad. Rugae on sides of pronotum al- 0.45-0.56 (0.55), EL 0.05-0.08 (0.08) (6-12 facets), most vertical to shoulder, then bend across the 44 Journal of Hymenoptera Research anterior face of the pronotum. Mesonotal rugae 0.81, SL 0.59-0.61, EL 0.10-0.11 (19-20 facets), PW longitudinal with occasional fusions and cross- 0.52-0.60, WL 0.93-1.05, SpL 0.17-0.20, PetL 0.43- CI OI 0.13- ridges. Wing venation (Fig. 30) most similar to that 0.45, PpetL 0.23-0.26mm, 0.87-0.88, SI PSI 0.18-0.20. of stigmatica. Radial sector and median veins long, 0.14, 0.75-0.79, N=4 nearly reaching wing margin; r-m vein present. The following supplements Kempf (1963). All Discussion. —See the tonduzi discussion for specimens at hand with 6 mandibular teeth de- lirata in size for a basal comparisons with that species. The name creasing basad, except large a refers to the characteristic rugae of the mesosoma, tooth. Eyes elliptical. Nuchal groove forms notch which resemble the ridges thrown up by a plow. in in lateral view of head. Anterior edge of Material Examined. —Holotype locality. CO- propodeum not marked by a transverse carina. LOMBIA: 7km N Leticia, forest litter, 10-25-11- Metapleural lobes low and broadly rounded or with 1972, #B-230 (S. and J. Peck) [MCZ]. triangular more narrowly rounded apex. Paratype localities. TRINIDAD: 3 workers, #191 Postpetiolar node from above like that of pellecta as biennis (N. A. Weber) [MCZ, MZSP]; 1 worker, Mayaro, (Fig. 32). Sting apparatus (Fig. 42), but shaft and lancets are less sclerotized Trinity Hills Reserve, 5-VIII-1976 (J. Noyes) sting (easily [BMHN]. GUYANA (British Guiana): 1 worker, R. twisted) and the lancets lack the barbule. Mazaruni Forest Settlement, 20-VIII-1935, #304 Laterodorsa of head predominantly rugose to (N. A. Weber) [MCZ]. COLOMBIA: 2 workers, rugose-areolate. Back of head areolate in a trans- holotype locality [mouthparts, sting; one coated versely arched pattern. Microsculpture vestigial; for electron microscopy] [CKC, MCZ]. PERU: 2 intervals with a shiny, effaced granular appear- workers, Loreto Department, 15km WSW ance. Anterior face of pronotum transversely ar- Yurimaguas, 5.59S 76.13W, 200m, 22-VII-1986, eolate; disc with diverging, longitudinal, vermicu- #8701-24 and #8701-25 (P. S. Ward) [mouthparts, late rugae with variable number of cross-ridges sting] [LACM, MCZ]. imparting a rugose-areolate appearance in places. Nontype localities. TRINIDAD: 1 queen, Sides of mesosoma predominantly longitudinally Nariva Swamp, 23-IV-1935, #140 (N. A. Weber) rugose, but with occasional cross-ridges making [MCZ]. BRAZIL: 3 queens, Mato Grosso State, elongate cells. Intervals in mesosoma Sinop, 55.37W 12.31S, X-1974 #12314 (Alvarenga), macrosculprure shiny, nearly smooth, especially #12576 (Alvarenga and Roppa) [MZSP]; 2 queens, on sides. Anterior and sometimes dorsal faces of Goias State, Jatai, XI-XII-1972, #8857, #8939 (F. M. nodes weakly sculptured. Nodes slightly dulled Oliveira) [MZSP]. by vestigial microsculprure. Head dorsum with erect-suberect hairs in

Rogeria lacertosa Kempf addition to the typical decumbent pilosity. Fig. 31 Color yellowish-brown; gaster slightly darker. Legs and, sometimes, antennae lighter, more yel- lacertosa 1963:194, 5-6. and Rogeria Kempf Figs Holotype lowish. Mandibles often slightly darker than head paratype workers, BRAZIL: Rio Grande do Sul State, capsule. Sinimbu (F. Plaumann) [MZSP] [All 4 paratypes exam- Discussion. —Because minensis has been ined; holotype not examined]. syn- onymized with gertnaini, some of Kempf's (1963) Additions to germaini-group diagnosis. WL list of characters that distinguish lacertosa are no 0.93-1. 05mm. with median notch. Clypeal apron longer valid, however, gertnaini and lacertosa are Metanotal weak to Petiolar keel groove strong. still distinguishable on the basis of a number of with lamellate carina. shaft has single Sting apex characters (see gertnaini discussion). Rogeria pellecta dorsal lancets acute; both weak. flange; (Figs. 32-33), also from southern Brazil, differs in not effaced on side of head below Macrosculprure petiole keel, sting and lancets, promesonotal sculp- Promesonotal and narrow like eye. rugae sharp ture, and pilosity. those on head. Sides of both nodes distinctly Distribution. —Rogeria lacertosa is known only macroareolate. Erect hairs on No decum- scapes. from the type material collected from 100-200m bent hair on little if on mesosoma gaster; any elevation in southern Brazil. No ecological data dorsum. are available. Workers.—TL 3.6-4.0, HL 0.83-0.91, HW 0.72- Material Examined. —BRAZIL: Rio Grande do Volume 3. 1 994 45

Sul Pardinho Sinimbu State, (F. Plaumann), (F. microsculpture vestigial, leaving nearly smooth, 4 workers intervals in Plaumann) [mouthparts, sting]. [MZSP]. shiny macrosculpture. Sides of peti- olar peduncle microareolate; nodes roughened by Rogeria pellecta Kempf obscure microsculpture and not as shiny as head Figs. 32-33 and mesosoma. Color yellowish-brown; gaster slightly darker. Rogeria pellecta Kempf 1963:191, Figs. 3-4. Holotype worker, Legs, and sometimes antennae, more BRAZIL: Santa Catanna State, Nova Teutonia (F. lighter, yel- lowish. Mandibles often darker than Plaumann) [MCZ, MZSP] [6 of 28 paratype workers slightly head 4 from examined, including holotype locality; holotype capsule. not examined] — Gynecoid Workers. As described by Kempf Diagnosis.—WL 0.90-0.99mm. Clypeus with (1963). Discussion. —In the southern median notch. Metanotal groove distinct. Brazil /Paraguay area occur related and Metapleural lobes prominent, triangular. Petiolar species germaini, lacertosa, sicaria. with these are found node arises gradually from peduncle. Sting and Comparisons species lancets in the "Species section and in the strong, acute; sting shaft with dorsal flange; Groups" germaini lancet and lacertosa discussions. with barbule. Laterodorsa of head longitu- Distribution. — is known dinally rugose. Promesonotum coarsely areolate Rogeria pellecta only from the 33 taken in southern to rugose-areolate. Back of petiolar node strongly type specimens Brazil from berlesate of leaf litter. areolate; postpetiole vestigially areolate. Scapes lack erect hair. Dorsa of head, mesosoma, nodes Material Examined.—BRAZIL: Santa Catarina and with erect State, Nova Teutonia (F. Plaumann) gaster and decumbent pilosity. [mouthparts, 6 workers, 1 Mandibles, palpal formula, eye, propodeal spines, sting]. gynecoid [MCZ, MZSP]. postpetiole, other aspects of petiole, sculpture, sicaria and pilosity as in germaini-group diagnosis. Rogeria Kempf 34 Workers.— TL 3.4-3.7, HL 0.81-0.89, HW 0.71- Fig.

0.78, SL 0.55-0.60, EL 0.09-0.11 (16-21 facets), PW

Rogeria sicaria Kempf 1962b:22, Figs. 20, 21 . worker, 0.50-0.55, WL 0.90-0.99, SpL 0.17-0.21, PetL 0.39- Holotype BRAZIL: Sao Paulo, Agudos [19km SSE Bauru] (W W. 0.46, PpetL 0.20-0.22mm, CI 0.87-0.88, OI 0.12- Kempf) [MZSP] [Holotype examined]. 0.14, SI 0.77-0.79, PSI 0.19-0.22. N=6 — The following supplements Kempf (1963). Diagnosis. WL 0.75mm. Eye very small. formula convex. Palpal 3,2. Anterior edge of clypeus weakly Clypeal apron Metanotal groove strong. emarginate. Eyes oval. Metanotal groove may be Propodeal spines very long, strongly inclined dor- sad bordered behind by a transverse costa. Posterior and divergent. Petiole long (PetL/PW 0.95), face of with keel. petiolar node vertical or slightly concave in prominent Sting apex strong, with dor- lateral view. sal Sides of Postpetiole widest anteriorly, as in flange. head below eye, mesosoma, or and node Fig. 32, evenly convex, as in Fig. 53. Sting petiolar areolate; coarsely so on like that of mesosoma. No erect hair apparatus inermis (Fig. 42), except for on scapes. Head, the sting (Fig. 33), which has larger anterolateral mesosoma dorsum, nodes, and gaster Tl with on a stiff, erect-suberect hairs and processes sting base, stronger, relatively shorter, more abun- thicker a sting shaft, and lower dorsal flange. dant, appressed-decumbent hair. Posterior head sculpture is rugose-areolate, Holotype Worker.—TL 2.8, HL 0.67, HW 0.56, with SL EL 0.03 rugae longitudinally diverging or transversely 0.48, (8 facets), PW 0.37, WL 0.75, SpL arching across back of head. Rugae of laterodorsa 0.22, PetL 0.35, PpetL 0.17mm, CI 0.83, OI 0.05, SI sometimes PSI 0.29. broken and with lateral spurs. Sides of 0.86, head areolate The around eye, but effaced and nearly following supplements Kempf (1962b). smooth ventrad. Dorsal face of propodeum trans- Mandible triangular, with 5 teeth decreasing in versely carinulate to coarsely areolate. Petiolar size to base. Clypeal apron evenly convex; body of node with areolate macrosculpture that becomes clypeus projecting over mandibles. Frontal lobes weaker and less defined narrow anteriorly. Ventral peti- (FLW/HW 0.29). Eyes small, circular, ole with a of somewhat in pair longitudinal carinae arising from sunken head. Petiolar peduncle long an anterior keel. Head and mesosoma (PetL/PW 0.95), with strong keel and dentate 4h Journal of Hymenoptera Research

face short face inferior process. Postpetiole highest in caudal half; longer anterior and posterior meet- dorsal view subtrapezoidal as in Fig. 49. ing at a narrow apex. Postpetiolar node from Posterior head areolate and more coarse than above as in utiguispina (Fig. 49). Postpetiolar ster- elsewhere on head. Intervals shiny, but dulled num short, weakly concave; anterior edge square. somewhat by vestigial microsculpture. Areolate Laterodorsa of head longitudinally rugose- sculpture on petiolar node not as well defined as areolate. Posterior head with rugae diverging from on head and mesosoma and even more effaced on midline or transversely arching; few if any cross postpetiolar node. Nodes roughened by vague ridges. Sides of head longitudinally rugose. Ante- microsculpture. rior pronotal disc transversely rugose to rugose- Color uniformly golden-brown. areolate; rest of promesonotal dorsum longitudi- Discussion. —This species is still known only nally rugose. Dorsal face of propodeum areolate- from the holotype. It seems related to germaini, rugose. Lateral mesosoma rugose; less regularly Brazil. pellecta, and lacertosa, also from southern so on meso- and metapleura. Microsculpture on See the "Species Groups" section for comparisons. head and mesosoma effaced; intervals quite smooth and shiny, especially on sides and back of head. Creightoni-Group and Related Species Petiole distinctly microareolate; sides of node vaguely rugose to vaguely areolate; posterior face Rogeria merenbergiana new species with strong ridges. Postpetiole effaced Figs. 46-48 microareolate and very sparsely and vaguely rug- ose.

Additions to creightoni-group diagnosis. WL Color very dark reddish brown to yellowish 0.69-0. 83mm. Palpal formula 3,2. Clypeal apron brown; appendages and often sides of mesosoma with median concavity. Eye small, elliptical; EL/ and gaster lighter. SpL 0.67-0.86. Anterior edge of pronotal disc with- Nontype Workers.— TL 2.6-3.0, HL 0.63-0.71, out a strong transverse carina. MHI 0.91-1.01. HW 0.53-0.60, SL 0.43-0.49, EL 0.08-0.09 (12-15 Promesonotum and metanotum with distinct pro- facets), PW 0.40-0.43, WL 0.69-0.77, SpL 0.10-0.13, files—promesonotum convex, metanotum usu- PetL 0.29-0.33, PpetL 0.16-0.18mm, CI 0.84-0.85, ally flat. Propodeal spines short, straight. OI 0.14-0.15, SI 0.79-0.82, PSI 0.14-0.17, MHI 0.95- Postpetiole from above subtrapezoidal. Mesosoma 1.01. N=3 and sides of head rugose (sides of head sometimes Little difference from types. Some slightly effaced rugose-areolate). Erect hair on scapes and smaller, but with slightly larger eye for size. Gen- middle and hind tibiae. erally higher mesosoma height index. Posterior Holotype and Paratype Workers. — TL 2.7-3.1 head sometimes transversely arching rugose-ar- (3.0), HL 0.69-0.71 (0.71), HW 0.57-0.62 (0.61), SL eolate. Sides of head rugose-areolate, but effaced 0.46-0.52 (0.50), EL 0.07-0.09 (0.08) (8-14 facets), posterior to eye. Promesonotal dorsum longitudi- PW 0.42-0.47 (0.45), WL 0.74-0.83 (0.80), SpL 0.10- nally rugose to rugose-areolate. Lateral mesosoma 0.14 (0.12), PetL 0.28-0.36 (0.335), PpetL 0.15-0.18 rugose-areolate. (0.16)mm, CI 0.83-0.88 (0.85), OI 0. 13-0. 15 (0.13), SI Paratype And Nontype Queens. —TL 3.3-3.7, HL 0.80-0.84 (0.83), PSI 0.13-0.17 (0.15), MHI 0.91-1.00 0.70-0.75," HW 0.61-0.64, SL 0.48-0.52, EL 0.16-0.17, (0.93). N=6 PW 0.58-0.60, WL 0.98-1.15, SpL 0.15-0.17, PetL Mandibles with 6-7 teeth decreasing in size 0.34-0.39, PpetL 0. 19-O.20mm, CI 0.86-0.88, SI 0.78- basad and basal tooth as large or larger than 0.80, PSI 0.14-0.16, MHI 1.13-1.19. N=3 penultimate basal. If 6 teeth, may have 1-2 addi- Mandible with 6-7 teeth; if 6, basal larger than tional denticles among basal teeth. Body of clypeus penultimate tooth. Clypeus evenly convex to not strongly produced. Nuchal groove weak. Pos- weakly emarginate. Posterior outline of head terior outline of head evenly convex. Propodeal strongly convex. Mesosoma and waist habitus spiracle one diameter, or slightly less, from edge of shown in Fig. 48. Sculpture mostly like worker. infradental lamella. Metapleural lobes low; edges Anterior pronotum areolate, becoming more rug- perpendicular. Petiolar node with a more or less ose on sides. Half to nearly all of ventral half of distinct angle between dorsal and posterior faces. mesopleura smooth; the rest of meso- and Petiolar peduncle with weak keel. Postpetiole with metapleura and meso- and metanota longitudi- Volume 3, 1994 47 nally rugose-carinate. Dorsal face of propodeum Belen, Santa Leticia, Finca Merenberg, 2300m, 9- Intervals on mesosoma rotten in transversely rugose. nearly 13-1-1978, jog pasture (C. Kugler and J. smooth; shiny. Petiole and postpetiole more Hahn) [1 mouthparts; 4 stings] [MCZ]. strongly microareolate— and weakly areolate. Paratype localities. COLOMBIA: 13 workers, 1 Discussion. Rogeria unguispina (Fig. 49) from queen, holotype locality [BMNH, LACM, MCZ, the mountains of Venezuela (1100-2000m) is also MZSP, USNM]; 2 workers, Huila, Las Cuevas de very similar in size, sculpture, pilosity and color to los Guacharos National Park, 1900-2300m, near merenbergiana, and the two species intergrade with Palestina, about 20km S Pitalito, 15-17-1-1978 (C. to respect mesosoma shape and petiole shape. It is Kugler and J. Hahn) [whole specimen slide quite possible that these are geographic variants of mounted] [MCZ]. the same species, but all unguispina specimens Nontype localities. ECUADOR: 2 workers, presently available have more elongate eyes and Pichincha Province, 16km E Tandapi, 2000m, in have a carina across the strong running shoulders, litter, 20-VI-1975, B-302 (S. and J. Peck); 3 workers, and specimens dissected differ in palpal formula. 2 queens, Pichincha Province, 3km E Tandapi, Most also 1 litter in unguispina specimens have downcurved 300m, wet ravine, VI-1975 (S. and J. Peck); propodeal spines, concave posterior face of peti- 4 workers, Pichincha Province, 20-30km ENE olar node, and a broad, shallow metanotal groove. Alluriquin on Chiriboga Road, 1400-1800m, 1975, See also the for nevadensis discussion comparisons B-301 (S. and J. Peck) [CKC, MCZ]. with that species from high elevations in Colom- bia. Rogeria unguispina new species Rogeria alzatei occurs in Colombia, but so far Fig. 49 has not been found above 1000m. These ants are usually smaller than merenbergiana (WL 0.51- Additions to creightoni-group diagnosis. WL 0.68mm), lack erect hair on the scapes, and lack a 0.64-0.83mm. Palpal formula 2,2. Clypeal apron distinct metanotal profile. with median concavity. Eye moderately large, Rogeria creightoni, so far only known from elongate-oval. Strong carina across shoulders. North and Central America, is related to Metanotal groove broad and shallow. Propodeal but has: less merenbergiana , generally 1) longer, spines usually with downcurved tips. Postpetiole inclined propodeal spines, 2) less distinct metano- from above subtrapezoidal. Sculpture on side of tum, and 3) stronger, rugose-areolate sculpture on head vestigial, leaving a small or extensive smooth sides of head (Figs. 51-54, 95-96). area. Pronotal disc and sides with weakly undu- Rogeria belti from Central America is the same lating rugae and nearly smooth, shiny interrugal size and also generally has a more or less definite spaces. Erect-suberect hair on scapes, and middle step in mesosoma profile at metanotal groove, but and hind tibiae. has: 1) larger eye, 2) longer propodeal spines, 3) Holotype and Paratope Workers. — TL 2.4-3.1 more rounded postpetiolar node, 4) more areolate (3.0), HL 0.60-0.72 (0.71), HW 0.50-0.61 (0.60), SL pronotum, and 5) mesosoma dorsum without two 0.40-0.49 (0.47), EL 0.10-0.11 (0.11) (15-19 facets), distinct of types hairs. PW 0.36-0.46 (0.46), WL 0.64-0.83 (0.81), SpL 0.09- This is species named in honor of Gunther 0.13 (0.13), PetL 0.26-0.34 (0.34), PpetL 0.14-0.19 Buch his and family, who have heroically main- (0.19)mm, CI 0.82-0.86 (0.84), OI 0.12-0.16 (0.13), SI tained of their part farm, Finca Merenberg, as a 0.79-0.82 (0.78), PSI 0.14-0.16 (0.15). N=7 nature preserve and permitted the collection of Mandibles with 5-6 teeth, basal larger than these and other ants.— penultimate tooth; some with a denticle between Distribution. Rogeria merenbergiana is found basal and penultimate teeth. Body of clypeus not at high elevations (1300-2300m) in the Andes of projecting over apron. Nuchal groove makes a Ecuador and southern Colombia. It has been taken notch in side of head. Pronotum rounded on front in litter in moist broadleaf forest sides. samples and and Propodeal spiracle large (Fig. 49) to bamboo-moss forest. A nest was found in a rotten moderate; less than one-half diameter from near- in a log pasture. est edge of propodeum. Free edges of metapleural Material Examined. — Holotype locality. CO- lobes form an oblique angle. Posterior face of LOMBIA: 14 1 workers, queen, Huila, 12km W petiole concave in holotype and some paratypes. 48 Journal of Hymenoptera Research

Petiolar peduncle lacks a distinct keel. Postpetiolar [MCZ]; 1 worker, Aragua State, Rancho Grande, node highest in posterior half. Postpetiolar ster- 14-VIII-1967, 1100m (R. W. Poole) [MCZ]; 2 work- num with a short, flat profile and square anterior ers, Lara State, vicinity of Caspo, 18km S Sanare, corner. 1620m, forest rotten wood, 6-XII-1985 (J. Lattke Laterodorsa of head rugose, sometimes ar- and W. L. Brown) [MIZA]. eolate caudad (holotype). Posterior head rugose- areolate in a more or less transversely arching Rogeria brunnea Santschi pattern. Interstices on head vaguely roughened by Figs. 50, 93 effaced microsculpture; especially shiny on sides brunnea Santschi 1930:79. work- and back of head. Promesonotal dorsum longitu- Rogena curvipubens Syntype ers, CUBA: Pinar del Rio, Sierra de los Organos, Rangel dinally rugose in front, becoming more vermicu- (A. Bierig) [NMB] [Both syntypes examined]. N. comb. late with partial or complete cross-ridges behind. Rogena scabra Weber 1934:27, Fig. 2. Syntype workers, queen, Pronotal sides with few longitudinally rugose CUBA: Cienfuegos, Soledad, Limones Seboruco cross-ridges. Meso- and metapleura with broken, (Darlington and Weber) (MCZ) [Queen and 1 of 2 worker irregular macrosculpture. Dorsal face of syntypes examined], N. syn. caraiba Santschi 1936:198, 5. workers, propodeum transversely areolate. Interstices on Rogeria Fig. Syntype CUBA: Habana, Playa de Mananao [NMB] [All 8 disc and sides of pronotum shiny, nearly smooth; syntypes examined]. N. syn. rest of mesosoma microareolate and vaguely quite Rogena cubensis Santschi 1936: 199, Figs. 3-4. Holotype worker, to shiny. Petiolar node weakly vestigially CUBA: Pinar del Rio, Sierra del Rosano (A. Bierig) N. microareolate; posterior face of node with longitu- [NMB] [Holotype examined]. syn. Rogena cubensis liabanica Santschi 1936:200. worker, dinal carinulae. Postpetiolar node nearly covered Holotype CUBA: Playa de Marianao (Bierig) [NMB] [Holotype with vague microsculpture to nearly smooth; al- examined]. N. syn. ways weaker than on petiole. Color blackish-brown with yellowish-brown man- Additions to creightoni-group diagnosis. WL dibles, flagellum of antennae, legs, and terminal 0.61-0. 91mm. Clypeal apron convex. Eye small, segments of gaster. elliptical. Nuchal groove weakly visible in lateral Discussion. —Rogena nevadensis is a similar ant view. MHI 0.84-0.99. Pronotal shoulders well from the mountains of northeastern Colombia, rounded. Propodeal spines straight, inclined. but nevadensis: 1) is usually smaller, 2) has smaller Propodeal spiracle less than 1 diameter from edge eyes, 3) has straight propodeal spines, 4) has more of infradental lamella. Postpetiole long (PpetL/ effaced sculpture, and 5) lacks erect hair on the PW 0.42-0.49), subtrapezoidal in dorsal view; ster- scapes and tibiae. See also the merenbergiana dis- num with long, flat profile and receding anterior cussion. edge (except Dominican Republic). Both macro- The name unguispina refers to the claw-like and microsculpture weak overall (stronger in shape of the propodeal— spines. Dominican Republic). Rugae fine, low on head Distribution. So far, unguispina is known only and mesosoma (Fig. 93); vestigial on petiole. Erect from Venezuelan rain forest between 200m and hair sometimes sparse and only slightly longer 2000m elevation. — than decumbent hair. Scapes have erect hairs; Material Examined. Holotype locality. VEN- extensor surfaces of legs do not. EZUELA: Miranda state, Rio Capaya at Quebrada Workers.— TL 2.3-3.3, HL 0.57-0.80, HW 0.45- El Bagre, near Salmeron, E of Caracas, 200m, nest 0.70, SL 0.36-0.52, EL 0.06-0.09 (9-18 facets), PW beneath log on forest floor in limestone gorge, 8-II- 0.35-0.49, WL 0.61-0.91, SpL 0.09-0.15, PetL 0.25- #238 1982, (J- Lattke) [MCZ]. 0.34, PpetL 0.15-0.22mm, CI 0.79-0.87, OI 0.12- Paratype localities. VENEZUELA: 9 workers, 0.16, SI 0.73-0.89, PSI 0.15-0.17, MHI 0.84-0.94. holotype locality [mouthparts, sting] [BMNH, N=16 CKC, LACM, MCZ, MZSP, USNM]; 1 worker, Mandibles with 5-7 teeth; basal tooth equal to Falcon State, Sierra de San Luis [mountain range or slightly larger than penultimate basal; a den- just south of Paraguana Peninsula], Haitoncitos, ticle may occur between basal teeth. Body of near Curimagua, 1180m, primary forest, 3-VIII- clypeus not projecting over apron. Posterior out- #271 1 1982, (J- Lattke) [MCZ]; worker, Aragua line of head convex to weakly concave. Mesosoma Rancho 4563a with State, Grande, 15-VI-1960, (J. qasti) broadly rounded shoulders; metanotal groove Volume 3. 1 994 49 a weak impression or absent; no ridge at anterior on the petiolar peduncle in Cuban specimen. of propodeum. Propodeal spines narrow, moder- PpetW/PpetL 1.05-1.14. Sculpture much as in ately long; a bisecting line passes dorsal to workers. Macrosculpture on sides of head weak. propodeal spiracle and through or below Most of ventral half of mesopleura smooth. Pilos- anteroventral corner of and color as in workers. pronorum. Metapleural ity — lobes low, angular. Petiolar peduncle with strong Discussion. Workers of this species have two keel; node large and somewhat bulbous. disjunct sizes. In the smaller size range (WL 0.62- Postpetiolar node with long, low profile and 0.70mm) are three specimens from Pinar del Rio weakly defined posterior and anterior faces (more Province (including brunnea syntypes), one from distinct in Dominican Republic). Soledad (scabra syntype), one from I. Pinas, three Laterodorsa and sides of head rugose-areolate; from Dominican Republic, and one from Baha- in most Cuban specimens ridges become very mas. The larger workers (WL 0.80-0. 91mm) are weak on sides, even absent in spots. Posterior from Sierra del Rosario (cubensis holotype), head with transversely arching rugose-areolate Soledad, Playa Marianao (habanica holotype), Si- macrosculp-rure. Anterior pronotal disc trans- erra Maestra, and Sierra Anale. Aside from size, versely rugose to rugose-areolate; rest of prono- the smaller workers show little consistent differ- rum longitudinally rugose, usually with incom- ence from the larger workers, and in ways that plete cross-ridges between rugae. Mesonotum often vary within Rogeria species: 1) slightly more more rugose-areolate. Pronotal sides basically rug- macrosculpture on petiolar node, and 2) slightly ose (areolate-rugose in Dominican Republic); more extensive microareolate sculpture on sides meso- and metapleura confused rugose-areolate of mesosoma. The specimens from Dominican with some smooth patches. Dorsal face of Republic differ from the rest in having a more propodeum transversely rugose. Vague rugosi- compact mesosoma, more distinct postpetiolar ties on sides and /or posterior petiolar node; node and sternum in side view, and slightly stron- postpetiolar node smooth. ger sculpture. Colorbrown to yellowish-brown, with lighter, Workers of creightoni from San Jose, Costa more yellowish appendages and frontoclypeal Rica have the same size as larger brunnea and region. much the same pilosity, nuchal groove, and shape Queens.—TL 2.8-3.4, HL 0.59-0.74, HW 0.52- of mesosoma and petiole. Rogeria creightoni speci- 0.65, SL 0.39-0.50, EL 0.11-0.15, PW 0.45-0.52, WL mens from Yucatan and Chiapas have size, eye 0.81-0.95, SpL 0.10-0.16, PetL0.26 (estimated)-0.40, size, postpetiole, and pilosity like the larger brunnea, PpetL 0.17-0.22mm, CI 0.86-0.89, SI 0.74-0.78, PSI and the Yucatan specimen has similar mesosoma 0.13-0.16, MHI (Cuban only; could not measure in shape. But brunnea workers can be distinguished Bahamian) 0.97-1.06. N=3 by the combination of characters in the diagnosis. Apparently collected as strays, so not defi- Material Examined.—CUBA: Pinar del Rio Prov- nitely associated with workers described above. ince, Sierra de los Organos, Sierra del Rosario The Bahamian queen distinctly smaller (all of the (Bierig), Las Acostas (E. O. Wilson), San Vincente smallest measurements above); Cuban queens (E. O. Wilson); Habana Province, Playa Marianao identical in size. Mandible with 6 nearly teeth; (Bierig); Cienfuegos Province, Soledad (P. J. basal and penultimate tooth subequal in size. Darlington; N. A. Weber; E. O. Wilson; D. Bates Clypeus evenly convex. Posterior outline of head and G. Fairchild); Las Villas Province, Caibanen concave. medially Mesosoma of Cuban queens (E. O. Wilson); Sierra Anale (Bierig); Sierra Maestra,

like that of scobinata I. robust, shaped (Fig. 62), but Uvero (L. Armas). Pinas, S. Casas (L. B. Zayas). larger and with more prominent, angular DOMINICAN REPUBLIC: 16km ENE Pedemales metapleural lobes. Mesosoma of Bahamian queen (P. S. Ward). BAHAMAS: Andros Island, Nassau seems like that of Cuban queens, but is obscured Island (W. M. Wheeler). 20 workers, 3 queens by legs. Waist as in workers, except for lack of keel [LACM, MCZ, NMB]. 50 Journal of Hymenoptera Research

Rogeria creightoni Snelling areolate; posterior head transversely arched ar- Figs. 51-54, 95-96 eolate-rugose. Pronotal sides longitudinally rug- ose to areolate-rugose; meso- and metapleura con- 1. and Rogeria creightoni Snelling 1973:2, Fig. Holotype fused rugose (Fig. 95). Dorsal face of propodeum worker, UNITED STATES: Texas, Cameron paratype marked anteriorly by a distinct transverse carina, County, La Fena (W. S. Creighton) [LACM] (Holotype then areolate and /or transversely rugose. Head and paratype examined]. and mesosoma microsculpture obscure; intervals Additions to creightoni-group diagnosis. WL in macrosculpture moderately shiny. Top and sides 0.63-0. 93mm. Clypeal apron usually with weak to of petiolar node effaced microareolate with very distinct median notch. Eyes small, oval to circular weak, indistinct macrosculpture; posterior face (EL/SpL 0.29-0.64). Metanotal groove weak or with weak longitudinal rugulae. Postpetiolar node absent. MHI usually 0.90-1.06. Propodeal spines with vague microsculpture on sides; smooth dor- long (PSI usually > 0.17), often nearly horizontal. sally. Postpetiolar node from above usually Color brownish-yellow to brownish-yellow; subtrapezoidal, sometimes subrectangular; ante- dorsa of head and gaster slightly darker, append- rior not Sides of head edge of sternum prominent. ages lighter.— distinctly rugose-areolate. Pronotal disc trans- Queen. Uncertain;— see discussion. versely rugose-areolate on anterior edge and lon- Discussion. This species becomes quite het- gitudinally rugose-areolate behind (Fig. 96). Scapes erogeneous with this revision. Ants from almost and extensor surfaces of tibiae with erect-suberect every locality are different from the others in some hairs. conspicuous way. Specimens from Belize have the Workers.— TL 2.4-3.5, HL 0.58-0.81, HW 0.51- longest propodeal spines (PSI 0.24-0.28), more 0.72, SL 0.38-0.55, EL 0.05-0.10 (8-16 facets), PW prominent metapleural lobes, weak or absent peti- 0.37-0.52, WL 0.63-0.93, SpL 0.11-0.20, PetL 0.27- olar keel and thicker ridges in macrosculpture. 0.40, PpetL 0.14-0.22mm, CI 0.85-0.91, OI 0.08- The Texas specimens have similar habitus, but 0.17, SI 0.68-0.80, PSI 0.16-0.28, MHI 0.87-1.08. have somewhat shorter propodeal spines (PSI 0.21- N=21 0.23), less prominent metapleural lobes, more dis- Mandibles with 5-7 teeth, if 5, may have 1-2 tinct petiolar keel and less thickened additional basal denticles. Basal tooth subequal or macrosculpture. A Tamaulipas specimen is simi- larger than penultimate basal. Palpal formula 3,2. lar to the Texas specimens, but has longer, dis- Body of clypeus projecting to anterior edge of tinctly inclined propodeal spines. Specimens from or Nuchal to are the apron beyond. groove weak strong. Yucatan, and Chiapas (Fig. 52) largest ( WL Posterior margin of head weakly concave to con- 0.78-0. 93mm), have a prominent clypeal body, vex. Mesosoma habitus variable (Figs. 51-53). somewhat longer scapes, small circular eye, strong Propodeal spines moderately long and angled to nuchal groove, relatively shorter propodeal spines very long and nearly horizontal, a bisecting line (PSI 0.19-0.22), and head macrosculpture with usually passes well above anteroventral corner of sharp ridges and unusually large areolae. Speci- pronotum. Propodeal spiracle not especially large mens from La Selva, Costa Rica are similar, but or prominent; located < 1/2 diameter to almost have normal clypeus and sculpture. Others from one diameter from edge of infradental lamella. Costa Rica (Fig. 53) have shorter, more inclined Metapleural lobes moderately to very prominent; propodeal spines (PSI 0.16-0.20), a more distinct corner broadly to narrowly rounded. Petiolar node petiolar node and keel, and subrectangular more or less distinct (Figs. 51, 53). Petiolar keel postpetiolar node. One specimen from Oaxaca is absent to moderately well developed. Sting appa- more like those from Belize; another is more like ratus of specimens from four localities like that of the San Jose specimens. With little material from inermis (Fig. 42), except for less angular most localities, I did not feel confident naming a anteroventral corner of spiracular plate, some- new species for each variant, but as collections what lower valve chamber height, and in Belize improve, I will not be surprised if this species and Oaxaca specimens the sting shaft is higher undergoes fission. and slightly upturned. At one time during the revision I also consid- Laterodorsa of head longitudinally rugose to ered describing cornuta, innotabilis, leptonana, and Volume 3, 1 994 51

alzatei as variants of creightoni. But now I believe (A. Newton); Yucatan, Chichen Itza [MCZ]. they can delineated. Comparison of creightoni with BELIZE: Intercepted in Brownsville, Texas on or- is found in chids nr. leptonana key couplet 40; comparisons (Heinrich); Belmopan (S. & J. Peck); Caves with conuita and in their discus- Branch alzatei, respective (S. & J. Peck) [mouthparts, sting] [BMNH, sions. All known specimens of innotabilis fall within CKC, MCZ, MZSP, USNM]. COSTA RICA: the geographic range of creightoni and the two Heredia Province, La Selva (Talbot & species are sympatric in at least one locality. All VanDevender) [sting] [LACM], (D. M. Olson) specimens of creightoni differ from innotabilis (Figs. [DMOC]; San Jose (H. Schmidt) [MZSP]; in several erect hair on 55-57, 97-98) ways: 1) Jurrucarres (A. Bierig) [MZSP]; no locality (F. scapes and tibiae, 2) palpal formula 3,2, 3) clypeal Nevermann) [MZSP]. 32 workers. apron with weak to distinct median notch, and 4) postpetiolar sternum not prominent. The regional Rogeria innotabilis new species variants differ in additional, but inconsistent ways. Figs. 55-57, 97-98 A single queen from Trinidad (N. A. Weber #129) is very much like the three Costa Rican and Additions to creightoni-group diagnosis. WL Colombian queens provisionally assigned to 0.66-0. 73mm. Mandibles triangular to slightly innotabilis (Fig. 57) in size, general habitus, trian- subtriangular. Palpal formula 2,2. Clypeal apron gular mandibles, convex clypeus, robust mesosoma convex or slightly flattened medially. Eye small, size (MHI 1.28), propodeal spine and shape, and elliptical. Nuchal groove clearly visible in lateral subrectangular postpetiolar node, but differs in view, but not notching ventral outline of head. erect hair on and little or having scapes tibiae, no Mesosoma with nearly flat dorsal profile; a strong keel on petiolar peduncle, undulate ventral profile carina marks anterior edge of propodeum; MHI of and postpetiole, longitudinally oriented sculp- 0.90-1.00. Propodeal spines straight, wide. Peti- ture on the posterior head. On the basis of the olar keel moderate to large. Postpetiole pilosity, I provisionally assign the Trinidad queen subrectangular in dorsal view; sternum weakly to to but its creightoni, because of great similarity in strongly prominent. Spiracular plate of sting ap- other respects to the innotabilis-like queens, and paratus with long spine on posteroventral corner. because neither set of queens is from an area Sides of head areolate-rugose; mesosoma predomi- where workers of creightoni or innotabilis have nantly rugose; microareolate sculpture distinct on been I collected, do not feel confident of these mesosoma sides. No erect hair on scapes or exten- assignments. sor surfaces of tibiae. See also discussions of alzatei, belti, brunnea, Holotype and Paratope Workers. — TL 2.4-2.7 and cornuta. (2.7), HL b.60-0.65 (0.65), HW 0.50-0.57 (0.54), SL Distribution. — Rogeria creightoni ranges from 0.40-0.45 (0.43), EL 0.06-0.08 (0.08) (7-13 facets), southern Texas to Costa Rica; from about 1,500m PW 0.37-0.41 (0.41), WL 0.66-0.73 (0.72), SpL 0.10- in of and Rica parts Chiapas Costa to sea level. 0.15 (0.13), PetL 0.26-0.30 (0.29), PpetL 0.14-0.16 Types were collected in a residential area formerly (0.16)mm, CI 0.83-0.87 (0.84), 010.11-0.15(0.15), SI mesquite-acacia savannah. Other specimens come 0.76-0.80 (0.78), PSI 0.15-0.20 (0.18), MHI 0.90-1.00 from riparian woodland, palm-thorn forest, rain (0.98) N=10 forest, pine- forest, cecropia forest, and cacao Mandible with 5-7 teeth. If 5 or 6 teeth, it may plantation. Most were taken in leaf litter on the have 1-3 denticles among proximal teeth. If 7 teeth, Some Belize ground. specimens were under a 4 proximals small. Basal tooth equal to or larger termite another in orchids. nest; than penultimate tooth. Body of clypeus not pro- Material Examined.—UNITED STATES: to Texas, jecting anterior edge of apron. Eye small, ellip- Cameron S. County (W. Creighton) [LACM]; Live tical. Posterior outline of head strongly to weakly Oak S. County (P. Ward) [mouthparts, sting] convex. Metanotal groove absent or weak. MEXICO: [MCZ]. Tamaulipas, Antiguo Morelos Propodeal spines wide, straight; a bisecting line & Vera (S. J. Peck); Cruz, Pueblo Nuevo (E. O. passes through or below anteroventral corner of Wilson); Oaxaca, 1 mi. E Reforma, nr. Tuxtepec (A. pronotum. Petiolar keel weak or absent. Petiolar Newton) [mouthparts, sting]; 9 mi. E El Cameron node evenly rounded from front to back, or apex 12 mi. (A. Newton); Chiapas, NW Ocozocoautla somewhat flattened (Fig. 55). Postpetiolar node 52 Journal of Hymenoptera Research — highest in posterior half; node in dorsal view Distribution. If the queens truly belong to this from to much as in Fig. 53. Sting apparatus like that of innotabilis, species ranges Chiapas inermis (Fig. 42), except for spiracular plate shape northern Colombia, but workers have not been collected south of Workers from (Fig. 56). Nicaragua. come Head macrosculpture areolate on laterodorsa; moist forest litter at 1000-1200m elevation. Collec- areolate to transversely arching rugose-areolate tion sites for queens range—from 10-1200m. on posterior head. Areolae on posterior head are Material Examined. Holotype locality. larger and ridges sharper than elsewhere on head. MEXICO: Chiapas State, 12 mi. NW Ocozocoautla, Head dorsum slightly dulled by vestigial 3200 ft., 4-5-IX-1973, forest leaf litter (A. Newton) microsculpture; intervals smooth and shiny on [MCZ], sides and back. Anterior pronotal disc with 1-2 Paratype Localities. MEXICO: 7 workers, ho- transverse rugae; rest of promesonotal dorsum lotype locality [mouthparts, sting, 1 coated for longitudinally rugose with lateral spurs (Fig. 98) SEM] [BMNH, MCZ, MZSP]; 10 workers, Chiapas, to rugose-areolate. Lateral pronotum areolate; 6-XII-1951, #51 11574 (Cary) [2 mouthparts, stings] meso- and metapleura mixed rugose and rugose- [USNM], NICARAGUA: 2 workers, Km 139 near areolate. Anterior edge of propodeum marked by Matagalpa, Hotel Selva Negra, ca. 1200m, 18-VI- forest leaf litter weak to strong transverse carina; dorsal face 1978, (C. Kugler & J. Hahn) [mouth- crossed by one or more transverse, sometimes parts, sting] [LACM, MCZ]. branching, rugulae. Dorsum with vestigial Nontype localities. COSTA RICA: Puntarenas microsculpture, but ventral sides quite distinctly Province, Monteverde, 1200m, 23-V-1979, #3496 microareolate (Fig. 97). Macrosculpture on sides (P. Ward) [MCZ], COLOMBIA: Magdalena De- and back of petiolar node very weakly areolate; partment, Parque Tayrona, 210m, l-X-76 (C. microsculpture vague and effaced. Postpetiolar Kugler) [MCZ]; Guajira, Don Diego, 25-50m, 22- node shiny, almost completely smooth. VI-76 (W. L. Brown & C. Kugler) [queen mouth- Color golden brown with yellowish append- parts, sting] [MCZ]. 3 queens. ages to brown with slightly lighter frontoclypeal area, mandibles, mesosoma sides, gaster, and legs. Rogeria alzatei new species Possible 0.59-0.61 99 Queens.—TL 2.6-2.9, HL , HW Figs. 58-60, SL EL PW 0.52-0.54, 0.40-0.41, 0.11-0.13, 0.44-0.50, — WL 0.72-0.83, SpL 0.13-0.14, PetL 0.28-0.31, PpetL Diagnosis. WL 0.51-0. 70mm, most < 0.65mm. 0.14-0.17mm, CI 0.88-0.89, OI 0.21-0.24, SI 0.75- Mandibles subtriangular (usually) to triangular. 0.77, PSI 0.16-19, MHI 1.15-1.22. N=3 Clypeus usually truncate. Eye usually moderately Collected outside known range of innotabilis large, oval (10-14 facets, OI 0.17-0.20 in types), but workers, but have innotabilis traits: shapes of occasionally half as large and elliptical. Nuchal clypeus, nuchal groove, and postpetiole, and lack groove weak to strong. Mesosoma compact (MHI of erect /suberect hair on scapes and tibiae. 0.90-1.07). Propodeal spines usually slender, in- Differing from the worker by normal caste clined. Petiolar peduncle with little or no keel. differences and the following. Parapsidal furrows Postpetiolar node from above usually indistinguishable from furrows in sculpture. subrectangular; anterior edge of sternum not Pronotum with 2-3 transverse rugae mesad, be- prominent. Macrosculpture areolate on head sides. to coming longitudinally rugose rugose-areolate Posterior head transversely rugose-areolate (Fig. on mesosoma sides; anterior katepisternum 99). Pronotal disc rugose to rugose-areolate. No smooth. Mesonotum longitudinally rugose. Ve- erect hair on scapes or legs. nation identical to that of belti and Workers. — TL 1.9-2.2 nearly — (Fig. 37). Holotype Paratype Discussion. For comparison with related spe- (2.2), HL 0.51-0.56 (0.54), HW 0.45-0.48 (0.47), SL cies, see creightoni, leptonana, and alzatei discus- 0.32-0.36 (0.35), EL 0.07-0.09 (0.09) (11-14 facets), sions. Remarks about the queens are found in the PW 0.32-0.37 (0.35), WL 0.51-0.61 (0.59), SpL 0.10- creightoni and curvipubens discussions. 0.12 (0.11), PetL 0.21-0.26 (0.23), PpetL 0.12-0.14 The name innotabilis (L., not remarkable) re- (0.13)mm, CI 0.87-0.88 (0.87), OI 0.17-0.20 (0.19), SI flects my inability to identify any one salient fea- 0.72-0.74 (0.74), PSI 0.19-0.21 (0.19), MHI 1.01-1.07 ture on which to base the name for this species. (1.05). N=ll Volume 3. 1 994 53

basal teeth Mandible subtriangular; 5-6 teeth, two vex clypeus, a distinct metanotal groove as in Fig. small, subequal. Palpal formula 2,2. Clypeal apron 46, and subtrapezoidal postpetiolar node as in Fig. truncate, with distinct corners; body of clypeus 51. does not project to edge of apron. Posterior outline Specimens from Peru, Brazil, and Paraguay of head continuously convex. Nuchal groove in- extend the description somewhat differently: Pal- distinct in lateral view. Mesosoma's evenly con- pal formula 3,2. Clypeal apron usually truncate vex dorsal profile ends abruptly fore and aft by with rounded corners; occasionally convex. Sev- rather sharp angles and by transverse sculpture. eral specimens with a somewhat prominent clypeal or absent. Metanotal groove weak Propodeal spines body (Fig. 59). Generally with smaller, elliptical straight, narrow; a bisecting line passes through or eyes (OI of most 0.10-0.16), but the four workers just above anteroventral corner of pronotum. Cor- from the Puerto Maldonado vicinity of Peru pos- of lobes a ner metapleural narrowly rounded right sess both the smallest eyes (Fig. 59) and largest angle. Petiolar node evenly rounded front to back. eyes of the species (4 and 21 facets, OI 0.08 and Pygidium in the five specimens dissected with no 0.20). Nuchal groove forms a distinct notch in the piligerous tubercles. head of one Peruvian worker (Fig. 59). Mesosoma Head laterodorsa rugose-areolate to areolate; generally less compact (MHI 0.93-1.03). Petiolar sides more coarsely areolate. Microsculpture ves- keel weak to distinct (Fig. 59). Some with tigial, producing irregular intervals on head dor- postpetiolar node slightly wider in anterior half and intervals sum nearly smooth on sides and and shape similar to Fig. 66. Posterior pygidium posterior (Fig. 99). Anterior edge of pronotal disc with small tubercles in a Paraguayan specimen with two transverse rugae. Promesonotal dorsum dissected. Sting apparatus with reduced longitudinally rugose with numerous lateral spurs anterodorsal corner of quadrate plate in a Brazil- that sometimes connect, forming areolae. No ian worker. macrosculpture on dorsal face of propodeum. Paratype and Nontype Queens. —TL 2.4-2.8, HL Lateral pronotum longitudinally rugose-areolate. 0.56-0.60,"hW 0.51-0.53, SL 0.35-0.38, EL 0.13-0.14, Meso- and metapleura with scattered, irregular PW 0.44-0.49, WL 0.70-0.79, SpL 0.14-0.15, PetL rugae. Vestigial rugae on sides and back of peti- 0.26-0.30, PpetL 0.15-0. 1 6mm, CI 0.88-0.92, OI 0.25- olar node. Vague microsculpture makes rough- 0.26, SI 0.70-0.72, PSIO.19-0.20, MHI 1.17-1. 18. N=2 ened intervals on mesosoma, petiole and much of Both collected in the same localities as alzatei postpetiole; apex of postpetiolar node smooth. workers (BCI, Panama; Quebrada Susumuco, Co- Body yellowish-brown to reddish-brown with lombia), but not in nest series. They most resemble slightly darker head and gaster dorsa (black to alzatei workers in characteristic shapes of clypeus, naked eye); legs and antennae lighter, more yel- mandible, mesosoma, propodeal spines, petiole lowish or yellowish-brown. and postpetiole, as well as sculpture and pilosity. Nontype Workers.— TL 1.9-2.5, HL 0.50-0.66, Mesoscutellum projects slightly beyond posterior HW 0.42-0.56, SL 0.31-0.46, EL 0.05-0.10 (4-21 fac- edge of metanotum. Posterior head rugose-ar- ets), PW 0.29-0.39, WL 0.51-0.70, SpL 0.09-0.13, eolate; no tubercles. Middorsum of head covered PetL 0.21-0.28, PpetL 0.11-0.15mm, CI 0.81-0.88, with erect hairs. OI 0.08-0.20, SI 0.71-0.85, PSI O.15-0.21, MHI 0.90- Discussion. —The extent of variation within 1.07. N=32 this species makes me question whether this is a Specimens from Guyana, French Guiana, and single species, but external characters that can be the Caribbean coast of Colombia extend the de- used in a key are either not marked enough to scription as follows: Mandibles triangular with 3- discount individual variation or not concordant. 4 teeth 4 basal plus denticles. Clypeal apron evenly Further complicating the picture is the surprising convex, sometimes with medial emargination. amount of variation within four specimens from with Some only 7-8 facets in the eyes. Some with the same region of Peru, and unique sets of fea- wider propodeal spines (Fig. 60) or slightly tures found in single specimens from northern downturned Petiolar tips. keel sometimes distinct Colombia, Rio de Janeiro, and Peru. Specimens (Fig. 60). Macrosculpture on head or pronotum from Dominican Republic strongly resemble more areolate. creightoni workers from the same localities, but Dominican Republic specimens have a con- differ in lacking erect hair on the scapes and tibiae 54 Journal of Hymenoptera Research and in having narrower propodeal spines. SW San Jose del Palmar, Rio Torito, Finca los Given the geographic variation within alzatei, Guaduales, 800m (C. Kugler) [mouthparts, sting] the species as a whole is difficult to distinguish [CKC, MCZ, MZSP]; Antioquia Department, clearly from creightoni, innotabilis, and leptonana. Providencia, Estacion Biologico, 600-800m, (C. But if one compares only those specimens oialzatei Kugler) [MCZ]; Cundinamarca Department, that are in sympatry or parapatry with those three Bogota to Villavicencio, Km 79 (W. L. Brown & I. species, the differences are clear. The Panamanian del Polania) [MCZ]; Meta Department, Bogota to alzatei workers differ from Central American Villavicencio, 23km NW Villavicencio, Quebrada in creightoni, innotabilis, and leptonana having: 1) Susumuco, 1000m (S. & J. Peck; C. Kugler) [1 subtriangular mandibles, 2) truncate clypeal apron, mouthparts, 2 stings, 1 whole specimen] [MCZ]. 3) larger, elongate-oval eye (OI 0.17-0.20), 4) more GUYANA (British Guiana): Oronque River (N. A. compact mesosoma (MHI 1.01-1.07), and 5) peti- Weber) [MCZ]. BRAZIL: Sao Paulo State, Agudos olar peduncle without a keel. They are also smaller (W. Kempf; C. Gilbert) [mouthparts, sting] [MZSP, than creightoni and innotabilis workers (WL 0.51- USNM]; Rio de Janeiro (T. Borgmeier) [MZSP]; 0.61mm vs. 0.66-0. 93mm in creightoni and 0.66- Federal District (C. A. C. Seabra) [MZSP]. PERU: 0.73mm in innotabilis). Additional differences from Madre de Dios Department, Puerto Maldonado leptonana are palpal formula (2,2 vs. 3,2 in leptonana), vicinity, 260m (C. Kugler) [CUIC, MCZ]. PARA- longer propodeal spines (PSI 0.19-0. 20 vs. < 0.16 in GUAY: Central Province (F. Baud, et al.) [mouth- leptonana), and lack of erect hair on scapes. parts, sting] [BMNH, MCZ (voucher), MHN]; See also the description of sibling species Concepcion, and Paraguari Provinces (F. Baud, et scobinata and the discussions of belli, curvipubens, al.) [BMNH, MHN]. 62 workers, 1 queen. micromma, tribrocca and merenbergiana. The species is named in honor of a courageous Rogeria scobinata new species Colombian intellectual— and friend, Isaac Alzate. Figs. 61-62, 100 Distribution. The range of alzatei is from — Panama to Paraguay, from western slopes of the Diagnosis. As inalzatei, except the following: Andes in Colombia to the Eastern coast of Brazil. Clypeal apron truncate (none emarginate). MHI All specimens were apparently taken as strays or 0.96-1.14. Posterior head with tuberculate in Berlese or Winkler sampling in moist forests macrosculprure (Fig. 100). Erect hair usually ab- from sea level to 1000m. — sent from head dorsum; if present, it is short and Material Examined. Holotype locality. usually confined to posterior margin. PANAMA: Canal Zone, Barro Colorado Island Holotype and Parati/pe Workers. — TL 1.9-2.5 (BCI), 2-VII-79, W. L. Brown [MCZ]. (2.2), HL 0.50-0.60 (0.55), HW 0.42-0.52 (0.475), SL Paratype localities. PANAMA: 3 workers, 1 0.31-0.40 (0.35), EL 0.06-0.09 (0.07) (10-14 facets), queen from holotype locality [mouthparts, sting, 1 PW 0.30-0.40 (0.35), WL 0.50-0.67 (0.585), SpL 0.08- worker coated for SEM] [CKC, MCZ]; 2 workers, 0.12 (0.095), PetL 0.19-0.26 (0.23), PpetL 0.12-0.16

BCI, IV-V-1942, #4953 (J- Zetek) [USNM]; 2 work- (0.14)mm, CI 0.82-0.86 (0.86), OI 0.14-0.18 (0.15), SI ers, BCI, VI-X-1943, #5105 (J. Zetek) [whole speci- 0.74-0.78 (0.74), PSI 0.16-0.19 (0.16), MHI 0.96-1.09 men slide mounted] [USNM]; 1 worker, BCI, I- (1.00). N=5 1960 (W. L. Brown & E. S. McCluskey) [MCZ]; 1 Nontype Workers.— TL 1.9-2.5, HL 0.50-0.61, worker, BCI, 7-III-1975, FP#10 (C. Toft & S. Levings) HW 0.44-0.52, SL 0.31-0.40, EL 0.06-0.09 (7-15 fac- [LACM]; 1 worker, Punta de los Chivos, W side ets), PW 0.30-0.40, WL 0.51-0.68, SpL 0.08-0.13, Gatun Lake, 3km SW Gahin, 100m, 3-9-VII-1979 PetL 0.19-0.28, PpetL 0.12-0.17mm, CI 0.84-0.88, (W. L. Brown) [MCZ]; 1 worker, Bocas del Toro, OI 0.14-0.18, SI 0.70-0.78, PSI 0.16-0.20, MHI 1.00- Pipeline Road, 300m, 8.53N 82.10W, 18-VII-1987, 1.14. N=32 #633 (D. M. Olson) [MZSP]. Mandibles subtriangular, 5-toothed (some- Nontype localities. DOMINICAN REPUBLIC: times with 1-2 additional basal denticles), decreas- 16km ENE Pedernales, 800m (P. S. Ward). CO- ing in size basad; basal tooth small. Palpal formula LOMBIA: Guajira, Rio Don Diego (W. L. Brown & 2,2. Median clypeus of some nontype workers C. Kugler); Magdalena Department, San Pedro, from Colombia like that of alzatei (Fig. 58), but type 550m Choco (P. Ward) [MCZ]; Department, 10km specimens with less prominent corners (Fig. 61) Volume 3, 1994 55 and other nontypes (Bolivia, some Brazil) have an the posterior rim and along the midline. almost evenly convex clypeal apron. Body of Color brown to golden brown. Legs and an- clypeus not projecting over clypeal apron. Poste- tennae generally lighter than rest of body; gaster rior outline of head weakly concave medially to sometimes darker. weakly convex. Nuchal groove clearly visible in Queens.—TL 2.4-2.6, HL 0.55-0.58, HW 0.46- side view. Eye oval to elliptical. Anterior and 0.52, SL 0.34-0.38, EL 0.11-0.13, PW 0.41-0.45, WL dorsal faces of pronotum may join smoothly, or in 0.68-0.74,SpL0.11-0.14,PetL0.25-0.27,PpetL0.14- a weak angle (Fig. 61). Metanotal groove broad, 0.16mm, CI 0.84-0.90, SI 0.70-0.75, PSI 0.16-0.19, slightly less to slightly more impressed than shown MHI 0.64-0.69. N=7 in a transverse carina. Fig. 61, bordered behind by Habitus shown in Fig. 62. Parapsidal furrows Propodeal spines inclined. Metapleural lobes indistinguishable from grooves in sculpture. An- moderately prominent; corner varies from sharply terior pronotum transversely rugose to rugose- angular (Ecuador, some Peru) to rounded as in areolate, becoming longitudinal on sides. Fig. 44 (some Paraguay). Ventral petiolar peduncle Mesoscutum with longitudinal, often diverging with a nonlamellate but some or usually weak, keel, rugae; mesoscutellum— rugose rugose-areolate. Ecuadorian specimens with distinct keel. Discussion. Rogeria alzatei is a sibling species Postpetiolar node in dorsal view subrectangular of scobinata, with which it is sympatric in Peru, as in Fig. 58. Pygidium with a pair of median, Brazil, and Paraguay, but can be distinguished by tubercles near columnar, piligerous caudal edge characters in the diagnosis and key. The pair of (barely visible in dissection microscope at 50X). columnar tubercles on the pygidium may also be Laterodorsa and sides of head densely ar- distinctive. eolate. Posterior head with short triangular, blunt The name scobinata, meaning having the na- tubercles in or less distinct more rows (Fig. 100). ture of a rasp, refers to the sculpture on the poste- Tuberculate sculpture usually extends across pos- rior head, which has rows of teeth like a rasp. terior — quarter of head, but in a few specimens from Distribution. Rogeria scobinata ranges from Leticia, Benjamin Constant, and Paraguay, the the north coast of South America to Paraguay at ridges between the tubercles are not always com- elevations below 1000m. All specimens were taken so the pletely lost, posterior head appears mostly as strays in tropical forest, mostly by Berlese or fragmented-rugose, with only a few of the trian- Winkler sampling of leaf litter, rotten wood, or gular tubercles. Interstices on most of head some- moss. what dulled indistinct areolate — by microsculpture, Material Examined. Holotype locality. PERU: but smoother and quite shiny between tubercles Madre de Dios Department, 3km N Puerto on back of head; sides sometimes rather strongly Maldonado, 260m, primary forest remnant by side microareolate. Anterior edge of pronotal disc with of road, berlesate of leaf litter and rotten wood, 13- 1-4 more or less transverse rugae. Rest of 16-VI-1981 (C. Kugler and R. R. Lambert) [MCZ]. promesonotum longitudinally rugose with fre- Paratype localities. PERU: 2 workers, 1 queen, quent incomplete lateral spurs. Mesosoma sides holotype locality [MCZ]; 22 workers, 2 queens, weakly and sparsely rugose to rugose-areolate, 5km E Puerto Maldonado on Rio Tambopata, Finca but more strongly microareolate than on pronotal Medina, 260m, primary forest berlesate, 13-16-VI- disc. Dorsal face of propodeum usually lacking 1981 (C. Kugler and R. R. Lambert) [mouthparts, but macrosculpture, rather strongly microareolate. sting, 1 worker coated for SEM] [BMNH, CKC, Rest of mesosoma with indistinct microareolate LACM, MCZ, MZSP, MHN, USNM]. Petiolar with sculpture. node broken vestigial Nontype localities. TRINIDAD: Nariva macrosculpture. Postpetiole without macrosculp- Swamp (N. A. Weber) [MCZ]. SURINAM: La ture; smooth on Sides of nodes with and nearly top. Poulle, Vank, Tambahredjo (J. van der Drift) weak microareolate that a sculpture imparts granu- [MZSP]. COLOMBIA: 7km N Leticia (S. and J. lar appearance; microsculpture usually weaker on Peck) [MCZ]. ECUADOR: Napo Province, Limoncocha postpetiole. and 20km S Tena (S. and J. Peck); Workers from Leticia have 8-10 erect hairs Pastaza Province, 25km N Puyo, (S. and J. Peck) along posterior rim of head and those from Ben- [MCZ]. PERU: Loreto Department, Ramon Castillo Constant have short erect hairs on jamin sparse, (S. and J. Peck), 15km WSW Yurimaguas (P. S. Sb Journal of Hymenoptera Research

Ward) [PSWC]; Pasco Department, near Pozuzo, Laterodorsa of head longitudinally rugose- (C. Kugler and R. R. Lambert); Madre de Dios areolate. Posterior head with transversely arching Department, Puerto Maldonado vicinity near Lake rugae. Intervals with weak microsculpture. Dor- Sandoval (C. Kugler and R. R. Lambert) [CKC, sal face of propodeum with 3 weak transverse CUIC, MCZ]. BOLIVIA: La Paz Department, rugulae. Mesosoma sides with sparse, indistinct, Caranavi (C. Kugler and R. R. Lambert) [MCZ]. longitudinal rugulae. Microsculpture on BRAZIL: Para State, (collector name unreadable); mesosoma even weaker than on head; interrugal Amazonas State, Benjamin Constant and vicinity spaces shiny, nearly smooth. Petiolar and (W. L. Brown); Mato Grosso State, Utiariti, Rio postpetiolar nodes nearly smooth, shiny. Saueruina [=R. Papagaio] (Lenko and Pereira); Mesosoma dorsum with eight pairs of erect Sao Paulo State, Piracununga Municipality, hairs (some matted down); nodes each with two Cachoeira das Emas (EEBP) (W. L. Brown); Rio de pairs of posterodorsally projecting hairs. Janeiro (T. Borgmeier) [MCZ, MZSP]. PARA- Color of body and appendages brownish-yel- GUAY: Alto dorsum of Tl darker. Parana, Amambay, Caaguazii, low; —gaster slightly Canendiyu, and Concepcion Provinces (F. Baud et Discussion. Boihciiri'ijntbens (Figs. 74-76, 101- alzatei also al.). 122 workers, 13 queens. 102) and (Figs. 58-60, 99), have been collected in Quebrada Susumuco or nearby Rogeria tribrocca new species Villavicencio environs. But they do not have a Fig. 63 three-toothed clypeal apron or a median spine on the third tergum of the gaster. Additions to creightoni-group diagnosis. WL The name tribrocca (brocca L., projection of 0.62mm. Eye small, elliptical. Clypeal apron with teeth) refers to the three tooth-like projections of three acute processes. Metapleural lobes very low, the clypeal apron. — broadly rounded. Postpetiolar node widest in an- Material Examined. Holotype locality. CO- terior half. Posterior edge of gaster T3 with me- LOMBIA: Meta Department, Quebrada Susumuko dian spine that seems to arise from a weak concav- [or Susumuco], 23km NW Villavicencio, 1000m,

ity. Sides of head smooth. Promesonotal dorsum forest litter, 5-III-1972, #B-234 (S. and J. Peck) with fine rugae diverging and branching from [MCZ]. anterior pronotum, becoming weak on meso- and metanota; interrugal spaces rather wide. No erect Rogeria carinata new species hair on scapes or legs. Fig. 64 Holotype Worker.— TL 2.3, HL 0.61, HW 0.51, Rogeria curvipubens: M. Smith 1937:851. Misidentification. SL 0.40, EL 0.05 (7 facets), PVV 0.36, WL 0.62, SpL PetL CI OI SI 0.10, 0.26, PpetL 0.13mm, 0.83, 0.10, Additions to creightoni-group diagnosis. WL MHI PSI 0.16. 0.78, 0.93, 053-0.67mm. Clypeal apron convex. Eye small. One mandible with six the other with teeth; Propodeal spines small. Mesosoma low (MHI 0.86- the basal two denticles penultimate replaced by 0.96). Postpetiole subrectangular in dorsal view; Nuchal forms a weak notch in (Fig. 63). groove sternum with distinct, but not projecting anterior lateral view of head. Pronotal shoulders seen from lip. Sides of head rugose. Promesonotal dorsum above 1 diameter subangular. Propodeal spiracles carinate, with broad, shiny intervals. Short erect- from nearest of infradental lamella. Ventral edge suberect hairs on scapes, but not on extensor sur- somewhat obscured but petiole by glue, probably faces of legs. with distinct keel and dentate Petiolar — process. Holotype and Parati/pe Workers. TL 2.0-2.5 node rounded in narrow; nearly symmetrically (2.2), HL 0.51-0.61 (0.55), HW 0.44-0.54 (0.46), SL lateral view. node with Postpetiolar broadly 0.35-0.42 (0.38), EL 0.05-0.06 (0.06) (7-10 facets), confluent anterior and dorsal faces and short pos- PW 0.31-0.41 (0.34), WL 0.53-0.67 (0.57), SpL 0.05- terior face. sternum with Postpetiolar perpendicu- 0.10 (0.08), PetL 0.19-0.24 (0.22), PetW 0.10-0.13 lar anterior corner and flat ventral profile (Fig. 63); (0.11), PpetL 0.11-0.14 (0.12), PpetW 0.14-0.18 node in dorsal view widest in anterior as in half, (0.15)mm, CI 0.86-0.89 (0.86), OI 0.11-0.12 (0.13), SI 74. shaft from Fig. Sting projecting gaster acute, 0.78-0.80 (0.83), PSI 0.09-0.15 (0.14), MHI 0.86-0.96 with low dorsal flange. (0.92). N=15 Volume 3. 1 994 57

Mandibles with 5 teeth and 0-2 denticles be- Rogeria nevadensis new species tween the two basal teeth; basal tooth larger than Fig. 65, 94 penultimate basal. Eyes oval to elliptical. Nuchal groove weak. Posterior outline of head weakly Additions to creightoni-group diagnosis. WL convex to weakly concave. Mesosoma from above 0.58-0.65mm. Clypeal apron emarginate. Eye small. broad, angular at shoulders, but strongly con- Shoulders rounded; promesonotum and propo- stricted behind; metanotum half as wide as shoul- deum meet without a strong angle. Postpetiolar ders. Propodeal spiracle small, one diameter from node widest in anterior half. Sides of head largely posterior edge of propodeum. Metapleural lobes smooth and shiny. Promesonotal dorsum longitu- small, rounded. Ventral petiole with weak keel. dinally rugose, with low, rounded ridges and no Postpetiole highest in posterior half. Sting appara- lateral spurs; interrugal spaces wide and nearly tus with slightly lower valve chamber than that of smooth. No erect hair on scapes or extensor sur- inermis (Fig. 42). faces of legs. of in Laterodorsa head rugose-areolate Puerto Holotype and Paratype Workers.—TL 2.2-2.5 (2.2), Rican specimens. Posterior head transversely HL 0.59-6.62 (0.59), HW 0.49-0.51 (0.49), SL 0.40- arched areolate-rugose. Interrugal spaces on head 0.42 (0.40), EL 0.06-0.08 (0.06) (7-10 facets), PW weakly and irregularly roughened on dorsum; 0.34-0.38 (0.34), WL 0.58-0.65 (0.58), SpL 0.09-0.10 smoother on sides and posterior. Anterior face of (0.09), PetL 0.22-0.25 (0.22), PpetL 0.12-0.14 pronotum with transverse carinules that continue (0.12)mm, CI 0.81-0.83 (0.82), 010.13-0.15(0.13), SI onto shoulders and curve dorsad. Meso- and 0.82, PSI 0.15, MHI 0.93-0.95 (0.93). N=3 metapleura with irregular longitudinal rugae; in- Mandibles with 5 teeth plus two denticles, or tervals smooth dorsally, but vestigially six teeth plus 1-2 denticles among basal teeth; microareolate below level of propodeal spiracle. basal tooth not larger than penultimate basal. Eye Dorsal face of propodeum mostly smooth with 0- elliptical. Nuchal groove weak. Promesonotum 3 transverse rugulae. Petiolar node vaguely slopes to join dorsal face of propodeum without an microareolate; macrosculpture absent or indis- angle or transverse carina; metanotal groove weak tinct on sides. Postpetiole vaguely microareolate or absent. Propodeal spines narrow; a bisecting on smooth venter; and shiny on dorsum. Gaster line passes through anteroventral corner of prono- smooth and shiny. tum. Propodeal spiracle within one diameter of Mesosoma dorsum with 10 pairs of erect hairs. edge of infradental lamella. Petiolar peduncle with Color brownish-yellow; head and mesosoma indistinct keel. Postpetiolar node with a short darker than rest of slightly body. Some Puerto vertical anterior face and long, sloping dorsal face; Rican reddish specimens slightly darker, with hue posterior face slightly concave in two paratypes. to head and mesosoma. Postpetiole widest in anterior half as in Fig. 66. Discussion. —The carinate macrosculpture of Postpetiolar sternum with flat ventral profile and the with promesonotum very shiny intervals serves somewhat prominent anterior edge. Sting shaft to differentiate carinata from others of the is like creightoni- projecting from apex that otinermis (Fig. 42). See also the group. discussion of nevadensis, which Laterodorsa of head rugose-areolate. Poste- has similar sculpture. rior head with transversely arching rugae. Head Material Examined.— BRIT- Holotype locality. shiny with effaced microsculpture, especially on ISH VIRGIN ISLANDS: Tortola, Long Bay, sides and back. Anterior pronotum crossed by berlesate from soil and 25-VII-1965 several leaves, (I. Proj. incomplete rugae. Lateral mesosoma (Fig. Staff) [MCZ]. 94) with little distinct macrosculpture and smooth, Paratype localities. 1 worker, holotype local- shiny intervals. Dorsal face of propodeum with PUERTO RICO: 3 7- ity [MCZ]. workers, Guanica, transverse rugae and effaced microareolate sculp- IV-1982, #75 (J. A. Torres) [LACM]; 1 worker, ture. Both nodes with weak, effaced microareolate #77 A. Cayo Ratones, 17-VII-1982, (J. Torres) sculpture and vague, vestigial rugose [LACM]; 8 workers, Cayo L. Pena, Humacao, 9-X- macrosculpture. 1982, #91 A. (J. Torres) [2 stings] [BMNH, CKC, Mesosoma dorsum with 10-11 pairs of erect LACM, MCZ, MZSP]; 1 worker, Ensenada, #1409, hairs; 2-3 pairs project posterodorsally from each (no collector) [USNM]. node. 58 Journal of Hymenoptera Research

Color brown to dark reddish-brown Rogeria leptonana new species (mesosoma and head dorsum may be darkest), Figs. 66-70 with brownish-yellow appendages and frontoclypeal area. Additions to creightoni-group diagnosis. WL Gynecoid Worker. —A nontype specimen is 0.54-0. 66mm. Median clypeal apron concave to circular. worker-like in most respects, but is larger (WL convex. Eye small, nearly Propodeal spines 0.70mm; PSI 0.19), has distinct mesonotal and small to absent. Mesosoma low, slender (MHI Petiole with metanotal sutures and partially formed wing at- 0.78-0.90; PW/WL 0.53-0.58). moder- tachments. Pronotum transversely rugose; ate to large keel. Postpetiolar node from above as mesoscutum longitudinally rugose. Sides of in Fig. 66. Anterior lip of postpetiolar sternum not Sides of macro- or postpetiolar node from above more evenly con- prominent. head distinctly vex. microsculptured. Mesosoma rugose, sometimes — microareolate sides. Erect hair Discussion. Rogeria merenbergiana (Figs. 46- with strongly on in and on extensor surfaces of tibiae. 48), which is also found at high elevations scapes usually — Colombia and Ecuador, has similar shapes of Holotype and Paratype Workers. TL 2.1-2.3 clypeus, mesosoma and propodeal spines, and (2.1), HL 0.50-0.55 (0.51), HW 0.43-0.46 (0.44), SL somewhat reduced sculpture, but: 1) are a little 0.32-0.34 (0.32), EL 0.05-0.07 (0.06) (7-10 facets), 0.57-0.61 0.07- larger than nevadensis (WL 0.69-0. 83mm), 2) have PW 0.32-0.36 (0.33), WL (0.57), SpL a metanotal impression, 3) have erect hair on 0.10 (0.07), PetL 0.20-0.23 (0.21), PpetL 0.12-0.13 second and third tibiae and scapes, and 4) are (0.13)mm, CI 0.83-0.86 (0.86), OI 0.11-0.15 (0.14), SI more distinctly sculptured, with sides of head 0.73-0.75 (0.73), PSI 0.12-0.16 (0.12), MHI 0.83-0.90 rugose. (0.83). N=4 5 teeth occa- Rogeria carinata (Fig. 64) from Puerto Rica and Mandibles triangular, with and Tortola have the same size, pilosity, shape of waist, sional denticle; basal tooth larger than penultimate and extensive shiny spaces on mesosoma, but: 1) tooth. Palpal formula 3,2. Median clypeal apron the metanotal-propodeal junction is more angu- concave. Body of clypeus projecting slightly over lar, 2) the sides of the head are distinctly rugose, 3) edge of apron. Posterior outline of head very the promesonotal dorsum has straighter parallel weakly concave or flat. Nuchal groove indistinct ridges and smoother intervals, and 4) the propodeal in lateral view. Mesosoma dorsal profile nearly spines are generally shorter (PSI 0.11-0.15). flat. Propodeal spines short with wide base; a Rogeria nevadensis takes its name from the bisecting line passes just above anteroventral cor- Sierra Nevada de Santa Marta, site of the type ner of pronotum. Spiracle about 1 diameter from localities. — nearest edge of infradental lamella. Petiolar node Material Examined. Holotype locality. CO- small. Postpetiole widest in anterior half (Fig. 66). LOMBIA: Magdalena Department, SE Santa Marta, Ventral profile of sternum concave; anterior lip Cuchilla San Lorenzo, vicinity of El Campano, not prominent. 1340m, rain forest, VI-1976 (W. L. Brown) [MCZ]. Head rugose-areolate to areolate on Paratype localities. 1 worker, same site as laterodorsa and sides, transversely areolate-rug- holotype [coated for SEM] [MCZ]. 1 worker, CO- ose on posterior head. Macrosculpture well de- LOMBIA: Magdalena, E Orihueca, 74.03W 10.55N, fined on posterior head, weaker on dorsum and San Pedro de la Sierra, 1300m, rotten log in coffee sides. Areolate microsculpture gives a granular plantation, 10-11-1977 (C. Kugler) [LACM]. appearance to dorsum and sides of head, where it Nontype locality. One gynecoid worker, largely obscures the weak macrosculpture; inter- berlesate at the San Pedro de Sierra site (C. Kugler) vals on posterior head nearly smooth. Anterior [MCZ]. 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- infradental lamella. Mesosoma low. Petiolar pe- pear granular. duncle with prominent lamellate keel; node short. Mesosoma dorsum with more than 12 pairs of Pygidium and sting apparatus as in workers. emergent erect hairs; nodes each with at least 2 Sculpture also as in workers. Microsculpture pairs of long posterodorsally projecting hairs. present but indistinct on head and mesosoma. Tibiae with erect hairs. Pronotum with 1-2 transverse rugae medially; Body yellowish-brown to brownish-yellow; laterally rugose. Meso- and metanota longitudi- frontoclypeal area and appendages lighter, more nally rugose, with nearly smooth, shiny intervals. yellowish. Meso- and metapleura longitudinally rugose, ex- Nontype Workers.— TL 2.2-2.5, HL 0.54-0.60, cept for smooth, shiny mesokatepisterna. Dorsal HW 0.43-0.50, SL 0.34-0.40, EL 0.04-0.08 (5-9 fac- face of propodeum smooth. Petiole, including ets), PW 0.33-0.35, WL 0.54-0.66, SpL 0.03 (no peduncle, and postpetiole nearly smooth. spines)-0.11, PetL 0.21-0.26, PpetL 0.12-0.16mm, There are differences, however. Palpal for- CI 0.80-0.83, OI 0.09-0.12, SI 0.77-0.83, PSI 0.05 (no mula 2,2 in the dissected Guajiran queen (vs. 3,2 in spines)-0.18, MHI 0.78-0.90. N=8 workers). One queen has a nearly subrectangular Type description extended as follows. Man- postpetiole. Neither has erect hair on the tibiae. dibles subtriangular in Dominican Republic speci- Males.—TL2.9,HL0.51-0.53,HW0.55-0.56,SL mens. Clypeal apron convex in the Chiriqui, 0.32-0.34, EL 0.25, PW 0.50-0.53, WL 0.91-0.95, Panama specimens and truncate in Chiapas and PetL 0.32-0.33, PpetL0.15-0.16mm, CI 1.04-1. 10, SI Dominican Republic. Body of clypeus not project- 0.58-0.61. N=4. ing beyond clypeal apron in Chiapas specimen. All four specimens from a nontype nest series, Basal tooth may equal penultimate basal in size. Cerro Pico Blanco, near San Jose, Costa Rica. Man- Scapes slightly longer (SI 0.77-0.83). The Ocosingo, dible with 4 teeth. Clypeal margin convex. Anten- Mexico specimen has slightly longer spines than nal flagellomeres 2-11 subequal in length and Cerro Pico Rica types (PSI 0.18); Blanco, Costa width; not twisted. Lateral habitus shown in Fig. have short specimens very spines (PSI 0.07-0.10) 69. May rian and parapsidal sutures present. Forew- those from Panama are devoid of armature venation as in (Fig. ing Fig. 30 hind wing as in Fig. 37. Short of 67). propodeal spines Costa Rican speci- Genitalia shown in Fig. 70. Mandibles smooth; mens are more inclined than in types; a bisecting clypeus nearly so. Head dorsum areolate; sides line the axilla. passes just below Postpetiole of behind eyes rugose; posterior head rugose-areolate. Chiapas and Pedernales, Dominican Republic Mesosoma longitudinally rugose on lateral prono- specimens narrower than others (PpetW/ PpetL tum and dorsal meso- and metapleura. Anterior 1.23-1.28 vs. 1.41-1.56); Pedernales postpetioles and ventral mesopleura smooth. Ventral are widest in half. posterior Microsculpture on metapleura diagonally rugose. Anterior portion head be less in may distinct than types. Panama- of mesonotum with vague effaced microsculpture; nian specimens lack erect hair on extensor sur- longitudinally rugose macrosculpture begins at faces of tibiae and have only 9 pairs of erect hairs level of wing and continues onto mesoscutellum. on the mesosoma dorsum. Propodeum areolate. Petiole microareolate, with a Possible Queens.—TL 2.6-2.8, HL 0.60-0.61, HW few weak, fine longitudinal rugae. Postpetiole and 0.52-0.53, SL 0.40, EL 0.12-0.13, PW 0.45, WL 0.73- gaster smooth. Abundant erect and decumbent 0.78, SpL 0.14-0.17, PetL 0.28-0.30, PpetL 0.14- pilosity on scapes, head, mesosoma, waist, and CI 0.16mm, 0.86-0.88, OI 0.20-0.23, SI 0.75-0.77, PSI gaster. Head brown, except for yellowish-brown 0.19-0.22, MHI 0.92-0.97. N=2 frontoclypeal area and yellow mandibles. Rest of Both collected queens on the north coast of body and appendages brownish-yellow; gaster Tl Colombia. Though workers of leptonana are un- and SI darker. known from South these — America, queens (Fig. 68) Discussion. The nontype specimens from the have strongest affinities to leptonana workers: high elevations of Costa Rica may belong to a Mandible with 5 teeth; basal as large as or larger distinct species defined by the complete absence than penultimate basal. Clypeal apron emargin- of propodeal spines, but presently they seem linked ate. Posterior outline of head flat. Propodeal spines to the types by the two specimens from Panama, wide. 2 Propodeal spiracle diameters from edge of which have intermediate spine length. 60 Journal of Hymenoptera Research

The different palpal formula of the workers parts, sting] [MCZ]; 1 queen, Magdalena Depart- and queens and the fact that workers and queens ment, Parque Nacional Tayrona, near Pueblito, specimens come from different countries cause 210m, ll-XI-1976 (C. Kugler) [MCZ]. DOMINI- doubt about whether the queens are correctly CAN REPUBLIC: 3 workers 16km ENE Pedernales, assigned. But most morphological characters fit 800m, 9-IX-1992 (P. S. Ward); 1 worker 4km NNW leptonana better than any other species. See also the Villa Altagracia, 200m, 12-IX-1992 (P. S. Ward) curvipubens discussion. [MCZ]. belti The range imwtabilis (Figs. 55-57, 97-98) over- Rogeria Mann laps that of leptonana at least in southern Mexico Figs. 35-39, 91-92 and Nicaragua, perhaps more, if queens are cor- Rogeria belti Mann 1922:31. Syntype workers, HONDURAS: rectly assigned. Workers of innotabilis differ as Progreso (Mann) [USNM] [2 of 12 syntypes examined]. follows: 1) no erect hair on scapes, 2) generally — larger (WL 0.66-0.73mm), 3) clypeus evenly con- Diagnosis. WL 0.61-0. 94mm. Clypeal apron vex, 4) palpal formula 2,2, 6) postpetiolar node usually with median notch, sometimes convex. widest midlength and sternum moderately to Eyes large, oval. Propodeal spines moderately strongly prominent. long, inclined (bisecting line usually passes well See also creightoni, curvipubens, alzatei, inermis, below anteroventral corner of pronotum). and neilyensis discussions. Promesonotal dorsum usually strongly areolate; The name from leptos (G., slender) and nanus postpetiolar node smooth. Scapes with erect hair. (L., dwarf) describes the habitus of this species. Hair on mesosoma dorsum decumbent to erect; Distribution. —Specimens range from Chiapas, not clearly sorting into two kinds. Head dorsum Mexico to the northern coast of Colombia; from and gaster Tl with short decumbent and long sea level to 1750m. Type specimens come from erect/suberect hair. No erect hair on extensor lowland rain forest, Colombian specimens from surfaces of tibiae. forest receiving about 2m strongly seasonal an- Workers.— TL 2.4-3.4, HL 0.56-0.80, HW 0.48- nual precipitation. Montane specimens from Costa 0.69, SL 0.39-0.55, EL 0.10-0.14 (20-36 facets), PW Rica were taken under large rocks in a pasture; 0.34-0.51, WL 0.61-0.94, SpL 0.11-0.20, PetL 0.25- those from Panama were collected in debris under 0.39, PpetL 0.13-0.21mm, CI 0.82-0.89, OI 0.18- a stump in unknown habitat.— 0.25, SI 0.76-0.89, PSI 0.14-0.24. N=31 Material Examined. Holotype and paratype Mandibles with 5-6 teeth; typically 5 teeth locality. PANAMA: Canal Zone, Barro Colorado evenly decreasing in size basad followed by a Island; 2 workers including holotype, H-III-1943, large basal tooth, or 4 teeth are followed by 1 or 2 lot #5059 (J. Zetek) [USNM, 43 3035]; 5 workers, denticles and a large basal tooth (Fig. 35). A few #5105 in scattered localities addi- VI-X-1943, (J. Zetek) [1 whole specimen specimens have an slide mounted] [USNM, lot 43-16534]; 4 workers, tional denticle between the third and fourth teeth. 14-11-1976 (A. Newton) [MCZ]; 1 worker, 27-11- Palpal formula 3,2. Clypeal apron medially emar- 1976 (A. Newton) [MCZ]; 1 worker, 1-1960, #B-9 ginate in most, but evenly convex in specimens (W. L Brown & E. S. McCluskey) [mouthparts, from Miami, Costa Rica. Body of clypeus not pro- sting] [MCZ]; 2 workers, 7-III-1975, FP#6 (C. Toft jecting over clypeal apron. Posterior outline of & S. Levings) [LACM]. head weakly convex, sometimes with a slight Nontype localities. MEXICO: 1 worker, median concavity. Eyes sometimes narrower than Chiapas, Ocosingo Valley, Finca El Real, 1-7- VII- shown in Fig. 35; widest in Miami, Costa Rica 1950 (Goodnight & Stannard). COSTA RICA: 6 specimens (Fig. 36). Pronotum from the side usu- workers, 4 males, Cerro Pico Blanco, near San Jose, ally subangular at junction of anterior and dorsal 1750m, 15-III-1987(W. Eberhard & W. L. Brown) [2 faces; from above, shoulders well rounded. Met- male genitalia] [MCZ], PANAMA: 2 workers, anotal groove either hardly visible, or a distinct Chiriqui Province, near Nueva California, W of narrow groove, or a small step in mesosoma pro- Finca Palo Santo, 1550m, 5-III-1959, #B-377 (H. S. file; usually marked behind by a transverse ridge Dybas) [MCZ]. COLOMBIA: 1 queen, Guajira, (Fig. 35). Propodeal spiracles small, slightly more near Don Diego, Quebrada Guacoche, 10m, 22- than 1 diameter from posterior edge of propodeum, VII-1976 W. L. Brown C. ( & Kugler) [queen mouth- facing slightly caudad. Metapleural lobes nar- Volume 3, 1 994 61 rowly to broadly rounded. Petiolar keel not lamel- (Fig. 48), but mesonotum slightly more convex late. Petiolar node usually dome-shaped (Fig. 35), and propodeal spines longer and narrower. Wings but sometimes slightly flattened on top. as in male (Fig. 37). Postpetiolar node highest in posterior half; dorsal Males.—TL 3.4-3.5, HL 0.59-0.61, HW 0.65- view subtrapezoidal as in Fig. 49; sternum long 0.70, SL 0.32-0.35, EL 0.32, PW 0.71-0.72, WL 1.12- and flat. Sting apparatus as in creightoni-group 1.19, PetL 0.40-0.41, PpetL 0.19-0.20mm, CI 1.11- diagnosis. 1.15, SI 0.50. N=2 Longitudinal rugae on frontal lobes become All five specimens from La Ceiba, Honduras. rugose-areolate by midlength of head; middle pair Mandibles with 5 teeth. Clypeus weakly emargin- of rugae usually remain free of cross-ridges until ate. Flagellomeres 2-11 straight, subequal in length posterior head. Cheeks and laterodorsa areolate to and width. Wings, habitus and genitalia as shown rugose-areolate. Sides of head strongly areolate in in Figs. 37-39. Rs vein curves away from leading most, but nearly smooth in Miami, Costa Rica edge of wing; r-m vein lost; M vein short. Man- specimens. Posterior head areolate with no clear dibles smooth except for piligerous punctures. transverse pattern in most, but transversely arch- Face very finely areolate, opaque. Posterior head ing rugose to rugose-areolate in ants from Miami, longitudinally rugose. Pronotum weakly punc- Costa Rica and Colombia. Promesonotum gener- tate. Mesonotum and posterior face of propodeum ally coarsely areolate to areolate-rugose (Figs. 91- finely rugulose and densely punctured. 92), but lower in relief and more rugose in speci- Mesopleura longitudinally striate right under mens from Miami, Costa Rica and some from wing, otherwise smooth and shiny with piligerous Colombia. Meso- and metapleura confused ar- punctures. Metapleura and sides of propodeum eolate to rugose. Dorsal face of propodeum pre- longitudinally rugose. Posterior face of propodeum dominantly transversely rugose, except for rugose-areolate. Petiolar node finely punctate to Boquete, Panama specimens, in which dorsal face smooth. Postpetiole and gaster smooth and shiny. is indistinctly areolate. Petiolar node not as heavily Hair as in worker, but more dense. Color brown- sculptured as mesosoma, but distinctly areolate to ish-black, with brown legs and antennae and weakly rugose-areolate; dorsal and anterior faces brownish-yellow— mandibles. often free of macrosculpture. Intervals irregular Discussion. The 18 specimens from Miami, and usually shiny. Microsculprure generally weak Costa Rica (Fig. 36) differ from typical belti in and indistinct, producing quite shiny intervals in having: 1) an evenly convex clypeus, 2) a larger macrosculpture, especially on sides of head, eye (31-36 facets vs. 20-30 in typical belti), 3) smaller mesosoma, and on petiolar node. Head dorsum propodeal spines (PSI 0.15-0.18 vs. 0.19-0.24 in and petiolar node of Miami, Costa Rica specimens typical belti), 4) generally narrower head, 5) head with more pronounced microsculprure, lending a and mesosoma more rugose than areolate, and 6) granular appearance. weaker macrosculpture in general, but Most dark reddish-brown with yellowish- microsculprure stronger on head and petiolar node. brown mandibles, antennae and legs. Specimens Queens from Miami, Costa Rica like their workers from Boquete, Panama and Miami, Costa Rica have narrower heads, shorter propodeal spines lighter, with more yellowish appendages and end and weaker sculpture than the four from Guate- of gaster. Mexican specimens reddish-black with mala, Panama and the Osa Peninsula of Costa brown appendages. Rica. I considered these Miami ants a new species Queens.—TL 3.0-3.4, HL 0.68-0.74, HW 0.58- until the arrival of new specimens from Costa Rica 0.64, SL 0.48-0.52, EL 0.18-0.20, PW 0.48-0.59, WL and Colombia. One worker from Heredia, Costa 0.88-1.00, SpL0.13-0.21,PetL0.29-0.37,PpetL0.18- Rica has a convex clypeus like the Miami ants, but 0.20mm, CI 0.84-0.85, SI 0.77-0.86, PSI 0.14-0.21. eye and propodeal spine size just within the typi- N=5 cal belti range. Other workers from Colombia have Like the worker except for normal queen struc- eyes as large as any of the Miami specimens, but an tures and the presence of predominantly rugose emarginate clypeus like typical belti, and they sculpture on the mesonotum. Parapsidal furrows have intermediate sculpture. While it is still pos- indistinguishable from furrows in sculpture. sible that the Miami population represents a dif- Habitus similar to that of merenbergiana queen ferent species, I am not confident of that at present. 62 Journal of Hymenoptera Research

Cerro El Valle Some alzatei (Figs. 58-60, 99) specimens Gaige); Campana (A. Newton); (H. belti Costa S. Canal Zone, Barro Colorado Island (J. strongly resemble specimens of from Dybas); A. COLOMBIA: Nariho Rica in size, habitus, spine length, and sculpture, Zetek; Newton). Depart- La 117 but alzatei: 1) is smaller (WL 0.52-0.65mm), 2) has ment, Guayacana (P. Silverstone) [LACM]. has 15 5 males a truncate clypeal apron in Central America, 3) workers, queens, [CUIC, JTLC, LACM, smaller eyes (11-14 facets), and 4) lacks erect hair MCZ, MZSP, USNM]. on the scapes. inermis Mann The geographic ranges of creightoni (Figs. 51- Rogeria 40-42 54, 95-96) and belti overlap in much of Central Figs. America and some members look quite similar in of inermis Mann 1922:32, Fig. 15. Syntype workers, size, sculpture, and pilosity. Workers creightoni Rogcrm HONDURAS: Progreso, Lombardia (Mann) [MCZ] [3of can be distinguished by: 1) generally smaller, el- 5 examined]. 19 syntypes liptical eyes with longer propodeal spines (< 0.29-0.64 vs. 0.62-1.23 in — facets; EL/SpL belti), 2) Diagtiosis. WL 0.60-0.85mm. Clypeal apron more horizontal propodeal spines, 3) promesonotal medially emarginate. Eye oval, moderately large. dorsum rugose to rugose-areolate, and 4) hair on Nuchal grooves weakly visible in lateral view. into two dis- mesosoma dorsum clearly sorting Propodeal spines short to absent. Propodeal spi- tinct kinds. racles rather large, prominent, face somewhat See discussions of inermis, tonduzi, caudad, usually < 1/2 diameter from edge of with merenbergiana, and exsulans for comparisons infradental lamella. Promesonotal dorsum rugose other similar species. to areolate-rugose; pronotal sides rugose, shiny. — is far from Distribution. Rogeriabelti so known Scapes with erect hair. Head dorsum with erect southwest southern Mexico to the Canal Zone and hairs and shorter, decumbent pilosity. Hair on to Colombia. Elevations range from 200m 1500m mesosoma and gaster Tl subdecumbent to erect; Labels in Mexico and 50m to 950m in Panama. not two distinct kinds. Pilosity on gaster rather data them collected in rain with ecological show sparse. Extensor surfaces of tibiae with decum- forest "in Philodendron," "frond sheaths on bent to suberect hairs. ground," leaf litter, around an epiphyte mat on a Workers.— TL 2.2-3.0, HL 0.55-0.71, HW 0.48- fallen branch, rotten wood, and in dead trees or 0.63, SL 0.38-0.50, EL 0.09-0.13 (15-28 facets), PW found a branches on the ground. Jack Longino 0.34-0.47, WL 0.60-0.85, SpL 0.03(no spines)-0.10, colony "nesting under loose bark of a still solid log PetL 0.25-0.37, PpetL 0.14-0.20mm, CI 0.82-0.91, rotten he lured fallen across trail." On another log OI 0.18-0.22, SI 0.73-0.88, PSI 0.04(no spines)-0.14. workers from under the bark to a freshly killed N=15 tabanid (Longino, unpublished field notes). Mandibles with 5 or 6 teeth, usually five de- — Cruz Material Examined. MEXICO: Vera creasing in size basad then a large basal tooth. Tuxtlas S. El State, Los (P. Ward); Chiapas State, Palpal formula 3,2. Median clypeal apron weakly Bobas Bosque (A. Newton). GUATEMALA: (W. to strongly emarginate. Posterior outline of head M. Mann). BELIZE [=British Honduras]: Caves convex. Metanotal groove weak to absent. Ante- Branch and HONDURAS: a (S. J. Peck). Progreso [or rior edge of propodeum sometimes marked by Lancetilla and La El Progreso] (W. M. Mann), transverse ridge. Propodeal spines usually short Ceiba L. 2 male (W. Brown) [mouthparts, sting; and acute (Fig. 40), but may also be quite long (Fig. between and genitalia]. EL SALVADOR: Apopa 41), blunt, or absent. Petiole (Fig. 40) with a dis- L. RICA: Miami Nejapa (W. Brown). COSTA (un- tinct, rounded node; peduncle slender, venter with known collector) [mouthparts, sting] [USNM]; strong tooth, but little or no keel. Postpetiolar Santa Clara Heredia Province, (J. Longino); node highest in posterior half; shape in dorsal = Farm and in 49. Anterior of [ Limon?] Province, Hamburg view subtrapezoidal as Fig. lip Parismina Br. Puntarenas Prov- (F. Nevermann); postpetiolar sternum square, not protruding. Py- Manuel Antonio National Park and Reserva ince, gidium as in creightoni-gioup diagnosis. Carara S. Osa Biologica (P. Ward), Peninsula, Sting apparatus as in Fig. 42. Spiracular plates Corcovado National Park P. S. (J. Longino; Ward). with incomplete medial connection; spiracle not M. PANAMA: Chiriqui Mountains, Boquete (F. enlarged. Anterodorsal corner of quadrate plate Volume 3, 1994 63

prominent, with medial and lateral lobes. Oblong Petiolar node with weak macrosculpture, plate anterior apodeme long, slender; ventral arm postpetiole smooth.— tall, with with long, slender fulcral arm. Gonostylus Discussion. The two specimens from the Osa with dorsoterminal and companion setae; no gap Peninsula of Costa Rica differ from the other speci- these other between and sensilla. Lancets strong, mens in having noticeably longer propodeal spines with a minute acute, single apical barb. Sting shaft and longitudinal, rather than transverse sculpture slender, straight; apex strong, acute, with low on the back of the head. These may be members of dorsal Valve flange. chamber and sting bulb large, a sibling species, but spine length is quite variable in and base much in inermis PSI 0.04-0.09 in subequal height length. Sting (e.g. one Berlese sample). lower than bulb I height, with weak anterolateral Thus, hesitate to name a new species until more processes. Furcula an inverted Y-shape. See Kugler specimens show consistent differences. for a (1978b) more complete description. Rogena leptonana (Figs. 66-70) from Cerro Pico Microsculpture on head and mesosoma re- Blanco, Costa Rica and Nueva California, Panama duced to in vague irregularities otherwise smooth have reduced or absent propodeal spines, but and of shiny interrugal spaces. Middorsum head differ from inermis in: 1) generally smaller size longitudinally rugose with lateral spurs, grading (WL 0.57-0.66mm; only 1 of 49 inermis workers to on areolate-rugose laterodorsa and sides. Sculp- with WL < 0.67mm), 2) smaller eye (6-8 facets; OI ture below and behind the eye sometimes vesti- 0.11-0.12), and 3) mesosoma and gaster Tl dorsa a area. Posterior head in most with 2 distinct of gial, leaving shiny types pilosity. Rogeria belti (Figs. specimens transversely arching rugose (with lat- 35-39, 91-92), often collected with inermis, has eral spurs) to areolate-rugose, but longitudinally longer propodeal spines and more areolate head in two workers from the Osa Peninsula of rugose and pronotal sculpture. See also the neilyensis dis- Costa Rica. Rugae on mesosoma are sharp, with cussion. broad interrugal spaces that are even smoother Distribution. —All specimens are from Guate- and shinier than on the head. Anterior pronotum mala to Panama, at elevations from 10m to 1250m. with 1-3 transverse rugae; disc with longitudinal Some were collected in fruit bat guano; others, in rugae; meso- and metanorum vermiculate-rugose Berlese samples from rain forest leaf litter. or areolate-rugose. Pronotal sides, mesopleura, Material Examined. —GUATEMALA: Alta and intervals metapleura predominantly rugose; Verapaz Department, Lanquin (S. and J. Peck). very shiny. Dorsal face of propodeum transversely HONDURAS: Progreso [or El Progreso]; to areolate. Sides of node rugose petiolar weakly Lombardia (W. M. Mann). COSTA RICA: (A. Bierig; to rugose rugose-areolate, producing an uneven, F. Nevermann); Heredia Province, Puerto Viejo, shiny surface; apex smooth. Postpetiolar node La Selva Field Station (C. Kugler; Talbot & smooth and with un- largely shining only slight VanDevender) [2 stings] ; Limon Province, Cahuita evenness of surface on sides. (P. S. Ward); Puntarenas Province, Osa Peninsula Color varies a from reddish-brown (almost (S. Peck), Manuel Antonio National Park (P. S. with black) body yellowish brown appendages to Ward), Carara Biological Reserve (P. S. Ward); a with darker yellowish-brown body gaster and Cartago Province, Turrialba vicinity (A. Bierig) L. brownish-yellow appendages. (W. Brown) [mouthparts, sting]. PANAMA: Queens.—TL 2.9-3.0, HL 0.62-0.64, HW 0.55- Bocas del Toro, 3 localities (D. M. Olson); Barro SL 0.57, 0.41-0.45, EL 0.15-0.16, PW 0.47-0.48, WL Colorado Island (W. L. Brown and E. S. McCluskey 0.81-0.84, 0.09, PetL 0.18- SpL 0.32-0.36, PpetL [mouthparts, whole specimen]; N. A. Weber; J. CI SI PSI 0.11. 0.19mm, 0.89, 0.74-0.79, N=3 Zetek); Gamboa (P. S. Ward); Panama City (W. L. Mandibles with 6 teeth, or 5 teeth and 2 den- Brown). 49 workers, 3 queens [CKC, CUIC, LACM, bdes. Parapsidal furrows not discernible. Sculp- MCZ, MZSP,USNM]. ture on posterior head of Panamanian specimens transversely rugose-areolate; more longitudinally Rogeria neilyensis new species oriented in Costa Rican specimens. Pronotum Fig. 43 transversely rugose-areolate. Mesonotal sculpture with — predominantly rugose, some closed cells. Diagnosis. WL 0.75-0.76mm. Clypeal apron Dorsal face of propodeum transversely rugose. convex, with weak median angle. Eye small. 64 Journal of Hymenoptera Research

Nuchal groove makes a deep notch in lateral view Color yellowish-brown, with slightly more of head. Propodeal spines short, wide. Propodeal yellowish appendages and anterior of head; first > 1 from of infradental of dark brown. spiracles diameter edge tergum gaster— lamella. Postpetiolar sternum not enlarged. Ter- Discussion. This species is similar in size and to Central minal segments of gaster rotated ventrad. Sides of reduced petiolar spines another Ameri- head with distinct areolate-rugose to rugose can species, biennis (Figs. 40-42), which differs as macrosculpture; mesosoma predominantly rug- follows: 1) clypeal apron medially emarginate, 2) ose. Scapes and legs lack erect hairs. Head dorsum nuchal groove weak, 3) eyes larger (OI 0.18-0.22; < with short, dense, decumbent pilosity and longer, 15-28 facets), 4) propodeal spiracle 1 diameter of erect more sparse, erect-suberect hairs. Mesosoma hairs from edge propodeum, 5) scapes with hair, range continuously from short, decumbent to long, and 6) mesosoma and gaster Tl with little, if any, hair. erect (i.e. not sorting into two distinct kinds as on decumbent head.). Hair on gaster Tl dense, subdecumbent- The leptonana (Figs. 66-70) specimens from suberect (decumbent hairs few or absent).— Cerro Pico Blanco, Costa Rica, have low, wide Holotype and Paratope Workers. TL 2.8-2.9 propodeal spines like neilyensis, but leptonana has: 0.58-0.60 1 an weak nuchal (2.8), HL 0.66-0.68 (0.67), HW (0.59), SL ) emarginate clypeus, 2) grooves, 0.45, EL 0.06-0.07 (0.06) (9-13 facets), PW 0.42-0.43 3) lower lower mesosoma profile (MHI 0.84-0.90), (0.42), WL 0.75-0.76 (0.75), SpL 0.08-0.10 (0.08), and 4) scapes with erect hair. CI The name refers to the PetL 0.30-0.34 (0.30), PpetL 0.18-0.19 (0.18)mm, neilyensis— type locality. 0.88-0.89 (0.88), OI 0.10-0.12 (0.10), SI 0.76-0.77 Material examined. Holotype locality. COSTA (0.76), PSI 0.11-0.13 (0.11), MHI 0.96-1.00 (0.96). RICA: Puntarenas Province, 3km N Ciudad Neily N=2 [=Villa Neily], 8.41N 82.57W, 210m, 31-VII-1985, Mandible 6-toothed; basal tooth larger than #7771-5 (P. S. Ward). From sifted litter (leaf mold penultimate tooth. On left mandible, fifth tooth and rotten wood) in second-growth rain forest replaced by 2 denticles. Body of clypeus slightly [MCZ]. prominent. Frontal lobes rather wide (FLW/HW Paratype locality. 2 workers, holotype locality 0.43-0.44). Anteroventral corner of pronotum pro- [LACM, MCZ]. jecting tooth-like. No distinct metanotal groove. Propodeal spiracles moderately large, weakly Rogeria exsulans Wilson and Taylor bulging from sides. Metapleural lobes low, broadly Fig. 44 rounded. Petiolar peduncle with large ventral keel exsulans Wilson and 60. and tooth. Postpetiolar node lacking a distinct Rogena Taylor 1967:74, Fig. Holotype worker, SAMOA: Upolu (T. E. Woodward) [Holotype posterior face; subtrapezoidal from above, as in MCZ; paratypes ANIC, MCZ] [Holotype and 30 Fig. 50. paratypes examined]. Longitudinal rugae on frontal lobes become — areolate-rugose halfway tobackof head. Posterior Diagnosis. WL 0.68-0. 84mm. Eye fairly large, head transversely areolate-rugose. Laterodorsa oval. Propodeal spiracle faces posterolaterally. areolate-rugose. Intervals in macrosculpture of Petiole with lamellate keel and dentate inferior head shiny, but wrinkled by vague, irregular process. Terminal segments of gaster rotated microsculpture. Anterior edge of pronotal disc ventrad. Sides of head areolate-rugose; posterior areolate; rest of promesonotal dorsum longitudi- head and promesonotum strongly areolate. Scapes nally rugose-areolate; more rugose mesad, more with decumbent and a few short, suberect hairs. areolate laterad. Dorsal face of propodeum trans- No erect hair on extensor surfaces of legs. Mostly versely rugose. Intervals on mesosoma generally abundant decumbent to subdecumbent hair on wider and smoother than on head. Petiolar node dorsa of head, mesosoma, nodes and gaster Tl vaguely areolate; posterior face and peduncle with (sparse erect hair on head and mesosoma); not longitudinal carinae. Surface of postpetiolar node sorting into two distinct kinds of pilosity. uneven, but predominantly smooth and shiny. Workers.— TL 2.5-3.0, HL 0.62-0.71, HW 0.52- Gaster Tl and SI smooth and shiny, but other 0.64, EL 0.08-0.11 (20-26 facets), PW 0.37-0.45, WL segments progressively dulled caudad by dense 0.68-0.84,SpL0.10-0.15,PetL0.26-0.33,PpetL0.15- microareolate sculpture. 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 and short on the /decumbent— pilosities gaster. Some mandibles with 6 teeth (5 decreasing in Distribution. Rogeria exsulans has been col- size then a large basal), but usually the small lected only at 600-700m on a single island in the penultimate basal tooth is replaced by 2 denticles, middle of the South Pacific. Most collections specify or by a gap and 1 denticle. Palpal formula 3,2. a rain forest habitat, where it has been found under Clypeal apron medially emarginate. Back of head dead bark, in rotten logs, in moss and ferns on in dorsal view convex or flat medially; temples trees and in berlesate of moss on logs and tree broadly rounded (Wilson and Taylor's Fig. 60). trunks. Promesonotal profile may be evenly rounded, but Material Examined.—WESTERN SAMOA: more often angular (Fig. 44). Metanotal groove Upolu, Afiamalu (T. E. Woodward; R. W. Taylor; usually distinct, producing a step between E. C. Zimmerman) [2 mouthparts, stings]. 31 work- promesonotum and propodeum, but weak in two ers [MCZ], specimens. Propodeal spiracles more than one diameter from nearest edge of infradental lamella. Rogeria cornuta new species Node short, evenly convex. Postpetiolar sternum Fig. 45 relatively long, flat; postpetiole from above — subtrapezoidal, as in Fig. 51. Sting apparatus as in Diagnosis. WL 0.93-1. 02mm. Eye relatively creightoni-group diagnosis. small. Nuchal groove makes strong notch in lat- Longitudinal rugae of frontal lobes and me- eral view of head. Propodeal spines very long dian head gradually change to strongly areolate (EL/SpL < 0.50); not inclined dorsad; distal por- on posterior head. Cheeks and laterodorsa longi- tions subparallel with midline. Propodeal spiracles tudinally areolate-rugose. Intervals on head dor- prominent, less than 1/2 diameter from edge of sum vaguely undulate, but shiny; intervals on infradental lamella. Metapleural lobes prominent. sides and posterior head smooth and shiny. Petiole with little or no keel. Sides and posterior Promesonotum for the most part strongly areolate, head strongly areolate; mesosoma predominantly but several elongate cells may occupy the midline rugose. Erect hair on scapes. Gaster Tl lacks de- of the pronotal disc, and often the pronotal sides cumbent hair. have weaker areolate sculpture. Meso- and Holotifpe and Paratype Workers. — TL (3.5)-3.8, metapleura confused areolate-rugose. Mesosoma HL (0.80)-0.89, HW (0>3)-0.78, SL (0.55)-0.58, EL microsculpture reduced; intervals in macrosculp- 0.10 (16-18 facets), PW (0.55)-0.60, WL (0.93)-1.02, ture uneven, but very shiny. Petiolar node with SpL (0.25)-0.27, PetL (0.42)-0.45, PpetL (0.21)- weak to vestigial areolate macrosculpture and 0.24mm, CI 0.88-(0.92), OI 0.13-(0. 14), SI 0.74-(0.75), vague microareolate background. Postpetiolar PSI 0.26-(0.27), MHI (0.98)-1.02. node shiny, nearly smooth throughout. Mandible with 6 teeth; basal larger than Head, mesosoma dark reddish-brown, waist penultimate basal. Clypeal apron medially emar- and anterior and posterior ends of gaster some- ginate; body of clypeus projecting slightly over what lighter. Mandibles, clypeus, antennae, and apron. Posterior outline of head broadly and legs yellowish-brown.— weakly concave. Sides of pronotum with anterior Discussion. The only other Pacific species, grooves for insertion of corners of head; shoulders stigmatica and megastigmatica are easily distin- from above angular. No meso- or metanotal guished from exsulans by their lack of a petiolar grooves. Paratype promesonotal dorsum less con- keel and inferior process, as well as other features vex and more angular in front and back than listed in the diagnoses. shown for holotype (Fig. 45). Propodeal directed The similar species most to exsulans is Central caudad so that a bisecting line would extend just American. Rogeria belti has some members with below shoulder; distal half of spines curve inward, similar size, habitus, eye size, clypeus shape, a almost paralleling the midline. Postpetiolar node distinct petiolar keel, strong areolate sculpture, somewhat flattened on top; subtrapezoidal in dor- and intergrading types of hairs on mesosoma. sal view. Postpetiolar node of paratype lower in However, belti workers still differ in having gener- front than behind. Postpetiolar sternum long, not ally longer propodeal spines (PSI 0.19-0.24), less projecting anteriorly. prominent petiolar keel, and distinct long/erect Head macrosculpture coarse (especially be- 66 Journal of Hymenoptera Research hind), with sharp ridges and shiny interstices. Curvipubens-Group and Related Species Longitudinal rugae on front break up at midlength Rogeria curvipubens Emery of head and give way to a transversely arching Figs. 74-76, 101-102 areolate-rugose pattern on the posterior head. Laterodorsa confused rugose-areolate. Mesosoma Rogeria curvipubens Emery 1894:190. Worker and queen also coarse with smooth inter- macrosculpture syntypes, U. S. VIRGIN ISLANDS: St. Thomas (Eggers) stices, but ridges are rounded. Anterior face of [MCSN] [Worker syntype examined]. pronotum transversely rugose-areolate, pronotal disc longitudinally rugose to vermiculate-rugose Additions to curvipubens-group diagnosis. vaulted and with small (holotype). Mesonotum vermiculate-rugose with Postpetiolar node strongly some cross-ridges. Pronotal sides areolate-rugose posterior peduncle. Anterior edge of postpetiolar (holotype) or broken and confused; rest of sides sternum not strongly produced; junction of poste- confusedly longitudinally rugose with few con- rior and ventral edges angular (Fig. 75). Sting shaft necting ridges. Anterior edge of propodeum and lancets weak, spatulate. Sides of head smooth marked by a sharp transverse ridge. Petiolar node and shiny. Promesonotum with vestigial weakly areolate on sides and posterior; smooth microsculprure, making interrugal spaces nearly along anterior and dorsal midline. Postpetiolar smooth and weakly to strongly shiny. Dorsal face node weakly areolate on sides; smooth along mid- of propodeum usually with 1-5 transverse rugulae. line. Workers.— TL 1.9-2.3, HL 0.48-0.58, HW 0.38- Scapes, head dorsum and tibiae with short 0.49, SL 0.32-0.39, EL 0.04-0.07 (4-11 facets), PW decumbent and long erect-suberect hair. Hair on 0.28-0.38, WL 0.50-0.63, SpL 0.05-0.11, PetL 0.20- mesosoma and waist ranges from decumbent to 0.24, PpetL 0.10-0.13mm, CI 0.80-0.86, OI 0.10- erect and varies in length, but not clearly segre- 0.15, SI 0.77-0.81, PSI 0.13-0.18, MHI 0.84-1.05. gated into two distinct types. Hair on gaster erect- N=22 suberect. Mandibles typically with 6 teeth, but occa- Color dark reddish-brown, with lighter sionally with an additional denticle or tooth; basal frontoclypeal area and ends of gaster; legs and tooth little if any larger than penultimate basal. antennae yellowish-brown.— Haitian specimens with 5 teeth. Anterior edge of Discussion. Some creightoni, also from Belize, clypeus evenly convex or with median concavity have the same habitus as cornuta, including long, (Haiti). Body of clypeus, though not especially horizontal propodeal spines, but these creightoni prominent, is generally the anteriormost point of members are much smaller (WL 0.63-0. 71mm) the head in full dorsal view; little or no shelf-like and have abundant decumbent pilosity on the apron. Posterior outline of head generally broadly gaster Tl. Other creightoni from La Selva, Costa convex, but sometimes with a weak median con- Rica are the same size as the cornuta holotype, but cavity (Fig. 74). Nuchal groove weak. Eye small, have shorter propodeal spines, slightly smaller elliptical. Pronotum weakly angular or rounded eyes, and abundant decumbent pilosity on the in front and on sides. Metanotal groove absent or gaster Tl. suggested by a broad, very shallow impression. The name cornuta means horned, referring to Anterior propodeum marked by transverse the horn-like carinula that often does not mesosoma long, propodeal— spines. interrupt Material Examined. Holotype locality. profile. Propodeal spines short, wide, with curved BELIZE (British Honduras): 2.5 mi. S Belmopan, 4- (Fig. 74) or straight edges; a bisecting line passes S. VIII-1972, and J. Peck, #242 [MZSP]. below the anteroventral corner of pronotum. Paratype locality. 1 worker, MEXICO: Chiapas Metapleural lobes low, subangular to rounded.

. like in State, Ocosingo, 2-VI-1967, J. M. Campbell [MCZ] Sting apparatus that of inermis (Fig. 42) 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 ever, if the two queens from Mexico and Guate- pronotal disc areolate to rugose-areolate; disc lon- mala are in fact cunnpubens, the range of cunnpubens gitudinally rugose (Fig. 102), sometimes effaced broadly overlaps that of sister species cuneola. The on meso- and metanota. Pronotal sides with one or two species are distinguished by characters in the more weak longitudinal rugae; meso- and cuneola diagnosis and description. confused to metapleura rugose rugose-areolate. Rogeria alzatei (Figs. 58-60, 99) and leptonana Petiole and postpetiole microareolate and devoid (Figs. 66-70) are very similar to curvipubens and of macrosculpture, except for small carinulae on cuneola and have been considered curvipubens in petiolar peduncle of some specimens. the past, but I believe they can be distinguished, weaker of Microsculpture on apices nodes, espe- usually by general habitus, but especially by more cially postpetiolar node, which is shiny and nearly abundant erect pilosity than described in the smooth. curvipubens-group diagnosis. Moreover, side-by- Head with 0-6 short suberect hairs; mesosoma side comparisons of sympatric specimens from dorsum with 2-7 pairs. Extent of erect pilosity on the Villavicencio vicinity of Colombia and from gaster Tl variable; usually covering whole sur- Barro Colorado Island reveal more differences: but little face, may cover as as the posterior third. the alzatei specimens have a truncate clypeal apron, Color uniformly yellow, to golden or light narrower propodeal spines, and generally stron- brown body with lighter appendages, ger sculpture. The leptonana specimens have an frontoclypeal region and ventral gaster. emarginate clypeal apron, lower mesosoma, and Queens.—TL 2.3-2.6, HL 0.51-0.59, HW 0.44- larger petiolar keel. The only question of identity 0.51, SL 0.34-0.39, EL 0.09-0.12, PW 0.38-0.45, WL arises on the north coast of Colombia, where a 0.63-0.71, SpL 0. 1 1-0.14, PetL 0.22-0.27, PpetL 0.12- single specimen has wider propodeal spines and 0.15mm, CI 0.83-0.90, SI 0.73-0.78, PSI 0.16-0.19, convex clypeus like curvipubens, which are com- MHI 0.96-1.17. N=6 in mon the area, but has the abundant pilosity and like in of Queens (Fig. 76) vary workers shapes stronger macrosculpture of alzatei. I call that speci- head, propodeal spines, petiole and postpetiole. men alzatei, since spine and clypeal shapes vary in Mandibles have 6 teeth in most; plus 2 denticles in that species. the Guatemalan specimen. Parapsidal furrows Assigning stray queens has also been prob- weak or indistinguishable. Wing venation as in lematic. In two localities on the north coast of belti (Fig. 37). Sting apparatus of a queen from the Colombia (Pueblito, Tayrona Park; Don Diego, north coast of Colombia like I that of workers from Guajira) collected three distinct kinds of queens. the region, including spatulate lancets. Sculpture, One set has a distinctly lower mesosoma (MHI and color also pilosity, vary similarly. Sides of 0.92-0.97) and larger petiolar keel than the other head be but may partly rugose, some portion two and has abundant erect pilosity on scapes, smooth. and Pro-, meso-, metanota rugose. Meso- whole head, mesosoma, waist, and gaster. These and metapleura longitudinally rugose dorsad; and other characters suggest those are leptonana anterior half of mesokatepisterna very smooth queens. The other two kinds of queens have a and as in shiny. Microsculpture workers. Num- more compact mesosoma (MHI 1 .05-1 .22), but one bers of erect hairs on various parts range as in set has no erect hair on scapes, head, waist, or first for one workers, except queen with 10 pairs on tergum of gaster, little or no erect hair on the mesosoma dorsum. mesosoma dorsum, more effaced sculpture with from Mexico Queens and Guatemala are not nearly smooth sides of head and transversely arch- associated with workers. Moreover, the Guatema- ing rugae on posterior head, and spatulate lancets, lan lacks a specimen postpetiole and gaster. Both just like the curvipubens workers that are common have the side of the head rather strongly rugose- in both areas. The third set of queens have distinct but I am that are areolate, guessing they cunnpubens, macrosculpture, distinctly areolate sides and pos- rather than cuneola on the basis of the strongly terior head, short erect hair on the whole head shiny mesokatepisterna and shape of the dorsum and more abundant erect /suberect hair sternum of the postpetiolar Mexican specimen. on mesosoma, waist and gaster and acute lancets. Discussion. —Workers come from the Carib- I have tentatively assigned them to innotabilis be- bean Islands and northern South America. How- cause of their evenly convex clypeus and 68 Journal of Hymenoptera Research posteroventral spine on the spiracular plate of the Nontype Workers.—TL 2.0-2.3, HL 0.50-0.56, sting apparatus. HW 0.43-0.48, SL 0.31-0.36, EL 0.04-0.06 (5-10 fac- See also micromma, and tribrocca discussions. ets), PW 0.30-0.37, WL 0.52-0.60, SpL 0.07-0.10, Distribution. —Rogeria curvipubens has been PetL 0.21-0.24, PpetL 0.11-0.13mm, CI 0.83-0.86, collected in by Berlese sampling in tropical rain OI 0.10-0.13, SI 0.72-0.78, PSI 0.14-0.17, MHI 0.93- forest, secondary growth forest, and dry tropical 1.03. N=9 forest. Collection sites have all been below 500m Like curvipubens, but differing in the follow- elevation. ing ways in addition to diagnosis. Relative widths Material Examined.—VIRGIN ISLANDS: St. of nodes with slightly different ranges (PetW/ St. PetL 1 .38-1 Thomas (Eggers; Balzan); Croix (I. Proj. staff). 0.56-0.70); PpetW/PpetL .52). Sting ap- HAITI: Aquin (no collector). JAMAICA: St. Eliza- paratus of specimens from Oaxaca (paratypes) beth Parish, Black River (no collector). MEXICO: and Vera Cruz, Mexico with strong, acute sting Vera Cruz, Cuatatolapan (M. Abarca). GUATE- shaft and lancets as in inennis (Fig. 42). MALA: Alta Verapaz, Trece Aguas (Schwarz and Posterior head with transversely arching rug- Barber). PANAMA: Barro Colorado Island (N. A. ose-areolate macrosculpture. Compared to Weber; W. L. Brown). COLOMBIA: Guajira, curvipubens, rugae on mesosoma dorsum with Serrania de Macuira and vicinity Rio Don Diego more lateral spurs that may connect rugae and (W. L. Brown and C. Kugler) [1 mouthparts, 2 create areolae on anterior pronotum and on met- stings, 1 whole specimen slide mounted]; anotum. Macrosculpture on mesosoma sides ab- Magdalena Department, Tigrera near Santa Marta sent or weakly rugose-areolate. Dorsal face of (W. L. Brown and C. Kugler), Tayrona National propodeum lacks macrosculpture. Park (C. Kugler) [worker and queen mouthparts, Head dorsum with 0-16 hairs suberect hairs; sting]; Meta Department, Caho El Buque near mesosoma dorsum with 1-8 pairs (usually 2-7). Villavicencio (Kugler). VENEZUELA: Bolivar Erect hair on gaster Tl usually limited to posterior State, Campamento Rio Grande (P. S. Ward); margin, but entirely absent from Jalisco specimen Monagas State, Parque de Laguna Grande (P. F. and entirely covering the tergum of the Yucatan Kukuk). SURINAM: Tambahredjo [queen sting], specimen. The Yucatan specimen is also unique in hairs Sidoredjo [sting], Dirkshoop, and La Poulle (I. van having some stiff, spatulate on head, der Drift). 57 workers, 13 queens [BMNH, CKC, mesosoma and gaster. CUIC, LACM, MCZ, MZSP, USNM]. 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, Rogeria cuneola new species PW 0.39-0.45, WL 0.65-0.72, SpL 0.11-0.14, PetL Figs. 77-78, 103 0.22-0.25, PpetL0.13-0.15mm, CI 0.85-0.89, SI 0.76- Additions to curi'ipubens-group diagnosis. 0.78, PSI 0.16-0.19. N=2 Postpetiolar node weakly vaulted and with no Queen as in curvipubens, except for shape of posterior peduncle. Anterior edge of sternum postpetiole, sting, and sculpture as in workers of strongly produced; posterior and ventral edges cuneola. Mandibles with 6 or 7 teeth. Sides of head merge insensibly (Fig. 78). Sting shaft and lancets and mesokatepisterna strongly microareolate and strong, acute; sting shaft with dorsal flange; lancet opaque. Paratype queen with erect-suberect hair with barbule. Sides of head and mesosoma with over whole gaster Tl. strong microareolate sculpture that obscures weak Discussion. —See micromma and minima dis- macrosculpture and makes intervals opaque. Dor- cussions for comparisons with other tiny Rogeria. sal face of propodeum without transverse rugulae. The name cuneola (L., small wedge) refers to the and Workers. — TL 2.0-2.1 of the in lateral view. Holotype Paratype shape postpetiolar— sternum (2.0), HL 0.50-0.54 (0.51), HW 0.43-0.46 (0.43), SL Distribution. These tiny ants are most often 0.32-0.35 (0.33), EL 0.05-0.06 (0.05) (6-7 facets), PW taken in berlesate of leaf litter and rotten wood. 0.31-0.33 (0.31), WL 0.51-0.56 (0.52), SpL 0.07-0.09 Some come from sittings under termite mounds (0.08), PetL 0.20-0.22 (0.20), PpetL 0.10-0.12 and one was collected in a Cattelya orchid. Habitat (O.ll)mm, CI 0.84-0.87 (0.84), OI 0.10-0.13 (0.13), SI of most specimens is rain forest or mesic forest, 0.73-0.78 (0.77), PSI 0.15-0.17, MHI 0.94-1.02 (1.02). either primary or secondary growth, but one speci- N=7 men was found in Yucatan thorn forest. 69 Volume 3, 1994

Material Examined. —Holotype locality. of erect hair; each node with 2 pairs of MEXICO: Oaxaca State, 1 mi. E Reforma, 15-VIII- posterodorsally projecting hairs. 0.37- 1973, litter, tropical evergreen forest (A. Newton) Workers.— TL 1.7-1.8, HL 0.44-0.51, HW [MCZ]. 0.45, SL 0.28-0.35, EL 0.02-0.04 (2-5 facets), PW PetL 0.17- Paratype localities. MEXICO: 14 workers, ho- 0.28-0.30, WL 0.45-0.52, SpL 0.07-0.08, 0.18, PetW 0.10-0.13, 0.10-0.11, 0.13- lotype locality [ BMNH, CKC, LACM, MCZ, MZSP, PpetL PpetW USMN]; 1 worker, 1 queen, Oaxaca State, 1 mi. E 0.15mm, CI 0.84-0.88, OI 0.05-0.08, SI 0.76-0.79, PSI Reforma, near Tuxtepec,12-15-VIII-1973, litter for- 0.15-0.16, MHI 0.92-0.96. N=3 est floor (A. Newton) [3 stings, 1 worker coated for The following supplements diagnosis and SEM] [MCZ]. Kempf (1961). Mandibles 5-toothed to 6-toothed; Nontype localities. MEXICO: 1 worker, San basal tooth not larger than penultimate basal. Luis Potosi State, El Salto, 6-VII-1969, #B-166 (S. Mesosoma profile of holotype interrupted by a weak metanotal followed two trans- and J. Peck) [MCZ]; 2 workers, Vera Cruz State, groove by Pueblo NuevonearTetzonapa,17-VIII-1953,#B.F. verse carinulae, but profiles of La Poulle and Bra- Brazil (E. O. Wilson) [MCZ]; 2 workers, Vera Cruz State, zil specimens uninterrupted. specimen with El Palmar near Tetzonapa, 8-VIII-1953, #173 (E. O. narrower propodeal spines than in types, and closer to of Wilson) [sting, whole specimen] [MCZ]; 1 worker, propodeal spiracle posterior edge Jalisco State, 6km N El Tuito, 31-XII-1987, #9327-6 propodeum. Petiole short (PetL//WL 0.35-0.38), keel tooth. (P. S. Ward) [MCZ]; 7 workers, 1 queen, Chiapas with ventral and L. In Surinam microareolate State, Palenque, 4-IX-1974 (E. M. and J. Fisher) specimens, vague [LACM]; 1 worker, Yucatan State, Uxmal, 27-VII- microsculpture densely covers head, dorsum of 1963, #118 (E. O. Wilson) [MCZ]. BELIZE: 4 work- mesosoma, dorsal face of propodeum, petiole and #B-243 and of a ers, Belmopan, 1-15-VIII-1972, (S. J. postpetiole types, producing weakly shining, Peck). GUATEMALA: 1 worker, 14-15-V-1946[?], granular appearance. Brazil specimen similar, ex-

SF-21214, 46-6134 (no locality orcollector) [USNM] . cept for smoother ventral petiolar peduncle and HONDURAS: 2 workers, La Lima, 4-IX-1961, UCF postpetiolar dorsum. Sides of mesosoma and pos- 217-128 (C. Evers Q.) [MZSP]. COSTA RICA: 1 terior surface of head more distinctly microareolate. worker, Guanacaste Province, Santa Rosa National Anterior of pronotal disc with 1 to 2 transverse Park, 14-VII-1983 (P. S. Ward) [CKC]; 3 workers, rugae followed by longitudinally rugose to ar- Puntarenas Province, Manuel Antonio National eolate-rugose macrosculpture that disappears into Park, 27-VII-1985, #7692-9 (P. S. Ward) [MCZ]. microsculpture on meso- and metanota. Sides of mesosoma with sparse, faint longitudinal rugulae Rogeria micromma Kempf on meso- and metapleura. Nodes microareolate; Fig. 71 more effaced on postpetiole. Erect hairs of Dirkshoop specimen all trichoid; Rogeria micromma Kempf 1961:509 (Figs. 12-13) . Holotype but thicker and stiff on the La Poulle specimen, at worker, SURINAM: Dirkshoop (J. vander Drift) [MZSP] least some cuneate; erect hairs on the Belem [Holotype and La Poulle paratype examined] speci- men seem intermediate. —WL 0.45-0. 52mm. Diagnosis. Clypeal apron Color brownish-yellow; legs, mandibles and flattened. node medially Eye tiny. Postpetiolar antennae sometimes slightly lighter. Frontocly peal in dorsal view and not subrectangular strongly region not lighter than rest of head. vaulted in side anterior of sternum not view; lip Discussion. — It may be that the three micromma in side view. prominent, posterior edge angular specimens are just unusual curvipubens or cuneola, Sides of head and mesosoma and dorsal face of but at present there are enough differences to with dense microareolate propodeum opaque provisionally retain this species. Eighteen Laterodorsa and sides of head also sculpture. finely curvipubens (Figs. 74-76, 101-102) workers from macroareolate. free Propodeum of macrosculpture. other van der Drift collections in Dirkshoop and No erect hair on or extensor surfaces of scapes La Poulle, are very similar to micromma in size and legs. Dorsa of head, mesosoma, nodes and but differ as follows: 1 gaster shape, ) clypeal apron evenly Tl with hairs and erect to short, appressed longer, convex, 2) sides of head and mesosoma shinier suberect hairs. Mesosoma dorsum with 8-10 pairs with effaced microsculpture, 3) sides of head with 70 Journal of Hymenoptera Research

rugose macrosculpture, and 4) reduced pilosity. Mandibles with 5 teeth; basal only slightly Some cuneola (Figs. 77-78, 103) specimens are larger than penultimate basal. Palpal formula 2,1. also very similar to micromma in shape and size, Clypeus torn; shape of apron unclear. Eyes mod- and one from Yucatan has stiff erect hairs on head, erately large, with about 20 facets. Mesosoma as waist and Tl but cuneola work- shown in 72. furrows cannot be mesosoma, gaster , Fig. Parapsidal ers differ in shape of the postpetiolar sternum, as discerned. Petiole with ventral tooth and well as in clypeal shape and pilosity. nonlamellate keel. Postpetiole wider (0. 18mm) than Some alzatei from Panama, Colombia, Guyana, long; sides of postpetiole seem to be convex in and French Guiana are only slightly larger (WL front, then tapered, as in many curvipnbens (Fig. 0.51-0.68mm) than micromma and have the same 74); sternum seems low and not prominent or pilosity and similar structure and sculpture, but wedge-shaped. Pygidial gland sculpture present, they generally have distinctly larger eyes with no tubercles on posterior surface. What is visible more than 10 facets, have a higher, more compact of the sting apparatus looks like that of inermis mesosoma (MHI 1.00-1.04), and generally nar- (Fig. 42), except that lancet apices lack barbules rower propodeal spines. Severalrt/za to from north- and sting shaft seems to have little, if any, terminal ern Colombia have reduced eyes (7-8 facets) and flange. Sinceboth sting shaft and lancets are folded, one has wider propodeal spines, but those Colom- they are probably weakly sclerotized. bian ants are larger, have a higher mesosoma, and Median head with diverging rugae continu- more abundant erect pilosity. See also the minima ing on posterior head as diverging rugose-areolate discussion. sculpture; laterodorsa areolate-rugose. No Material Examined. — SURINAM: Dirkshoop (J. macrosculpture on median pronotum; sides van der Drift), La Poulle 0- van der Drift). BRAZIL: weakly and incompletely areolate, especially Para, Pirelli Plantation (Iriboca) nr. Belem, (W. L. ventrad. Mesonotum longitudinally rugose. Dor- Brown). 3 workers [MCZ, MZSP]. sal half of mesopleura longitudinally rugose; most of mesokatepisterna smooth, but with some weak Rogeria minima Kusnezov areolate sculpture along posterior margins. Figs. 72-73 Metapleural and propodeal sides confused ar- eolate-rugose. Mesosoma lacks microsculpture, Rogeria minima Kusnezov 1958:44, Figs. 1-3. Holotypedealate except on median pronotum, sides of propodeal queen, ARGENTINA: Tucuman [FML] [Holotvpe ex- and lobes and amined]. spines, metapleural (metanotum Known only from a single queen mounted on a microscope median propodeum could not be examined). All slide dorsal head in half showing (nearly split length- surfaces of petiole, at least venter and sides of wise), ventral maxillae and labium, lateral mesosoma postpetiole, and at least sterna of gaster minutely and petiole, and a mostly ventral view of postpetiole and shallowly areolate with sharp, thin partitions. and gaster. Much of dorsal of difficult to — aspect gaster see, but Diagnosis. WL of worker probably < 0.60mm. at least anterior and lateral portions of Tl, and the Mandibles triangular. Palpal formula 2,1. Sting terminal terga are also shallowly microareolate. shaft and lancets spatulate. Postpetiole widest in Body covered with appressed to decumbent anterior half; anterior lip of sternum not promi- setiform pilosity. In addition, dorsa of head, nent. Mesosoma predominantly rugose. No mesosoma nodes, and gaster Tl and SI with erect little microsculpture on head dorsum, on cuneate-fimbriate hairs (Fig. 73). Mesosoma dor- mesosoma sides; but microareolate sculpture sum with more than 12 pairs of erect hairs; petiolar present on gaster Tl and SI. No erect hairs on node with 2 pairs of posterodorsally projecting scapes or extensor surfaces of legs. Head dorsum erect hairs (postpetiolar hairs hidden). All hair on with abundant erect hair; mesosoma dorsum with terminal segments of gaster are setiform. more than 12 pairs; petiolar node with 2 pair Discussion. —A worker of micromma from La projecting posterodorsally. Most erect hairs cu- Poulle, Surinam and a worker of cuneola from neate-fimbriate. Uxmal, Yucatan come very close in size, sculpture, TL in Queen.— 2.3, HL 0.55, HW 0.53, SL 0.34, EL and having cuneate pilosity, but the gasters of 0.10, WL 0.65, SpL 0.12, PetL 0.22, PpetL 0.11mm, both are smooth except for piligerous punctures. I CI 0.96, SI 0.64, PSI 0.18. cannot tell if any hairs are fimbriate. 71 Volume 3. 1 994

smooth on sides of head. Microareolate Fore/i-Group nearly and and pattern distinct on meso- metapleura on dorsal face of indistinct Rogeria foreli Emery generally propodeum; on and and vestigial on Figs. 79-82, 104-105 promesonotum petiole over- postpetiole (Figs. 104-105). Microsculprure 1894:191. worker, VIRGIN Rogeria foreli Emery Holotype lain by very fine longitudinal rugulae on lateral ISLANDS: St. Thomas (Eggers) [MHN] [Holotype ex- middorsum and clypeus, cheeks, frontal lobes, amined). head with Forel 1914:617. worker, CO- sometimes laterodorsa; posterior very Rogeria foreli gaigei Holotype N. or LOMBIA (Gaige) [MHN]) [Holotypeexamined], syn. fine transverse diverging rugulae. 1973:4, 1. and a brownish Rogeria huaciiucana Snelling Fig. Holotype Color yellow with slightly gaster USA: Arizona, Cochise County paratype worker, to chestnut-brown with brownish-yellow or light and 1 exam- (Snelling) [LACM] [Holotype paratype brown appendages. ined]. N. syn. Queens.—TL 2.4-3.1, HL 0.54-0.69, HW 0.48- Additions to foreli-gioup diagnosis. Basal 0.59, SL 0.36-0.49, EL 0.10-0.16, PW 0.38-0.51, WL mandibular teeth abruptly smaller than apical 0.64-0.87, SpL0.12-0.17,PetL0.24-0.32,PpetL0.13- with a faint PSI 0.17-0.20. teeth. Clypeal apron convex, often 0.20mm, CI 0.86-0.93, SI 0.72-0.83, or facets. median angle. Eyes usually 10 more N=8 0.43- in the normal Workers.— TL 1.9-2.9, HL 0.50-0.71, HW Differing from the workers queen 0.62, SL 0.32-0.51, EL 0.06-0.10 (7-20 facets), PW attributes and in the following. One specimen with the 0.30-0.45, WL 0.50-0.80, SpL 0.07-0.15, PetL 0.20- with 3 teeth and 5 denticles; the others SI workers. Posterior outline 0.32, PpetL 0.12-0.19mm, CI 0.83-0.89, 0.74-0.85, same variation as in the OI 0.12-0.17, PSI 0.14-0.20. N=25 of head with weak median concavity in some. Mandibles with 4-7 teeth and 0-3 denticles. Parapsidal furrows are barely discernable. ru- Generally teeth 1-4 decrease in size gradually, Mesoscutum in all specimens longitudinally smaller and to areolate- then teeth 5-7 (if present) abruptly gulose; mesoscutellum vaguely rugose or more denticles. possibly interspersed with one rugose. of the Sometimes basal tooth is distinctly larger than Discussion. —Side by side comparison often revealed that the penultimate tooth. Body of clypeus project- types of foreli and gaigei only Pos- is but the ing slightly over the anterior clypeal margin. difference between them size; gaigei type terior outline of head flat to weakly convex. Eyes is well within the size variation offoreli specimens. small, oval. Nuchal grooves inconspicuous in lat- Although the types of huachucana come from Ari- eral view. Pronotal shoulders well rounded. Met- zona, far from other known foreli specimens, and but differ anotal groove generally absent (Fig. 79), may were collected in an unusual habitat, they visible. Anterior in 1 7-8 facets be weakly to distinctly (Fig. 80) from theforeli holotype only having: ) a metanotal border of propodeum not marked by ridge. in the eyes (vs. 12), and 2) a weak groove rounded. Peti- size and metanotal Metapleural lobes small, broadly (vs. none). Since eye groove olar node shape varies between extremes shown development vary continuously and not concor- nodes are as as from Central and South in Figs. 79 and 81; smaller long dantly in foreli specimens wide, larger nodes are longer than wide. America, and since equally small eyes and even Postpetiolar node peaks in posterior half; more distinct metanotal grooves are present in the subrectangular in dorsal view. Anterior lip of those specimens, I am unable to distinguish or at this postpetiolar sternum small (Fig. 79), prominent huachucana specimens as a distinct species (Figs. 81, 104). time. Mandibles, median clypeus, legs, posterior While working in Northern Colombia for two sides of kinds of which face of propodeum, gaster, and sometimes years I regularly collected two foreli, for minute the of petiolar peduncle smooth, except I could distinguish at a glance by shapes of piligerous punctures. Rest body densely their postpetiolar sterna (Figs. 79, 81). Specimens microareolate or microcolliculate, often appear- from Barro Colorado Island, Panama, and Trinidad ing granular at low magnification. Microareolate also have the same two types of sterna. In Colom- sculpture on head is more distinct near antennal bia and Panama, both morphs have been taken insertions and more effaced caudad, sometimes from the same locality, but it is not known whether 72 Journal of Hymenoptera Research the two morphs come from different colonies or like the foreli with nonprojecting clypeus, weak or not. I considered calling the specimens with a absent metanotal groove, and nonprojecting more prominent, shelf-like postpetiolar sternum a postpetiolar sternum, but differ in the following new species, but the difference is sometimes subtle ways: 1) Mandible with five large teeth, second to and individuals difficult to assign. Without cor- fifth subequal (Fig. 82), 2) clypeus with a median roboration from another I decided smaller than in most character, notch, 3) eyes — foreli. against erecting a new species at this time. Material Examined. PARAGUAY: Misiones Rogeria foreli is closely related to bruchi, which Province, 30km S. of San Juan Bautista (F. Baud et at present is known only from Argentina and al.); Concepcion Province, 50km E. of San Lazaro Paraguay, much farther south than known foreli (F. Baud, et al.) [mouthparts, sting]. ARGENTINA: specimens (see the— bruchi description). Buenos Aires Province, Monte Veloz (C. Bruch); Distribution. Rogeria foreli has a disjunct dis- Misiones Province, Posadas (A. Breyer). 5 workers tribution: Panama and northern South America, [MCZ, MHN, MZSP, NHM], the Caribbean, and Southwestern United States. Specimens come from Om to 610m in Panama, 0m ACKNOWLEDGEMENTS to 240m in Colombia and Venezuela generally from berlesate of leaf mold and rotten wood. The I am very grateful to the following people for loans or of Cesare Baroni Urbani Claude Arizona specimens were under stones in an oak- gifts specimens: (NMB), Besuchet (MHN), Barry Bolton (BMNH), Carlos Roberto juniper woodland at 1783-1814m. Brandao (MZSP), William L.Brown, Jr. (MCZ), John E.Lattke, Material Examined. —USA: Arizona, Cochise Jack Longino, Alfred Newton (MCZ), David M. Olson, Roberto County (R. R. Snelling). PUERTO RICO: Rio Piedras Poggi (MCSN), David Smith (USNM), Roy R. Snelling S and E. O. Wilson (J. A. Torres). ST. CROIX: Buck Island. ST. THO- (LACM), Philip Ward, (MCZ). Jack also sent me his field col- MAS (Eggers). TRINIDAD: (N. A. Weber). Longino generously unpublished lections notes. PANAMA: Barro Colorado Island (W. L. Brown, Lily Fainter of the Virginia-Maryland Regional College W. L. Brown and E. S. A. Newton, N. McCluskey, of Veterinary Medicine and Bob Honeycutt of Virginia Poly- L. A. Weber, J. Zetek); Gahin (W. Brown); Ancon technic Institute and State University, Department of Forest Products assisted with the electron (no collector) [sting]; Gamboa (P. S. Ward); Colon scanning microscopy. Reed R. Lambert helped collect ants in Peru and Bolivia. Province, Frijoles (J. Ventocilla); Cerro Azul (A. Richard L. Hoffman, W. L. Brown, Jr., Andre Francoeur, and Newton). COLOMBIA: no locality (Gaige); R. W. Taylor provided valuable advice. Special thanks to Roy Park (C. Magdalena Department, Tayrona Kugler, Snelling, who made many helpful criticisms of a draft. None P. S. L. Ward) [mouthparts, sting]; Tigrera (W. of these bear any responsibility for the product. The work was Brown and C. Kugler) [queen mouthparts, whole supported financially by NSF grant #DEB- worker 8022177 and Radford University. specimen; mouthparts, sting] ; Guaj ira, Don Diego (W. L. Brown and C. Kugler). VENEZU- ELA: Barinas State, near Ciudad Bolivia (P. Ward); Sucre LITERATURE CITED State, 32km W. Campano (S. & J. Peck). 106 workers, 17 queens, 4 males [CKC, CUIC, LACM, B. 1973. The ant of West Africa: A MCZ, MCZ, MZSP, USNM]. Bolton, genera synonymic synopsis with keys (Hymenoptera: Formicidae). Bulle- tin of the British Museum (Natural History), Entomology Rogeria bruchi Santschi 27(6):319-368. B. 1982. of the ant Fig. 82 Bolton, Afrotropical species myrmicine genera Cardiocondyla, Leptothorax, Melissotarsus, Messor and Cataulacus (Formicidae). Bulletin of the British Mu- Rogeria bruchi Santschi 1922:352. Holotype worker, ARGEN- seum (Natural History), Entomology 45(4):307-370. TINA, Buenos Aires, Monte Veloz (C. Bruch) [NHM] Bolton, B. 1987. A review of the and [Holotype examined). Solenopsis genus-group revision of afrotropical Monomorium Mayr (Hy- Bulletin the British Museum Workers.— TL 2.2-2.6, HL 0.56-0.65, HW 0.48- menoptera: Formicidae). of (Natural History), Entomology 54(3):263-452. 0.57, SL 0.38-0.46, EL 0.04-0.06 (5-7 facets), PW Borgmeier, T. 1953 (1951). Algumas formigas do genero 0.35-0.40, WL 0.58-0.70, 0.10-0.17, PetL 0.21- SpL Macromischa Roger. Arquwos do Museu Nacwnal Rio de 0.29, PpetL 0.13-0.17mm, CI 0.61-0.70, OI 0.10- /amero 42:107-112. 0.12, SI 0.76-0.81, PSI 0.17-0.24. N=5 Brown, W. L. 1952. Synonymous ant names. Psyche 58(3): 124. Brown, W. L. 1953. Characters and the The five known specimens of bruchi are just synonymies among Volume 3. 1 994 73

genera of ants. Part II. Brevwra no. 18:1-13. Entomologica 15(l-4):3-344. Brown, W. L. 1988. Data on malpighian tubule numbers in Kempf, W. W. 1975. Miscellaneous studies on neotropical ants 17-27 in: C. ants. VI. (Hymenoptera: Formicidae). Pp. J. (Hym. Formicidae). Studia Entomologica 18(1-

in . Leiden: E. Trager (ed.) Advances myrmecology J. Brill, 4):341-380. xxvii+551 pp. Kugler, C. 1978a. Pygidial glands in the myrmicine ants Brown, W. L., and W. L. 1949. Wing venation and the phylog- (Hymenoptera, Formicidae). Insectes Sociaux 25(3):267- eny of the Formicidae. Transactions of the American Ento- 274. mological Society 75(3-4):113-134. Kugler, C. 1978b. A comparative study of the myrmicine Emery, C. 1894. Studi sulle formiche della fauna neotropica. sting apparatus (Hymenoptera, Formicidae). Studia V1-XV1. Bullettnio della Societa Entomologica Italiana26:\37- Entomologica 20:413-548. 242 + 4 pi. Kugler, C. 1980. The sting apparatus in the primitive ants C. 1896. sulle Emery, Studi formiche della fauna neotropica. Nothomyrmecia and Myrmecia. Journal of the Australian XVII-XXV. Bullettmo della Societa Entomologica ltaliana Entomological Society 19:263-267. 28:33-107. Kugler, C. 1986. Stings of ants of the tribe Pheidologetini Emery, C. 1897. Formicidarum species novae vel minus (Myrmicinae). Insecta Mundi l(4):221-230. cognitae in collectione musaei nationalis Hungarici quas Kugler, C. 1992 (1991). Stings of ants of the tribe Ectatommini in Nova-Guinea, colonia germanica, collegit L. Biro. (Formicidae: Ponerinae). Insecta Mundi 5(3-4):153-166. Fuzetek 20:571-599 + 2 N. Termeszetrajzi pi. Kusnezov, 1958. La posicion sistematica del genero Rogeria, C. 1914. Les fourmis de la et Emery, Nouvelle-Caledonie des con descripcion de una nueva especie. Acta Zoologica lies F. Loyalty. P 415 In: Sarasin and J. Roux (eds.) Nova Lilloana 15:41-45. Caledonia recherches scientifiques en Nouvelle Caledonie et Linneaus, C. 1758. Systema Naturae Regnum Animate ed. 10, aux iles Loyalty (Zoologie). Wiesbaden: C. W. Kreidels Tom 1. Stockholm, VI+824pp. Verlag, l(4):389-436+13 pi. Mann, W. M. 1919. The ants of the British Solomon Islands. C. 1915. Emery, Noms de sous-genres et de genres proposes Bulletin of the Museum of Comparative Zoology, Harvard la sous-famille des Modifications d la + pour Myrmicinae. 63(7):273-391 2 pi. classification de ce 1921. groupe (Hymenoptera Formicidae). Mann, W. M. The ants of the Fiji Islands. Bulletin of the Bulletin de la Societe Entomologique de France 1915:189- Museum of Comparative Zoology, Harvard 64(5):401-499. 192. Mann, W. M. 1922. Ants from Honduras and Guatemala. 1899. Forel, A. Formicidae. Biologia Centrali-Americana. Insecta. Proceedings of the United States National Museum 61(13):1- Hymenoptera 3:1-160+4 pi. 54. A. 1914. Forel, Einige amerikanische ameisen. Deutsche Mann, W. M. 1925. Ants collected by the University of Iowa 1914:615-620. Entomologische Zeitschrift Fiji-New Zealand expedition. University of Iowa Studies Harris, Rick A. 1979. A glossary of surface sculpturing. Occa- m Natural History ll(4):5-6. sional Papers in Entomology. Department of Food and Agri- Mayr, G. L. 1886. Notizen ueber die Formiciden-Sammlung culture, State of California no. 28:1-31. des British Museum in London. Verhandlungen der B. and H. 1978. Holldobler, Engel. Tergal and sternal glands Zoologisch-Botanischen Gesellschaft in Wein 36:353-368. in ants. 1863. Psyche 85(4):285-330. Roger, J. Die neu aufgefuehrten Gattungen und Arten Holldobler, B., R. Stanton, R., and Engel. 1976. A new exo- meines Formicidenverzeichnisses, nebst Ergaenzungen crine gland in Novomessor (Hymenoptera: Formicidae) einiger frueher gegebenen Beschreibungen. Berliner and its as a possible significance taxonomic character. Entomologische Zeitschrift 7:131-214. 32-41. Psyche 83(1 ): Santschi, F. 1922. Myrmicines, dolichoderines et autre B. and E. Holldobler, O. Wilson. 1990. The Ants. Cambridge, formicides neotropiques. Bulletin de la Societe Vaudoise Massachusetts: Belknap Press, xii+732 pp. Sciences Naturelles 54(205):345-378. W. W. 1961. Kempf, A survey of the ants of the soil fauna in Santschi, F. 1923. Description de quelques nouvelles fourmis Studia Surinam. Entomologica 4(l-4):481-524. du Bresil. Reinsta (Museu Paulista, Sao Paulo, Brazil) Kempf, W. W. 1962a (1961). Remarks on the ant genus Irogera 13:1253-1264, Emery, with the description of a new species (Hy- Santschi, F. 1930. Quelques fourmis de Cuba et du Bresil. menoptera, Formicidae). Rei'istaBrasileiradeEntomologia Bulletin de la Societe Royale de Botamque d'Egypte Mai 21(4):435-441. 1930(2-3):75-83. Kempf, W. W. 1962b. Miscellaneous Studies on Neotropical Santschi, F. 1936. Contribution & l'etude des fourmis de Ants. II. (Hymenoptera, Formicidae). Studia Entomologica l'Amerique du Sud. Revista de Entomologia, Rio de Janeiro 5(l-4):l-38. 6(2):196-218. W. W. 1963. Additions to the Kempf, neotropical ant genus Santschi, F. 1941. Quelques fourmis japonaises inedites. with a Rogeria Emery, key to the hitherto recorded South Mitteilungen der Schweizerischen Entomologischen American species (Hym., Formicidae). Remsta Brasileira Gesellschaft 18:273-279. de Entomologia 23(2):189-196. Smith, F. 1858. Catalogue of the hymenopterous insects in the W. W. 1964. Kempf, Miscellaneous studies on neotropical collection of the British Museum. Part VI. Formicidae. Lon- ants. III. Studia 216 + Entomologica 7(l-4):45-71. don, pp 14 pi. W. W. 1965. Nota sobre Kempf, preliminar algumas formigas Smith, M. 1937. The ants of Puerto Rico. University of Puerto descritas neotropicas, por Frederick Smith (Hy- Rico, Journal of Agriculture 20(4):819-875. menoptera, Formicidae). Revista Brasileira de Entomologia Snelling, R. R. 1973. Two ant genera new to the United States. 25(2):181-186. Contributions in Science, Natural History Museum of Los W. W. 1972. Kempf, Catalogo abreviado das formigas de Angeles County no. 236:1-8. regiao neotropical (Hym. Formicidae). Studia Snelling, R. R. 1989. Untitled comment. Notes From Under- 74 Journal of Hymenoptera Research

ground: A Myrmecological Newsletter. MCZ Laboratories, of Washington D.C. vi + 108 pp. G. C. 1977. Harvard, MA no. 3, pp. 7-8. Wheeler, and J. Wheeler. Supplementary studies for of ant larvae: Transactions the American Trager, J. C. 1989. A pitch standardized orientation the on Myrmicinae. of ant head, and related matters. Notesfrom Underground: A Entomological Society 103:581-602. 1986. Myrmecological Newsletter. MCZ Laboratories, Harvard, Wheeler, G. C and J. Wheeler. Supplementary studies MA. no. 1, pp. 5-7. on ant larvae: Myrmicinae (Hymenoptera: Formicidae). Weber, N. A. 1934. Notes on neotropical ants, including the journal of the New York Entomological Society 94(4):489- descriptions of new forms. Revista de Entomologia, Rw de 499.

Janeiro 4(l):22-59. Wheeler, G. C. and J. Wheeler. 1988. Notes on ant larvae: Wheeler, G. C. 1989. Vive la myrmecologie. Notes from Under- Myrmicinae. Transactions of the American Entomological ground: a Myrmecological Newsletter. MCZ Laboratories, Society 114:319-327. Harvard, MA. no. 3, pp. 6-7. Wheeler, W M. 1911. A list of the type species of the genera

Wheeler, G. C. and J. Wheeler. 1953. The ant larvae of the and subgenera of Formicidae. Annals of the Neiv York myrmicine tribe Pheidolini. Proceedings of the Entomo- Academy of Sciences 21:157-175. logical Society of Washington 55(2):49-84. Wilson, E. O 1955. A monographic revision of the ant genus Wheeler. 1973. larvae of four tribes: Lasius. Bulletin the Wheeler, G. C. and J Ant of Museum of Comparative Zoology, Second supplement (Hymenoptera: Formicidae: Harvard 113(1):1-201. Myrmicinae). Psyche 80(l-2):70-82. Wilson, E. O. and R. W. Taylor. 1967. The ants of Polynesia C. and 1976. Ant larvae: Review and Wheeler, G. J. Wheeler. (Hymenoptera: Formicidae). Pacific Insects Monograph synthesis. (Memoires of the Entomological Society of 14:1-109. Washington, no. 7). Washington: Entomological Society

Posterior head

Middorsum

Laterodorsum >,/ Nuchal groove Frontal area Side

Lateral clypeus Venter Body of clypeus

Denticles Clypeal apron

Erect Suberect

1 Appressed

Fig. 1. Illustration of some terms and measurements used in descriptions. Dotted lines on the heads show approximate of the is boundaries regions. The dashed line an extension of a line bisecting the propodeal spines. Where it crosses the outline a of = = mesosoma gives rough measure 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 furcula. Below lateral furcula 4. sting, right: sting, (Fu). Paratype queen lateral mesosoma. Figs. 5-6. Rogeria gibba. 5. Holotype dorsal head. 6. profile; Paratype sting with furcula in lateral and ventral views; lateral view of lancet apex. 76 Journal of Hymenoptera Research

14 prominula

7-14. 7. Figs. Figs. 7-12. Rogeria stigmatica. 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 with furcula Western 10. sting (Falepuna, Samoa). Queen mesosoma profile (Viti Levu, Fiji). 11. Male Viti 12. Genitalia (Sigatoka, Levu). (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

15-23. Figs. Figs. 15-16. Rogeria besucheti. 15. Holotype profile; dorsal head. 16. Paratype ventral and lateral views of sting with furcula. 17. blanda Fig. Rogerui 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. 21. Fig. 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

24-31. 24-27. 24. Profile hairs had been matted down dorsal dorsal Figs. Figs. Rogeria germaini.— (mesosoma somewhat), head; 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

32-39. 32-33. 32. Figs. Figs. Rogerm pellecta. Paratype profile; dorsal postpetiole. 33. Paratype lateral sting, furcula. Fig. 34. sicaria Rogeria 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. 44. exsulans 45. cornuta Fig. Rogeria holotype profile. Fig. Rogeria holotype profile; dorsal head. Figs. 46-48. Rogeria 46. 47. merenbergiana. Holotype profile. Nontype mesosoma dorsum profile (hair omitted) showing the least distinct metanotum (Cuevas de los Guacharos, Col.). 48. lateral mesosoma Paratype queen and waist. Same scale for all figures. 82 Journal of Hymenoptera Research

unguispma

creightoni

49-54. 49. dorsal dorsal waist. 50. brunnea dorsal Figs. Fig. Rogena unguispma— holotype profile; head; Fig. Rogena profile; 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. 55-57. innotabilis. 55. dorsal head. 56. Figs. Rogeria Holotype profile; Paratype spiracular plate from sting apparatus (Chiapas, Mex). 57. 58-60. Nontype queen profile (Parque Tayrona, Col.). Figs. Rogeria alzatei. 58. Holotype profile, dorsal head; dorsal waist. 59. (Pto. Maldonado, Lake 60. Nontype profile Sandoval, Peru). Nontype profile (R. Don Diego, Col.). Figs. 61- 62. scobinata. 61. dorsal head. Rogeria Holotype profile; 62. Paratype queen profile (holotype loc). Same scale for all external views. K4 Journal of Hymenoptera Research

tribrocca

carinata nevadensis

leptonana

dorsal head. 64. carinata dorsal head. Figs. 63-70. Fig. 63. Rogeria tribrocca holotype profile; Fig. Rogena holotype profile; Fig. head. 66-70. 66. dorsal head; dorsal 65. Rogeria nevadensis holotype profile, dorsal Figs. Rogeria leptonana. Holotype profile; 68. 69. Male postpetiole. 67. Nontype profile (nr. Nueva California, Pan.). Nontype queen profile (Parque Tayrona, Col.). (Cerro Pico Blanco, C. R). 70. Male genitalia (same loc). Same scale for all external views. 85 Volume 3. 1 994

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

79. 80. mesosoma dorsum metanotal Figs. 79-82. Figs. 79-81. Rogerm foreli. Holotype profile. Nontype profile showing strongest dorsal mandibles groove (Cerro Azul, Pan.) 81. Lateral waist (Don Diego vie, Col). Fig. 82. Rogeria bruchi holotype profile; and clypeus. All to same scale Volume 3. 1 994 87

;W3fe|

86 tonduzi I N« l» :""'

lirata

83-90. 83-84. blanda 83. Lateral waist. 84. Figs. Figs. Rogeria (L. Sandoval, Peru). mesosoma, Dorsal head, mesosoma, petiole. 85-86. tonduzi Selva C. 85. Lateral Figs. Rogeria (La Station, R.). mesosoma and waist. 86. Dorsal mesosoma. Figs. 87-88. Rogerw (Alto Parana 87. Lateral mesosoma. 88. germaini Prov., Para.). Dorsal mesosoma, posterior head. Figs. 89-90. Rogeria lirata paratype. 89. Lateral mesosoma. 90. Dorsal mesosoma. Scale lines = 0.10mm. 88 Journal of Hymenoptera Research

mm'*/® '''*IWmm

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

99-105. 99. alzatei dorsal Figs. Fig. Rogeria paratype posterodorsal head, pronotum. Fig. 100. Rogeria scobinata paratype posterodorsal head. Figs. 101-102. Rogeria curvipubens (R. Don Diego vie, Col). 101. Posterodorsal head. 102. Dorsal mesosoma. 103. cuneola dorsal Fig. Rogeria paratype mesosoma. Figs. 104-105. Rogeria foreli (Parque Tayrona, Col.). 104. Lateral waist anterior of showing prominent lip 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 genera. Kieffer's and subsequent authors' con- cepts of the bethyline genera needed modification The aculeate family Bethylidae (Chrysidoidea) to accommodate the degree of morphological varia- is a moderately large family of about 1,900 de- tion which we had discovered. Computerised scribed species with probably at least as many phylogenetic analysis was selected as the most undescribed. Bethylids are mainly gregarious objective method of assessing character states ectoparasitoids, the Bethylinae mostly develop- within the bethyline genera, while providing a ing on larval Lepidoptera, the remainder mostly more stable classification of the subfamily. The on Coleoptera, although the hosts are known for following study was carried out primarily to set only a small proportion of the family. Bethylids new limits to some of these genera, and to facilitate are frequently encountered as parasitoids of crop their identification. pests, especially in tropical areas, and several spe- Here we address the internal phylogeny of the cies have been used in attempted biological con- Bethylinae. In comparison to the other bethylid trol (Gordh & Evans 1976; Hempel 1934). Success- subfamilies, the Bethylinae have been little af- ful identification of Bethylidae is therefore impor- fected at the generic level since Kieffer's (1914) tant for many economic entomologists, but has revision. Only one genus, Neoclystopsenella Kurian been extremely difficult because the most recent (1955), was assigned to the Bethylinae since revision was published eighty years ago (Kieffer Kieffer's work. Neoclystopsenella was synonymised 1914) and contains many errors. Even at the ge- with Tapinoma (Formicidae) by Brown (1987), but neric level identification can be very difficult, par- mistakenly retained by Gordh & Moczar (1990) in ticularly for the Palaeotropical species. For the the Bethylidae. Gordh & Moczar (1990) also mis- Nearctic and Neotropical species this problem has takenly transferred Odontepyris Kieffer to Epyrinae, been alleviated largely by the work of Evans (1964, thus assigning a total of eight genera to the 1978). Bethylinae (Table 1). When attempting to identify our on During independent studies African Bethylinae genera using Kieffer's (1914) key, the and Asian Bethylinae we have each discovered primary source of confusion is his treatment of the related undescribed species that donotfitKieffer's genera Goniozus Fbrster, Parasierola Cameron and (1914) concepts of the bethyline genera, but which Perisierola Kieffer. The latter two genera have since we do not regard as meriting description in new correctly been synonymised with Goniozus (Evans 92 Journal of Hymenoptera Research

are for de- 1978), but from our studies it is clear that Kieffer venation characters which important Current intended something different from what he inad- ducing the phylogeny of the subfamily. of the Anoxus and vertently achieved when he keyed and diagnosed interpretations genera Bethylus differ from each other in whether the are these genera in his revision. Goniozus (sensu Evans only eyes or not. This is a character that we have 1978) contains species that either possess, or do setose observed to so one not possess, a closed discoidal cell (areola, areolet) vary intragenerically, only with a dis- of was included in (Fig. 1). Kieffer assigned those species representative species Bethylus coidal cell to Perisierola, and those without one to the analysis. The taxa selected are given in the data Goniozus. Kieffer (1907, 1911, 1914), Brues (1907a matrix below (Table 2) in the generic combinations 1907b) and Muesebeck (1940) all mistakenly se- which have resulted from this study. Former ge- lected Parasierola Cameron (1883) to accommodate neric combinations can be found in the treatments cell with of and species with both a closed discoidal and Bethylus Odontepyris (see below). one or more longitudinal carinae on the propodeum, this lattercharacter being absent from Selection of Characters Kieffer's concepts of the other two genera. Fur- thermore, Kieffer (1914) transposed his concepts We consider the following to represent the of Parasierola and Perisierola when going from his ground plan characters of the subfamily Bethylinae. list is based on the work of key to genera (1914:238) to his generic diagnoses This character partly of but on our (1914:533, 542). Thus began 80 years of confusion Evans (1964) and Sorg (1988) largely surrounding these bethyline genera. own independent assessments.

METHODS Ground plan characters within Bethylinae

Selection of taxa Antennae 13-segmented; clypeus with a well- developed keel, and frontal streak present; maxil- labial To clarify the status of the genera of Bethy linae lary palps 6-segmented; palps 3-segmented; we analysed 11 taxa of Bethylinae for 22 morpho- notauli and parapsidal lines present; scutellum with small at its corners logical characters using the parsimony programme flat, two grooves proximal carinae see be- 'Hennig86' (Farris 1988). To polarise characters (Fig. 2); propodeum without (but the genus Lytopsenella Kieffer was selected as the low); fore femora expanded; prostigma absent; all discoidal cell and outgroup. Lytopsenella possesses the characters (areolet) present, submarginal common to all the remaining bethyline genera in marginal cells present; subdiscoidal cell absent their hypothetical plesiomorphic conditions (see (=discoidal cell of Evans 1964); sternum of petiole character selection). Characters that are prone to with a complete keel, sexual dimorphism limited to characters and head reduction (e.g., number of antennal segments, genital shape. maxillary and labial palp segments, and wing venation) are found at their maxima within Characters Analysed Lytopsenella. Lytopsenella has previously been cho- all of those sen as a basal group, not just for Bethylinae but for The following characters include Bethylidae as a whole (Evans 1964; Sorg 1988). which have been used previously for the discrimi- Representatives of each of the currently valid nation of genera within the Bethylinae (except eye bethylid genera were included in the analysis. In setation, see above) as well as some which have cases of existing doubt or controversy surround- not been used previously. We generally agree ing the limits of some genera, type species of both with Sorg (1988) concerning both the selection and current genera and formerly recognised genera polarization of characters. However, we disagree were examined. Particular attention was paid to with Sorg's polarization of the scutellar foveae previous authors' concepts of Goniozus, (character 6, below; Figs. 3, 4). Sorg considers that Odontepyris, Parasierola, Perisierola and Trissomalus. the occurence of scutellar foveae in the Three species of Goniozus (in the broad sense, i. e. Embolemidae (Sorg 1988: p 30) suggests that of Evans 1978) were selected to cover the plesiomorphy. The probability is that, at least in range of known variation in propodeal and wing the Bethylinae, they are represented in their Volume 3, 1 994 93

Table 1: Genera of Bethylinae

Kieffer (1914) 94 Journal of Hymenoptera Research

pterostigma marginal cell position of prostigma

median cell

submedian cell

subdiscoidal cell (absent in Bethylinae)

of venation Fig. 1. Bethylidae: terminology wing

these have apparently become enlarged to propodeum is characterised by a smooth, trans- area which shows form pits, often connected by a broad unsculptured triangular or modifications in several taxa verse groove (Figs. 4, 6, 7). 0= scutellar pits (Figs. 7, 10, 11). It is in the basal The foveae present as grooves; 1= scutellar foveae absent groups. develop- character in lin- enlarged. ment of this some bethyline Expansion of mesopleuron: Kieffer (1905) eages should not be confused with the characterised his genus Odontepyris by the "propodeal triangle" in Apidae and possession of dentate processes on the Sphecidae (Brothers 1976). 0= smooth, trian- area absent from dorsal mesopleura (Figs. 6, 14). Although an gular propodeum 1= area apomorphic character in one lineage, it shows (Figs. 2, 5, 6, 8, 9); smooth, triangular varying degrees of development in related present on dorsal propodeum (Figs. 7, 10, taxa. 0= mesopleuron not expanded to form 11). our a dentate process; 1= mesopleuron moder- 13. Petiole ventrally keeled: To knowledge, this has not been used ately expanded; 2=mesopleura with dentate character previously the of but is processes. in defining genera Bethylinae, Presence of posterior transverse propodeal important for separating Goniozus from su- similar taxa in carina (Figs. 2, 6, 7, 9, 10): 0= posterior trans- perficially Odontepyris (see a verse propodeal carina present; 1= posterior key to genera, below). In most Bethylinae, transverse propodeal carina absent. complete longitudinal keel is present on the Presence of median longitudinal propodeal ventral petiole (Fig. 13). This keel is reduced taxa 0= with carina (Figs. 2, 5, 6, 8, 9): 0= median longitu- in several (e. g. Fig. 15). petiole dinal propodeal carina present; 1= median a complete ventral keel; 1= petiolar keel re- longitudinal propodeal carina absent. duced; 2= petiolar keel absent. 10. Presence of discal longitudinal propodeal 14. Expansion of the fore femora: Expansion of for is often en- carinae (Figs. 2, 5, 6, 7, 8, 9, 10): 0= discal the fore femora fossorial use longitudinal propodeal carinae present; 1= countered among the Bethylidae, and also discal longitudinal propodeal carinae absent. occurs in related chrysidoid taxa and other 11. Presence of median propodeal pits or foveae aculeates. Other modifications of the fore in the Aculeata. 0= fore (Figs. 4, 7, 10): The presence of these struc- femora are frequent tures is characteristic of the genus Prosierola. femora strongly expanded; 1= fore femora as 0= median propodeal pits absent; 1= median less strongly expanded, half as wide long. propodeal pits present. 15. Development of the prostigma: The expan- 12. Development of a smooth, triangular area sion of the junction of the subcostal and basal on the dorsal propodeum: In several veins into a secondary pterostigma is charac- bethyline taxa, particularly Goniozus and teristic of the genus Goniozus (Figs. 20, 25), Prosierola, the proximal dorsum of the but also occurs elsewhere within the subfam- Volume 3. 1994 95

Table 2. Character matrix: 96 Journal of Hymenoptera Research

of mesosoma. Fig. 2. Bethylidae: terminology

The character matrix was analysed using the Characters Supporting Monophyly of the of Genera / Clades 'implicit enumeration' option Hennig86, Bethyline procedure which computes the maximum pos- + + sible number of cladograms, or 'trees'. Multistate 1. (node A, Fig. 26): {(Prosierola Odontepyris) 5: + + characters (3: number of maxillary palps; pres- ((Sierola Bethylus) Goniozus))). the ence of notauli; 7: expansion of the mesopleuron; Monophyly of this clade is supported by 3: reduction of 13: reduction of the petiolar keel) were treated following synapomorphies: both as unordered (non-additive) and ordered number of maxillary palp segments from six loss of (additive). to five; 5: loss of notauli; 20: submar- in ginal cell; 21: marginal cell open (reversed Sierola).

RESULTS 2. (node B, Fig. 26): (Prosierola + Odontepyris) Monophyly of this clade is supported by the the 6: of A single cladogram (tree) resulted from following synapomorphies: expansion the scutellar 7: of the analysis, having the following characteristics: pits; expansion 71 retention 8: of a trans- length 35 steps, consistency index and mesopleura; presence posterior effect verse 10: of discal index 80 (Fig. 26). Ordering the data had no propodeal carina; presence on the topology of the tree, only increasing its carinae; 16: rs+m shorter than rs; 17: rs longer the than m. length by one step, and reducing consistency index by two. 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 8. (node H, Fig. 26): Sierola following synapomorphy: 9: presence of a Monophyly of Sierola is supported by 3: max- median propodeal carina. illary palps 4-segmented, as well as by the cell is 4. (node D, Fig. 26): Prosierola closed marginal (21 ) which here a rever- Monophyly of Prosierola is supported mostly sal. by convergent characters, except for the 9. (node I, Fig. 26): Bethylus autapomorphic character (11) presence of Monophyly of Bethylus is supported by the 1: of propodeal pits. following synapomorphies: reduction 5. (node E, Fig. 26): ((Sierola + Bethylus) + Goniozns)) the number of antennal segments from 13 to Monophyly of this clade is supported by a 12; 13: petiolar keel absent. single synapomorphy: 4: loss of the frontal streak. DISCUSSION

6. (node F, Fig. 26): Goniozus Monophyly of Goniozus is supported by a The analysis supports monophyly of our modi- single synapomorphy: 13: reduction of the fied concept of Odontepyris, including Trissomalus petiolar keel. Character 12, presence of a Kieffer, and Parasierola Cameron sensu Kieffer smooth triangular area on the propodeum is a (1914), not Cameron (1883). The currently accepted synapomorphy for Goniozus which occurs as a definitions of most of the remaining genera are homoplasy in Prosierola. also supported by the analysis. 7. (node G, Fig. 26): (Sierola + Bethylus) Trissomalus (Kieffer 1905) was characterised Monophyly of this clade is supported by the in a key, but not formally diagnosed until Kieffer's following synapomorphies: 2: reduction of (1914) generic revision. Odontepyris (Kieffer 1904) the number of labial palp segments from three was characterised mainly by the possession of to two; 22: radial vein sharply angled. dentate processes on the enlarged mesopleura 98 Journal of Hymenoptera Research

mesosoma. 7, Prcsierola mesosoma. Figs. 5-7. 5. Eupsenella sp.: mesosoma. 6, Odontepyris sp.: sp.:

described and Trissomalus Kieffer 1905: 105. Type species Goniozus (Fig. 14). We have examined many transvaalensis Du Buysson, 1888: 354 by subsequent des- undescribed species belonging to the Odontepyris ignation. Holotype female: [SOUTH AFRICA:] I Trissomalus and conclude that the group, degree Transvaal, Hamman's Kraal 1893 (E. Simon) (MNHN) in the of the of variation development mesopleural [examined] syn. n. reduction in processes, coupled with a successive cell — the size of the discoidal (Figs. 22-24) suggests Diagnosis. Antennae 13-segmented. Palpal for the group should be treated as a single genus, formula 5:3. Frontal streak present. Notauli ab- which Odontepyris Kieffer is the oldest available sent. Scutellar foveae well developed. Mesopleuron to treat as new name. The alternative would be expanded, sometimes developed into a dentate all intermediate or discal genera species species-groups. prominence (Fig. 14). Median, and posterior Examination olAnoxus specimens shows that transverse carinae present on propodeum, the in the extent of the genus differs from Bethylus only longitudinal carinae occasionally reduced. Petiole setation of the eyes, other characters are identical with a complete ventral keel. Prostigma present or in the two genera. Furthermore in some Bethylus absent, rs + m reduced, rs elongate. Discoidal cell setation can be found on the species some eyes (e. present, reduced or absent. Marginal and submar- B. amoenus Since we know that the de- curved. g. Fouts). ginal cells absent. Radial vein smoothly is a character that varies — gree of eye setation clearly Included species. Odontepyris argyriae Kurian; interspecifically within other bethyline genera, O. batrae Kurian; O. cameroni (Kieffer) (comb. n. as a distinct we do not see how retaining Anoxns from Trissomalus, holotype examined); O. cirphi genus can be justified, and it is therefore Kurian; O. erucarum (Szelenyi) (comb. n. from ex- synonymised below. Parasierola); O. flavinervis (Kieffer) (holotype amined); O. fuscicrus (Kieffer) (comb. n. from GENERIC SYNONYMY Trissomalus, holotype examined); O. hypsipylae (Kurian) (comb. n. from Goniozus); O. indicus ODONTEPYRIS Kieffer Kurian (comb. n. from Trissomalus); O. moldavicus (Nagy) (comb. n. from Prosierola); O. peringueyi Odontepyris Kieffer 1904: 378. Type species Odontepyris (Kieffer) (comb. n. from Trissomalus); O. flavinervis Kieffer 1904 by original designation. Holo- quadrifoveatus (Muesebeck) (comb. n. from type female: SUMATRA: Pangherang-Pisang x. 90 e iii. Parasierola, holotype examined); O. ruficeps Kieffer; 91 (E. Modigliani) (MCSN) [examined]. 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 carinae 5 Propodeum without well-developed lateral (fig 11). Scutellum without large foveae, with small in Petiole with a forked keel Goniozus — grooves (as Fig. 3). ventrally reduced, (Fig. 15) 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 Species transferred from Anoxus Goniozus, holotype examined); O. zvaterhousei

B comb. n. B. n. , (Kieffer) (comb. n. from Parasierola, holotype ex- boops (Thomson) ; coniceps (Kieffer) comb.

B (Kieffer) comb. n. . amined); O. xanthoneurus (Kieffer) (comb. n. from pilosus Parasierola, holotype examined). Comments. — It might appear questionable to Comments. —Tryapitsin (1978) and Terayama with Bethylus without (1990) mistakenly included the Neotropical synonymise/iHOXKS having examined the type species of either genus. How- bethyline genus Prosierola in their respective keys ever, there has not been any controversy sur- to Russian and Japanese bethylid genera. In each these since Kieffer's case the error was due to misidentification of an rounding genera (1914) ge- neric revision, and we have examined sufficient Odontepyris species. material conforming to the original descriptions and Kieffer's of Anoxus and BETHYLUS Latreille interpretations Bethylus. Bethylus Latreille, 1802: 315. Type species Omalus fusacorms Junne 1807: 301 by subsequent designation (Interna- DISTRIBUTION OF BETHYLINAE tional Commision on Zoological Nomenclature, opin- ion 153). Holotype 'female [not examined, ?lost). Anoxus Thomson 1862: 451. Type species Anoxus boops The most primitive bethyline, Lytopsenella, is Thomson 1862: 452. female Monotypic. Lectotype known only from two extant species, both from (Naturhistonska Rijksmuseet, Stockholm) [not exam- Chile. Three fossil species are known from Baltic ined] syn. n. amber (Evans 1964). Eupsenella is known so far For full synonymy see Gordh & Moczar 1990. only from Australia. —Antennae Diagnosis. 12-segmented. Palpal Odontepyris is predominantly Palaeotropical, formula 5:2. Frontal streak absent. with or Eyes but its distribution spans southeastern Europe to without long hairs. Notauli absent. Scutellar foveae South Africa and northern Queensland. The small narrow, not groove-like. Mesopleuron expanded. genus Prosierola is primarily Neotropical although without carinae. Petiole with the ven- Propodeum one species ranges into the extreme southern tral keel absent. and discoidal cells Prostigma United States; a fossil species from Baltic Amber is absent. and cells absent. Marginal submarginal certainly incorrectly assigned to Prosierola (Brues Radial vein usually sharply angled. 1933). Goniozus, with about 150 described species, 100 Journal of Hymenoptera Research m

Prosierola Goniozus Figs. 8-11. 8, Eupsenella sp.: propodeum. 9, Odontepyris sp.: propodeum. 10, sp.: prapodeum. 11, sp.: propodeum. Volume 3. 1 994 101

12-15. ventral sternite 2. Figs. 12, CWoHf£7n/nssp.:head. 13, Odontepyns sp.: petiole and 14, Odontepyris sp.: side of mesopleuron showing dentate process. 15, Goniozus sp.: ventral petiole and sternite 2. is cosmopolitan although only one species is known described from Baltic and Dominican amber. from the Pacific Islands. Approximately three Sierola contains almost 200 described species, dozen each species were described from the Ori- mostly from Hawaii, with three endemic species ental, Nearctic and Neotropical regions, and about in Australia. The genus underwent tremendous a dozen each from the Palaearctic, Ethiopian and speciation in Hawaii, and a secondary small spe- Australian regions. A number of species have ciation in the Marquesas Islands. Evans (1978) been introduced into other countries as biological suggested that the single Californian species might control agents. A few fossil species have been be introduced, and this is possibly also true of the 102 Journal of Hymenoptera Research

17

19

25

16-25. Fore of various Figs. wings Bethylidae: 16, Lytopsenella sp. 17, Eupsenella sp. ; 18, Sierola sp. 19, Bethylus sp. 20. Coniozus Prosierola (sensu stricto) 21, sp. 22-24, Odonttpyris spp. 25, Goniozus ("Parasierola") sp. Volume 3, 1 994 103

F

26 104 Journal of Hymenoptera Research records of bethylids reared from pupae are dubi- females, and sibling mating is common (Hardy ous, and the most probable explanation is that 1992). Maternal care of the larvae has been ob- "pupal parasites" were reared from larvae which served in Bethylus and Prosierola, and the mother had spun cocoons but had not yet pupated. may subsequently mate again with one of her Evans (1962) reported that Bethylus amoenus male offspring. Maternal care has also been ob- Fouts was reared from an olethreutid moth and a served in Goniozus (Hardy & Blackburn 1991). nitidulid beetle. The latter record was later omit- Adults of both sexes feed on honey in culture, and ted from Evans' (1978) list of host records, and to females have been observed feeding on us seems doubtful. A species tentatively identified haemolymph exuding from the paralysed host. as Goniozus gestroi (Kieffer) was reported by Richards (1955) as being reared from larvae of the ACKNOWLEDGMENTS anobiid beetle, Lasioderma. The rearing was not This work was carried out while the first author was in questioned, only the specific identity of the of a from the Directorate General for Interna- Goniozus, and we therefore accept this record as receipt grant tional Cooperation of the Netherlands Government (DGIS). authentic. Goniozus morindae Kurian (1952), de- thank the for We following the loan of material: J. Casevitz- scribed from a was reared from the single male, Weulersse and M. LaChaise (Museum National d'Histoire of morindae Mani gall acecidomyiid , Asphondyla Naturelle, Pans MNHN), E. De Coninck (Musee Royal de in ofMorinda tinctoria. The record is anoma- l'Afrique Centrale, Tervuren, Belgium), V Rainen (Museo Civico di Stona Naturale, Genoa, MCSN) and M lous, and is the first for a dipteran as a host of a Italy Terayama (Biological Laboratory, Toho Institute of Educa- bethylid. The most recent authentic record of a tion, Tokyo). J. M. Carpenter (American Museum of Natural host for a line is that of Melo hymenopterous bethy History, New York), H. E. Evans (Dept of Entomology, Colo- and Evans (1992) who reported Goniozus rado State University), I. C. W Hardy (Dept of Population microstigmi Evans as being reared from a brood Biology, University of Leiden), A. Menke (Systematic Ento- Lab. and M. mology , USDA) Mickevich (Maryland Center cell of the sphecid Micros tigmus xylicola Melo, for Systematic Entomology, University of Maryland) are a predator of nymphal Thysanoptera. The nests thanked for their comments on the manuscript, and R. W. were in abandoned beetle in beams of an galleries Hodges (Systematic Entomology Laboratory, USDA) for as- a G. exposed roof, and dozen microstigmi females sistance with Lepidoptera systematics. were collected while walking near and entering nests of M. xylicola and M. similis Melo. So far as is known, all Bethylinae are gregari- LITERATURE CITED ous parasitoids, laying a clutch of eggs on each D. 1976. Modifications of the host . The host is stung, sometimes repeat- Brothers, J. metapostnotum and of the in edly, behind the gula. Paralysis of the host larva origin 'propodeal triangle' Hymenoptera Aculeata. Systematic Entomology 1: 177-182. may be temporary or permanent, depending upon Brown, W. L. 1987. Neoclystopsenella (Bethylidae) a synonym the of The number of clutch species wasp. eggs per of Tamnoma (Formicidae). Psyche 94: 337. varies to the size of the as well as according host, Brues, C T. 1907. Notes and descriptions of North American interspecifically (Gordh & Evans 1976), ranging parasitic Hymenoptera IV. Bulletin of the Wisconsin Natu- from one or two to as many as 40. Placement of the ral History Society 5: 96-1 1 1 Brues, C. T. 1933. The parasitic Hymenoptera of the Baltic eggs depends upon the species, eggs being depos- Amber. Part 1. Bernstein-Forschungen 3 (1932): 4-172. ited either intersegmentally or longitudinally, and Cameron, P. 1883. Descriptions of new genera and species of either or hatch dorsally, ventrally laterally. Eggs Hymenoptera. Transactions of the Entomological Society of about two days after oviposition, and the larvae London 1883: 187-197. Du R 1898. de M. E. Simon dans complete feeding in 2 to 5 days. Each larva spins a Buysson, Voyage l'Afrique australe Annates de la Societe cocoon on the substrate near the host remains. The (janvier-avril 1893). Entomologiaue de France 66: 351-363. varies from 8-14 pupal period days, depending Evans, H. E. 1962. The genus Bethylus in North America upon the species and ambient temperature. Males Breviora 16: 1-12. emerge a day or two before females are ready to Evans, H. E. 1964 A synopsis of the American Bethylidae the Museum Com- eclose; they have been observed chewing into a (Hymenoptera: Aculeata). Bulletmof of parative Zoology, Harvard Unwersity 133:67-151. cocoon containing a female and mating with her Evans, HE. 1978. The Bethylidae of America north of Mexico. while she is still teneral. The from a clutch progeny Memoirs of the American Entomological Institute 27: 1-332. usually consists of a single male and a number of Volume 3. 1994 105

S. 1988. Version 1. User's Farns, J. Henmg86 Reference. 5. Aculeata). Proceedings of the Entomological Society of Wash- Manual 17 pp. ington 89: 356-358. Gordh, G. and H. E. Evans. 1976. A new species of Gomozus Kurian, C. 1952. Descriptions of four new and record of one imported into California from Ethiopia for the biologi- known Bethyloidea (Parasitic Hymenoptera) from In- cal control of pink bollworm and some notes on the dia. Agra University journal of Research 1: 63-72. taxonomic status of Parasierola and Gomozus (Hy- Kurian, C. 1955. Bethyloidea (Hymenoptera) from India menoptera: Bethylidae). Proceedings of the Entomological Agra University Journal of Research 4: 67-155. Society of Washington 78: 479-489. Latreille, P. A. 1802. Histoire Naturelle des Crustaces et Insectes. Gordh, G. and L. Moczar. 1990. A catalog of the World Tome 3, 1-467. Bethylidae (Hymenoptera: Aculeata). Memoirs of the Melo, G. A. R. and H. E. Evans. 1993. Two new Mtcrostigmus American Institute 46: 1-364. Entomological species (Hymenoptera, Sphecidae), with the descrip- I. C. W. 1992. sex allocation Hardy, Non-binomial and brood tion of their parasite, Gomozus microstigmi sp. n. Proceed- sex ratio variances in the parasitoid Hymenoptera. Oikos ings of the Entomological Society of Washington 95: 258-263. 65: 143-158. Menke, A. S. 1993. Notauli and parapsidal lines: just what are Hardy, I. C. W. and T. M. Blackburn. 1991. Brood guarding they? Sphecos 24: 9-12. in a bethylid wasp. Ecological Entomology 16: 55-62. Muesebeck, C. F. W. 1940 Two new hymenopterous para- 1934. Hempel, A. A Prorops nasuta Waterston no Brasil. sites of sugarcane borers in India. Proceedings of the do 5: Archizvs lnstituto Bwlogico Sao Paolo 197-212. Entomological Society of Washington 42: 120-122. L. 1807. Junne, Nouvelle methode de classer les Hymenopteres et Nagy, C. G. 1976. Bethylidae (Hymenoptera) parasitizing les 1-319. Dipteres. J. J. Paschaud, Geneva. orchard caterpillars. Revue Roumame de Biologie, Serie de Kieffer, J. J. 1904. Description de nouveaux Dryininae et Bwlogie Animate 21: 103-108. du Musee de Genes. Annali Museo O. W. 1955. the of Israel. Bethylinae Civique Richards, On Bethyloidea (Hymen. ) Civico di Storm Naturale di Genova 41: 351-412. Bulletin of the Research Council of Israel 4: 357-359. 1905. de Kieffer, J. J. Description nouveaux proctotrypides Sorg, M. 1988. Zur phylogenie und systematik der Bethylidae exotiques. Annales de la Societe Saentiftque de Bruxelles29: (Insecta: Hymenoptera: Chrysidoidea). Geologisches 95-142. lnstitut der Umversitaet zu Koeln Sonderveroeffentlichungen Kieffer, J. J. 1907 (1906). Beschreibung neuer in 63: 1-146. Naturhistorischen Museum zu Hamburg aufbewahrter Terayama, M. 1990. Keys to the Japanese Bethylidae (Hy- Proctotrypiden and Evaniiden. Berliner Entomologische menoptera, Aculeata) I. Subfamilies and Genera. Bulle- Zeitschrift 51: 258-278. tin of the Toho Gakuen 5: 19-43. 1911. Kieffer,]. J. Nouveaux Bethylides et Dry mides exotiques Thomson, C. G. 1862. Svenges Proctotruper (Fortsattning). British du Museum du Londres. Annales de la Societe Svenska Vetenskapsakadenuen. Stockholm. Oefrersigt af K. Scientifique de Bruxelles 35: 200-233. Academinens. Forhandlmgar. 18: 451-453. 1914. Kieffer, J. J. Hymenoptera, Proctotrupoidea, Bethylidae Tryapitsyn, V. A. 1978. [Identification of the insects of the Das Tierreich 41: 1-595. European part of the USSR. Volume 3. Hymenoptera. Krombein, K. V. 1987. notes on the Faune Synonymic Bethylidae Bethyloidea, Bethylidae. ] Opredeliteh SSSR 120: 3- described by V. de Motschulsky (Hymenoptera: 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 . — were in confined that Haemolymph proteins, especially vitellogenin ( Vg), investigated Apis mellifera workers, 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 control exerted by the queen over the workers represents an additional factor influencing Vg The of the A inhibits vitellogenin—(Vg) Apis mellifera, — synthesis. queen pheromone oocyte precursor of vitellin the major protein of yolk development in the workers. As a result Vg is not is a glycolipoprotein produced in the fat body incorporated into the oocytes, although Vg is de- (Harnish and White 1982; Trenczek and Engels tected in the workers haemolymph. However, in 1986; Shapiro et al. 1988; Wheeler and Kawooya queenless colonies the Vg titre of workers in- 1990). In most insects, Vg synthesis may be con- creases, reaching a level similar to that observed in trolled by the diet, which obviously provides the the queens, followed by oocyte growth and ovipo- materials and energy needed for this process sition (Engels 1974; Engels and Fahrenhorst 1974). (Bianchi and Pereira 1987; Bownes 1989; Bownes How Vg synthesis is regulated in A. mellifera and Reid 1990). However, other factors in addition still remains an interesting question, and aspects to diet are also involved. Many studies have been of this regulation have been studied in queens, conducted on the role of neurosecretory cells (Elliott workers and drones. In queens Vg synthesis does and Gillott 1978) and neurohormones (Keeley and not depend on functional corpora allata and is not Mckercher 1985; Keeley et al. 1988; Girardie et al. mediated by JH (Engels and Ramamurty 1976; 1992), and of juvenile hormones and ecdysteroids Kaatz 1985). Similarly, JH topical application does in the regulation of Vg synthesis (Adams et al. not increase Vg synthesis in drones (Trenczek et al. 1985; Borowsky et al. 1985; Hagedorn 1985; 1989). But in workers, Rutz etal. (1976) and Flurief Schwartz et al. 1985; Wojchowsky and Kunkel al. (1977) observed that low JH doses applied on 6 1987; Adams and Filipi 1988; Keeley et al. 1988; Ma day old workers stimulate Vg synthesis whereas et al. 1988; Roseler and Roseler 1988; Wyatt 1988; high doses have an inhibitory effect. Furthermore, Bownes 1989; Bownes and Reid 1990; Davis et al. Rutz et al. (1976) observed a correlation in vivo 1990; Hatakeyama and Oishi 1990; Yin et al. 1990; between low JH titre and Vg synthesis. Within the Agui et al. 1985, 1991; Don-Wheeler and Engelman first few days after worker emergence, character- 1991;Sochaeffl/. 1991). ized by low JH titres in haemolymph there is an In Apis mellifera, a highly eusocial insect, the increase in Vg synthesis. After this period Vg 108 Journal of Hymenoptera Research

and or synthesis decreases while JH titre increases. This bread) and candy (powdered sugar honey), increase in JH titre was confirmed by Huang et al. only candy. (1991) and the temporal changes in Vg titre ob- In three experiments, three naturally mated served by Rutz et al. (1976) were similar to those queens aged 60-90 days were removed from the confined with of related by Engels et al. (1990). respective colonies and groups The investigation of factors that affect A. 150-200 newly emerged workers. These groups to formed workers and a were confined for mellifera Vg synthesis can help understand how by queen fed on the mixture of beebread and this protein is regulated. In the present work, the 15-16 days, and effect of some factors such as changes in social candy. Water was also supplied. environment (confinement of workers with or without a queen) and diet (protein-rich or not) and Treatment of confined workers JH or PPN (pyriproxyfen, 2-[l-methyl-2(4- with JHIII, JHI and PPN phenoxyphenoxy) ethoxyl] pyridine, a JH ana- for logue, were investigated not only on worker Vg Some groups of 150-200 workers confined or synthesis but also on other haemolymph proteins. 6 days were treated with JHIII, JHI PPN applied topically to the abdominal cuticle. Each worker MATERIALS AND METHODS received 1 ul of a hormone solution in acetone at a concentration of 1 ug/ul, administered in two the first after Apis mellifera equal doses, immediately emergence, before the introduction into the cages, and the We used "wild type" Africanized Apis mellifera other on the third day after rapid anesthesia with m. worker were treated bees (hybrids of European A. m. ligustica, A. gaseous nitrogen. Two groups carriica, A. m. mellifera and the African bee A. m. with JHIII, two groups with JHI and nine groups scutellata) from the Experimental Apiary of the with PPN. Department of Genetics, Faculty of Medicine of Three control groups were prepared in paral- Ribeirao Preto, University of Sao Paulo. lel: the first consisted of marked workers reintro- Combs containing workers ready to emerge duced into the colony (control a), the second formed were removed from colonies and placed in an by confined workers treated with two 1 ul doses of incubator whose temperature (34°C) and R.H. acetone on the first and third days, respectively (80%) were similar to those in the colony. The (control b). This group was also submitted to a workers that emerged within 15-20 hours were rapid anesthesia with gaseous nitrogen at third collected. About 100 newly emerged workers were day, immediately before acetone treatment. The marked on the thorax and put back into a small third group consisted of untreated confined work- colony, formed by a queen, approximately 3000 ers (control c). All worker groups, except control a, workers (hive and forager bees), brood (eggs, lar- were allowed to feed ad libitum on the mixture of vae and pupae) and, sometimes, a few drones. beebread and candy. Water was also supplied. Presence of and pollen into the combs were also checked. Haemolymph

Confinement For collecting haemolymph the workers were anesthetized with gaseous nitrogen and immobi- The remaining workers were separated into lized on dissection plates. Haemolymph was ex- groups of 150-200 and immediately submitted to tracted through a small superficial incision in the confinement in 8 x 11 x 13 cm wooden cages with dorsal cuticle between the 2nd and 3rd tergites. a sliding glass door and a screened bottom. The Haemolymph was withdrawn from 6 day old workers in these cages were placed in an incubator and 15-16 day old confined workers that were fed at 34°C and 80% R.H. and confined for 6 days on different diets, treated or not with hormones or (short confinement) or 15-16 days (long confine- acetone and mantained with or without a queen confined bees received food ment). The water and (Table I). Haemolymph was also extracted from ad libitum. The diet consisted of 50% sugar in water colony reared workers (6 or 15-16 day old) and (syrup), a mixture of pollen from the comb (bee- from newly emerged workers before confining or Volume 3, 1 994 109 returning them to the colony. RESULTS Haemolymph pools were prepared from groups of at least 80 workers obtained from the Confinement blocks the increase same confinement cage. Similar pools were pre- of haemolymph Vg pared with the haemolymph of at least 20 workers of the same age maintained in the colony and of 1 2 The confinement of workers for a period of 6 newly emerged workers. Phenylthiourea was days after emergence prevented the increase of Vg that in added to the pools. Haemolymph pools were cen- titre in haemolymph normally occurs work- trifuged at 3080 g for 10 min at 0°C and the super- ers in colonies. This occurred even when bee- natant was stored at -20°C. bread, the natural source of bee protein was sup- plied to the confined bees. In some of the Rocket Immunoelectrophoresis haemolymph pools an extremely low Vg titre was detected (Fig. 1, wells 5 and 6), but, on average, Vg Immunoelectrophoresis was used for the quan- titres in these pools corresponded to 6.0 ± 2.1% titative determination of the vitellogenin fraction (Table II) of that present in workers maintained in anti- the colonies natural conditions 1 wells in haemolymph. Monospecific vitellogenin under (Fig. , serum produced in rabbits (Simoes 1980) was 1 and 2). added at a 1% concentration to 1% agarose gels A small increase in Vg titre was observed in prepared with 0.06 M Tris-HCl buffer, pH 8.6. 15-16 day confined workers (Fig. 1, wells 3 and 4). Immunoelectrophoresis was carried out at 10°C This only occurred when the workers were fed for 16 hours, at 0.08 V/cm gel. The same buffer beebread and candy. Even with this diet, the Vg used in the gel was used in the electrode compart- concentration was much lower than in workers of ments at a concentration of 0.3 M. The gels were the same age maintained in the colonies (Fig. 1, stained with Coomassie Blue R-250. The height of wells 1 and 2). the peaks detected (reported as mm) was propor- tional to the amount of antigen. The values ob- Confinement changes the protein pattern of tained for confined workers were compared to haemolymph obtained by SDS-PAGE those for workers of the same age maintained in colonies. The Vg peaks detected in colony reared The pattern of soluble haemolymph proteins workers (control c) was considered to be 100%. from six day confined workers (Fig. 2, lane 4) differed from that observed in workers main- Total protein tained in colonies during the same period of time 2, lane 3). In the confined bees, in addition to Protein concentration in the haemolymph (Fig. the fact that the Vg band (as determined by pools was determined using bovine serum albu- Trenczek et al. 1989) was very weak, the a and c min as a standard (Bradford 1976). polypeptides were not observed, whereas the b formed a band. This SDS-PAGE polypeptide strong polypep- tide corresponded to a weak band in the workers Soluble haemolymph proteins were separated living in colonies. These were the most obvious by SDS-PAGE according to the method of Laemmli differences, however differences between low (1970) except that SDS was not used in the separat- molecular weight polypeptides, were also ob- ing and stacking buffers. A 5-12% acrylamide served. gradient was used on a 0.7-mm thick gel. Electro- In Fig. 2, lanes 1, 2 correspond to the phoresis was run at 12mA constant current at 10°C haemolymph protein pattern of newly emerged until bromophenol blue tracking dye reached the workers, collected immediately before the bees bottom of the slab. were confined or returned to the colony. This in Haemolymph samples (5 \i\) from confined pattern changed as the bees maintained the and newly emerged workers (diluted 1: 2, v/v, in colony aged (Fig. 2, lane 3). However, the changes sample buffer) and from workers maintained in depended on the social environment as shown by the colonies (diluted 1: 20 or 1: 40, v/v) were the protein pattern of confined workers (Fig. 2, applied to the acrylamide gel. lane 4). 110 Journal of Hymenoptera Research

^*t|l'"iWWwwi«AM*' vg

b c

J |

4 5 6 7 8 9 M(kDa) 2 3 4 5 6

1. Rocket of 2. Fig. Immunoelectrophoresis worker Fig. SDS-PAGE (5.0-12%). Coomassie Blue staining. Pat- 150 antiserum in 15 ml haemolymph. |il queen egg agarose terns of worker haemolymph proteins. (1, 2) Newly emerged 2 ul of a well. gel. Antigen: haemolymph pool per Staining: workers. Haemolymph diluted 1: 2 (v/v) in sample buffer; (3, Coomassie Brilliant Blue R-250. Adult workers (6 days) main- 7) 6-day old workers maintained naturally in the colony tained in the confined for naturally colony (1, 2), 15-16 days Haemolymph diluted (3) 1: 20 (v/v) and (7) 1! 40 (v/v) in and fed beebread (3, 4) and candy and confined for 6-days (5, sample buffer. Workers confined for 6 days after emergence and fed the diet. 6) same and treated with PPN (10), JHI (5,9), JHIII (6), or acetone (8, or 11), untreated (4). Haemolymph diluted 1: 2 (v/v) in sample buffer. (M) - molecular in kDa to The haemolymph pattern of workers confined weights according marker Note: Columns should be numbered 1-11. for proteins. 15-16 days and fed candy and pollen (Fig. 3, lanes 5 and 6) did not differ from that obtained when fed on the after a short confinement (Fig. 2, lane 4). However, beebread and candy mixture, as can be seen a discrete increase was observed in the Vg band of by Immunoelectrophoresis Fig. 1, wells workers confined for 15-16 days. The value of diet 3 and 4. However if a queen is confined together with a of 150-200 protein content for protein synthesis is clearly group workers during this same time from shown in Fig. 3, lanes 3, 4, corresponding to the interval, emergence until 15-16 days, protein pattern of a haemolymph pool from work- the workers will not have Vg in the haemolymph. the ers maintained on a carbohydrate (syrup) diet for Thus, queen effect on Vg synthesis could be observed even in a 15-16 days. In the same figure, the protein pattern different environment condi- of haemolymph from confined workers can also tion, i.e. that established by confinement. The be compared with that of newly emerged ones haemolymph protein SDS-PAGE pattern of these (lanes 7, 8) and with that of workers maintained in workers was similar to that of bees confined in the absence of a the colony for 15 days (lanes 1, 2). queen, except for a weak Vg band present in the latter (results not shown). There is no Vg in workers confined for 15-16 days with a queen PPN induces Vg titre increase in workers confined for 6 days Workers confined for 15-16 days in the ab- sence of a queen had Vg in their haemolymph Workers confined for six days after emer- 111 Volume 3. 1994

gence, fed beebread and candy and treated with PPN consistently showed a significant increase (p<0.001) in haemolymph Vg concentration (Fig. wells ^-Htm* 4, wells 3, 4, and Fig. 5, 3, 4) when compared 180 vg Wt* to confined workers treated with acetone (Fig. 4 wells 1, 2, 7, 8; Fig. 5 well 1) or untreated (Fig. 4, 116 wells 9, 10). This increase corresponded to ap- .97 proximately 17% of the Vg levels in the b haemolymph of workers naturally maintained in •66 the colonies (Fig. 4 wells 11, 12; Fig. 5, well 5). To calculate Vg percentage, we used a total of 9 pools derived from 9 experiments, -45 haemolymph each consisting of 150-200 workers treated with PPN. In this case, Vg concentration in control a — 29 (workers maintained in the colonies) was consid-

ered to be 100% (Table II). Table II also shows that I 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 12 3 4 5 6 7 8 M(kDa) (6.0 ±2.1%). The Vg titre detected in workers treated with Blue Pat- similar to that detected in workers con- Fig. 3. SDS-PAGE (5.0-12%), Coomassie staining. PPN was old terns of worker haemolymph proteins. (1, 2) 15-day fined for 15-16 days and feeding on pollen and workers maintained naturally in the colony. Haemolymph concentra- candy (Fig. 1, wells 3 and 4). This Vg diluted 1: 20 (v/v) in sample buffer; (3, 4) workers confined tion, however, was never detected in controls c for 15-16 days after emergence and fed syrup or (5, 6) bee- confined for six and on bread and candy; haemolymph diluted 1: 2 (v/v) in sample (workers days feeding diluted or confined for six buffer; (7, 8) Newly emerged workers. Haemolymph pollen and candy) b (workers - molecular in kDa 1: 2 (v/v) in sample buffer. (M) weights days, feeding on pollen and candy and treated according to marker proteins.

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

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) I under the same experimental conditions as those o with PPN did not increase Vg concentration in 1 2 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 wells 8 and Fig. 5. Rocket Immunoelectrophoresis of worker (Fig. 4, 1, 2, 7, 150 ul antiserum in 15 ml haemolymph. queen egg agarose Fig. 5, well 1 as showed in Table IV). However, the 2 of a well. gel. Antigen: (il haemolymph pool per Staining: Vg peaks obtained from workers treated with JHI, Coomassie Brilliant Blue R-250. Adult workers confined for a hormone not synthesized by Apis mellifera 6 days after emergence, fed beebread and candy and treated (Robinson et al. 1987), were found to be with acetone (control b, well 1), JHI (well 2), or PPN (wells 3 slightly than the obtained for workers treated and 4). Well 5 corresponds to the Vg of adult 6-day old higher peaks workers maintained naturally in colonies. with acetone or with JHIII, the natural hormone of these bees. This result should be considered with with acetone). Thus, under conditions of prolonged caution since it is based only on the data obtained confinement, there is an increase in Vg concentra- for two haemolymph pools from workers treated tion in haemolymph during the second week of with JHI or JHIII. But we do not exclude the confinement, but this increase never occurs dur- possibility of this homologue (JHI) being more ing the first week of confinement unless the work- effective on Apis mellifera than JHIII. ers are treated with PPN. JHI and JHIII did not change the SDS-PAGE PPN does not change the SDS-PAGE protein pattern protein pattern The haemolymph protein pattern of workers The protein pattern (Fig. 2) from workers con- confined for six days and treated with JHI (Fig. 2, fined for six days and treated with PPN (lane 10) lanes 5, 9) or JHIII (Fig. 2, lane 6) was similar to that was similar to that observed in the acetone-treated observed in the controls treated with acetone (Fig. (lanes 8, 1 1) and untreated controls (lane 4), except 2, lanes 8, 11) or untreated (Fig. 2, lane 4). by an increase in Vg band. Volume 3, 1 994 113

DISCUSSION produced Vg when fed beebread and candy. How- ever, when the workers were confined with a diet on this Effect of confinement, queen presence and queen during same period of time, no Vg was haemolymph protein composition detected in the haemolymph. Perhaps, the experi- mental conditions employed by Engels et al. 1990, Our results show that the normal Vg titre in with a piece of comb (with brood?) within the workers mainly depends on social environment confinement cage, provided a more favorable en- established in the colony. When workers were vironment (maybe more similar to that of a normal removed from the colony, and maintained during colony) that permitted Vg synthesis even in the 6 days in an appropriate environment (where presence of a queen. For comparison, other condi- mortality was practically zero) and on a protein- tions employed in both experiments should also rich diet, profound physiological changes occurred, be taken into account, such as the age of the queens that inhibited Vg titre and probably Vg synthesis. used, the size of the worker population and the The initiation of Vg synthesis in these workers can time of confinement. be observed if confinement is lenghthened for 15- Confinement affects not only Vg synthesis, 16 days. But the haemolymph Vg peak detected by which is also controlled by the queen and by the immunoelectrophoresis in these workers (after food available, but also the synthesis of other 15-16 days of confinement) is smaller than that haemolymph proteins, as determined by SDS- observed in colony reared workers. The onset of PAGE. The polypeptide a showed marked reduc- Vg in the haemolymph of 15-16 days confined tion, the polypeptide c was not detected, whereas workers is on administration of a dependent pro- another, b, showed markedly increased titer (Figs. tein-rich diet: Vg is not detected in these confined 2 and 3). This indicates that confinement can si- workers fed a diet without protein such as syrup, multaneously provoke contrasting gene expres- or a low protein content diet, such as candy made sions. This was observed in confined workers fed with honey. Besides if a queen is confined with the proteic or non-proteic diets. Workers maintained workers for 15-16 days no Vg is produced al- on a syrup diet had a lower protein content, but a rich diet and similar though protein (beebread pollen) protein pattern in the haemolymph (Fig. 3) available. had been when compared to workers maintained on a proteic The influence of factors related to the social diet. Thus the social interactions and not the pro- environment, on Vg synthesis has been studied in tein supply are responsible for the contrasting social Hymenoptera, especially the interaction gene expressions mentioned above. among individuals of different castes in the colo- nies. Reproductively active queens characteristi- Action ofPPN, JHIII and ]H1 on haemolymph protein cally inhibit Vg synthesis and egg-laying in other composition changed by confinement females capable of reproduction. This fact has been well in al. documented A. mellifera (Engels et A significant (p<0.001) increase in Vg titre 1990), Melipona (Engels and Imperatriz-Fonseca ocurred in workers with confinement-induced Vg 1990), and Camponotus festinatus (Martinez and deficiency, after treatment with PPN. However Wheeler other social insects. 1991) among We the observed induction was partial. Apparently that A. verify mellifera queens can inhibit Vg syn- other factors are also involved in the regulation of thesis in workers even under conditions of con- normal Vg synthesis. The action of these factors finement, i.e., far from the normal colony environ- may be prevented or impaired by the confinement ment. Engels et al. (1990) maintained groups of 25 conditions employed. workers for three weeks confined from emergence PPN seems to be specific for Vg synthesis, on a of in piece comb, the presence and absence of since the haemolymph SDS-PAGE pattern of the a detected queen. They that both worker groups other proteins in confined workers apparently did had in the Vg haemolymph, but the Vg titre of not change when treated with this JH analogue. workers confined in the absence of a queen was However as the small increase in Vg titre induced than that of higher workers confined in the pres- by PPN was not reflected in total protein measure- ence of a In our that queen. experiments, we found ment we can not exclude the possibility that PPN workers confined without a queen for 15-16 days also partially inhibits other haemolymph 114 Journal of Hymenoptera Research

protein(s). ACKNOWLEDGMENTS Contrary to what is observed with PPN, the natural JHIII (Robinson et al. 1987) and its homo- This work was supported by grants from the Fundacao de Amparo a Pesquisa do Estado de Sao Paulo (FAPESP) logue JHI did not significantly increase Vg con- Proc. 91/2587-8. centration in haemolymph. JHIII and JHI also did We thank Dr. H Oouchi, Sumitomo Chemical Co., not seem to the modify haemolymph protein pat- Ltd, Osaka, Japan, for providing the pynproxyfen, and PR. tern induced by confinement. We should consider Epifanio and L. R. Aguiar for the technical assistance We that PPN may have caused an increase in Vg titre thank also Dr. David De Jong for correcting the English. in haemolymph by being a more potent analogue. PPN is considered to be one of the most effective juvenile hormone analogues known for locusts LITERATURE CITED (De Kort and Koopmanschap 1991), with a strong T. S. and P A. 1988. juvenilizing effect on these insects. PPN has a Adams, Filipi. Interaction between hormone, the - strong effect when ]uvenile 20-hydroxyecdysone, corpus morphogenetic topically ap- - cardiacum allatum complex, and the ovaries in regulat- plied to A. mellifera larvae. Larvae treated with 1 fig ing vitellogenin levels in the housefly, Musca domestica. but the show dras- undergo metamorphosis, pupae journal of Insect Physiology 34: 11-19. tic in in the changes pigmentation, especially eyes Adams, T. S., H. H. Hagedorn and G. D. Wheelock. 1985. and thorax, and can also die before emergence Haemolymph ecdysteroid in the housefly, Musca domestica, and its with (Bitondi et al., unpublished data). Similarly to during oogenesis relationship vitellogenin levels, journal of Insect 31: 91-97. what occurs in Locusta, the effect of PPN on A. Physiology Agui, N., S. Izumi and S. Tomino. 1985. The role of ecdysteroids mellifera larvae and is more drastic when pupae and juvenoids in vitellogenin levels and follicle devel- to the effects induced compared by JHIII or JHI. opment in the housefly, Musca domestica. Applied Ento- Kaatz (1985) proposed a model of regulation mology and Zoology 20: 179-188. Agui, N., T. Shimada, S. Izumi and S. Tomino. 1991. Hor- of Vg synthesis in A. mellifera queens. According to monal control of vitellogenin mRNA levels in the male this model, Vg synthesis may be influenced by JH and female housefly, Musca domestica. Journal of Insect but also and a by ecdysteroids by haemolymph Physiology 37: 383-390. factor. These factors mentioned Kaatz by (1985) Bianchi, A. G. de and S. D. Pereira. 1987. Time of synthesis of may be involved in the regulation of normal Vg Musca domestica vitellogenin during the first gonotrophic and synthesis in workers. However, the simple adap- cycle. Comparative Biochemistry, Physiology 86B(4): 697-700. tation to workers of the model proposed for queens Borovsky, D., B. R. Thomas, D. A. Carlson, L. R. Whisenton should be considered with caution since some and M. S. Fuchs. 1985. Juvenile hormone and 20- data obtained in studies on Vg synthesis regula- hydroxyecdysone as primary and secondary stimuli of tion in queens indicate that this regulation may v\te\\oger\esism Aedes aegypti. Archwes of Insect Biochem- and 75-90. differ between the two castes (Engels et al. 1990). istry Physiology 2'. Bownes, M. 1989. The roles of juvenile hormone, ecdysone In summary the absence of one or more types and the ovary in the control of Drosophila of social interaction induced confinement vitellogenesis. by Journal of Insect Physiology 35(5): 409-413. caused a reduction of total content in protein Bownes, M and G. Reid. 1990. The role of the ovary and haemolymph, induction and repression of protein nutritional signals in the regulation of fat body yolk in synthesis and impairment of Vg synthesis in A. protein gene expression Drosophila melanogaster. jour- nal of Insect Physiology 36(7): 471-479. mellifera workers. Only haemolymph Vg titer can Bradford, M.. 1976. A rapid and sensitive method for the be partially recovered by PPN treatment of con- quantification of microgram quantities of protein utiliz- fined workers. ing the principle of protein dye binding. Analytical Bio- chemistry 72: 248-254. R. T. Davis, E., 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 Neuroendocrine and juvenile hormone effects on fat - male produced vitellogenins of Leucophaea maderae. body protein synthesis during the reproductive cycle in Journal of Insect Physiology 37(12): 869-882. female Blaberus discoidalis cockroaches. General Com- Elliot, R. H. and C. Gillot. 1978. The neuro-endocrine control paratwe Endocrinology 72: 364-373. of protein metabolism in the migratory grasshopper, Laemmli, U. K. 1970. Cleavage of structural proteins during

Insect 24: the of the head of T . Nature Melanoplus sangumipes. Journal of Physiology assembly bacteriophage 4 119-126. 227: 680-685. -Z. Engels, W. 1974. Occurrence and significance of vitellogenins Ma, M, J. Zhang and H. Gong. 1988. Permissive action of in female castes of social Hymenoptera. American Zoolo- juvenile hormone on vitellogenin production by the gists 14: 1229-1237. mosquito Aedes aegypti. Journal of Insect Physiology 34: - Engels, W. and H. Fahrenhorst. 1974. Alters - und kasten 593-596. - spezifische Veranderungen der Haemolymph Protein Martinez, T. and D. Wheeler. 1991. Effect of the queen, brood - Spektren bei Apis mellifica. Wilhelm Roux' Archiv 174: and worker caste on haemolymph vitellogenin titre in 285-296. Camponotus festinatus workers. Journal of Insect Physiol- Engels, W. and V. L. Imperatriz-Fonseca. 1990. Caste devel- ogy 37(5): 347-352. opment, reproductive strategies, and control of fertility Robinson, G. E., Strambi, A., Strambi, C, Paulino-Simoes, in In: honey bees and stingless bees. Engels, W. (ed.) Z.L., Tozeto, SO., Barbosa, J. M. N. 1987. Juvenile hor- Social Insects. An evolutionary approach to castes and repro- mone titers in Africanized and European honey bees in duction. Springer-Verlag, Berlin, ppl67-230. Brazil. Gen. Comp. Endocrinol. 66: 457-459. Engels, W., H. Kaatz, A. Zillikens, Z. L. Paulino-Simoes, A. Roseler, P. F. and I. Roseler. 1988. Influence of juvenile Trube, R. Braun and F. Dittrich. 1990. Honey bee repro- hormone on fat body metabolism in ovanolectomized duction: vitellogenin and caste-specific regulation of queens of the bumblebee, Bombus terrestris. Insect Bio- fertility. In: M. Hoshi and O. Yamashita (eds.) Advances chemistry 18(6): 557-563. in invertebrate reproduction 5, 495-502. Elsevier, Rutz, W., L. Gerig;, H. Wille and M. Luscher. 1976. The Amsterdam. function of juvenile hormone in adult worker honey- Engels, W and P. S. Ramamurty. 1976. Effects of carbon bees, Apis mellifera. Journal of Insect Physiology 22: 1485- dioxide on vitellogenin metabolism in unmated queen 1491. Research 15: 3-10. B. honeybees. Journal of Apiculture Schwartz, M. B., T. J. Kelly, R. Imberski and E. C. Rubenstein. Fluri, P., H. Wille; L. Gerig and M. Luscher. 1977. Juvenile 1985. The effects of nutrition and methoprene treatment hormone, vitellogenin and haemocyte composition in on ovarian ecdysteroid synthesis in Drosophila winter worker honey bees (Apis mellifera). Experientia 33: melanogaster. Journal of Insect Physiology 31: 947-957. 1240-1241. H. Shapiro, J. P., J. Law and M. A. Wells. 1988. Lipid transport A. Girardie. 1992. Girardie, J., O. Richard and Time-depen- in insects. Annual Rei'iew of Entomology 33: 297-318. dent variations in the activity of a novel ovary maturat- Simoes, Z. L. P. 1980. Estudo da vitelogenina e da vitelina em

ing neurohormone from the nervous corpora cardiaca Apis mellifera L. (Hymenoptera : Apoidea). Tese de in the during oogenesis locust, Locusta migratoria doutoramento, Universidade de Sao Paulo, pp.: 1-108. Insect migratorwides. Journal of Physiology 38: 215-221. Socha, R., J. Sula, D. Kodrik and I. Gelbic. 1991. Hormonal Hagedorn, H. H.. 1985. The role of ecdysteroids in reproduc- control of vitellogenin synthesis in Pyrrhocoris apterus tion. In Insect and Insect Comprehensive Physiology, Biochemistry (L.) (Heteroptera). Journal of Physiology 37( ): 805- Pliarmacology. (Edited by G. A. Kerkut and L. I. Gilbert), 816. vol. 205-262. 8, pp Pergamon Press, Oxford. Trenczek, T. and W. Engels. 1986. Occurrence of vitellogenin D. Harnish, G. and B. N. White. 1982. Insect vitellins: identi- in drone honeybees. International Journal of Invertebrate fication, purification, and characterization from eight Reprodution and Development 10: 307-311. orders. Journal of Experimental Zoology 320: 1-10. Trenczek, T., A. Zillikens and W. Engels. 1989. Developmen- Hatakeyama, M. e K. Oishi. 1990. Induction of vitellogenin tal patterns of vitellogenin haemolymph titre and rate of and maturation of synthesis transplanted previtellogenic synthesis in adult drone honey bees (Apis mellifera). eggs by juvenile hormone III in males of the sawfly, Journal of Insect Physiology 35: 475-481. Athalia rosae. Insect 36: 791-797. D. E. K. Journal of Physiology Wheeler, and J. Kawooya. 1990. Purification and Z. G. E. S. S. K. C. Huang, Y., Robinson, Tobe, J. Yagi, Strambi, characterization of honey bee vitellogenin. Archives of A. Strambi and B. Stay. 1991. Hormonal regulation of Insect Biochemistry and Physiology 14: 253-267. behavioral in the bee is development honey based on Wojchowski, D. M. and J. G. Kunkel. 1987. Purification of two changes in the rate of juvenile hormone biosynthesis. distinct oocyte vitellins and identification of their corre- Journal of Insect Physiology 37(10): 733-741. sponding vitellogenins in fat body and hemolymph of H. H.. 1985. Kaatz, Initiation und Regulation der Vitellogenin- Blaberus discoidalis. Insect Biochemistry 17: 189-198. bei der Synthese Bienenkbnigin (Apis mellifera L.). Inau- Wyatt, G. R.. 1988. Vitellogenin synthesis and the analysis of gural-Dissertation, Universitat Tubingen, pp. 146-159. juvenile hormone action in locust fat body. Canadian L. L. S. C. Keeley, and McKercher. 1985. Endocrine regula- Journal of Zoology 66: 2600-2610. tions of ovarian in the maturation cockroach Blaberus Yin, C. -M„ B. -X. Zou, S. -X. Yi and J. G. Jr. Stoffolano. 1990. discoidales. Comparatwe Biochemistry and Physiology 80A: Ecdysteroid activity during oogenesis in the black blow- 115-121. fly, Phormia regina (Meigen). Journal of Insect Physiology Keeley, L. L., S. M Sowa, T. K. Hayes and]. Y. Bradfield. 1988. 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 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

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

— is are a is for Abstract. The agile species group of Pison redescribed, the 12 included species reviewed, and key provided Palearctic is identification. The agile group is restricted to the Oriental and eastern Regions, although one species, koreense, 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 CAS California Academy of Sciences, San Francisco,

California (W. J. Pulawski). The genus Pison, which contains nearly 200 NMNH Nationaal Natuurhistorisch Museum, Leiden, The Netherlands van described species (Bohart and Menke 1976; Menke (C. Achterberg). OUM Hope Entomological Collections, University Mu- 1988), is well represented in all faunal regions seum, Oxford, England (C. OToole). except North America where only an adventive USNM U.S. National Museum, Washington, DC, USA (K. is known. species, koreense, V. Krombein, A. S. Menke). In this paper I review the agile species group ZIN Zoological Institute, Russian Academy of Sciences,

St. Russia I. which contains 12 species, five of which are new. Petersburg, (V. Tobias). ZMK Museum, Denmark Members of this assemblage have been placed in Zoological Copenhagen, (O. Lomholdt). the subgenera Pisonoides and Krombeiniellum, but Menke (1988) used species groups rather than The specimens from the collection of the Zoo- subgenera for infrageneric groups. The agile group Museum of the Moscow State is restricted to the eastern Palearctic and the Orien- logical University (ZMUM) were also used in the review. tal Regions, except for the east Asian species koreerxse which was introduced into North America after World War II (Krombein 1958a). presumably THE AGILE GROUP Morphological terminology used here follows Bohart and Menke and Menke The (1976) (1988). This group is characterized as follows: com- abbreviations are used in the text: following OOD pound eyes densely setose, antenna clavate, oc- = ocello-ocular distance; OD = ocellus diameter; cipital carina complete or nearly so, anterior POD = distance between ocelli. posterior pronotal pit small, subomaulus present, episternal The museum and institutions lent following sulcus straight, forewing with only two submar- for this are used specimens study (abbreviations ginal cells (true second submarginal cell lost in the text): through diminution), hindcoxa without dorsolat- eral carina, lamelli- AUZM Universiteit van Amsterdam, Zoologisch Museum, metapleural flange usually Amsterdam, The Netherlands (W. Hogenes). form, propodeum without lateral carina or crenu- BMNH The Natural History Museum, London, England late ridge, male sternum VIII narrow, gonostyle (C. R. L. Ficken). Vardy, simple, volsella small, penis valve compressed 120 Journal of Hymenoptera Research

laterally and without teeth or notched ventrally. (really 3rd) usually not petiolate; recurrent veins Menke (1988) regarded the setose eyes, the received by 1st and 2nd submarginal cells or 2nd clavate antenna, the two submarginal cells, the recurrent vein interstitial between 1st and 2nd presence of a subomaulus, the straight episternal submarginal cells; hind coxa dorsum with low sulcus, the lamelliform metapleural flange and inner carina, without outer carina; legs finely sculp- claw shape as apomorphies of the agile group. Its tured, without stout spines on tibiae and species apparently represent a monophyletic as- tarsomeres; all tarsomeres IV with small plantulae; semblage. Morphologically they are very similar tarsal claw thick to just before apex; propodeum and differ mainly in leg color, tergal bands, punc- rounded, without lateral carinae or lines of foveae tation, vestiture and proportions of tergum I. Male and crenulate ridges, punctate with smooth genitalia are also similar. One of the new species, interspaces; propodeal dorsum not delimited pulaivskii, stands apart from the others in the group (browni with shallow lateral sulcus), with medium because its metapleural flange is narrow. I regard furrow containing short to complete ridge; abdo- it as a plesiomorphic state. Two other species have men compact; tergum I simple or with preapical in unique autapomorphies: browni the propodeal transverse depression (Figs. 15-19); apical bands dorsum is delimited a laterally by shallow, broad of terga often translucent and with silvery or golden sulcus and tergum I has a distinct preapical de- pubescence; male sternum VIII long, narrow, in pression; agiloides submarginal cell II is open rounded or weakly notched apically; genitalia distally. compact, compressed laterally; volsellae small, —Inner Description. orbits of eyes moderately rounded and weakly setose; gonostyle triangular, emarginate, parallel (eyes equidistant at vertex simple, with long, coarse lateral setae curved be- and or clypeus) slightly converging below (rarely neath; penis valve compressed laterally, consider- above); eyes covered densely with short, erect ably widened apically, without teeth or notches setae (Fig. 3); clypeus convex, in female rounded ventrally. in male or Included — (Figs. 4a, 5a), angulate prominent apically Species. agile (Smith), agiloides sp. n., frons antennae (Figs. 4b, 5b); convex; clavate, com- browni (Ashmead), chrysoptilum sp. n., differens paratively short, with distal flagellomeres wider Turner, erythropus Kohl, hissaricum Gussakovskij, than labrum truncate or long; subquadrate, slightly koreense (Radoszkowski), ningxjuenfuense sp. n., carina a emarginate apically; occipital complete pulaivskii sp. n., rothneyi (Cameron), vechti sp. n. almost — (or complete) circle, narrowly separated Biology. Information is available for two spe- from hypostomal carina; male mandible simple, cies of the group, koreense and erythropus. The that of female with inner tooth slightly distad of former was studied by Iwata (1964) in Japan and midpoint; pronotum with small round pit anteri- by Sheldon (1968) in North America. The Indian without orly, lamellae; scutum and mesopleuron species, en/thropus, was studied by Home (1870). moderately, uniformly punctate; episternal sul- These wasps construct small, clay cells, placing cus almost straight, not curved forward ventrad; them separately or in groups on variable surfaces, subomaulus recurved ventrad; omaulus and ac- but not within linear cavities (borings in wood or etabular carina sulcus absent; mesopleural paral- empty stems of plants). The groups may include leled anteriorly by a row of foveolae; metapleuron up to 21 cells which do not merge in a common smooth; metapleural flange usually broadly lamel- mass but keep their independence. Prey consists liform of posteriorly (Figs. 6-8); tegula entirely punc- 6-31 paralyzed, small, usually immature spi- tate; forewing media diverging after cu-a; forew- ders. ing with two submarginal cells (Figs. 10-13), 2nd

KEY TO SPECIES OF THE AGILE GROUP

—1 All femora completely reddish 2 Mid- and hindfemora brown, forefemora mainly brown 6 narrow India —2(1) Metapleural flange (Fig. 9); 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 of middle almost equal to diameter hindtibia; Malaysia, Indonesia vechti sp. n.

Abdominal I same as —6(1) tergum polished, sculpture nearly 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 cell II Sri Lanka n. — I; submarginal open distally (Fig. 14); agiloides sp. 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 (OOD

female I often II-III tergum absent, narrow on terga (hardly broader than diameter of hindtarsomere I), bright golden; tergum I dull due to dense microsculpture; metapleural flange with narrower, densely pubescent lamella (Fig. 11); Russian Far East, Korea, eastern China, Japan, United States koreense (Radoszkowski) — Abdomen II-III 1 .5 comparatively longer, terga times as wide as long; translucent apical band of tergum

I those on II-III I present, terga whitish-yellow; tergum weakly shiny, with sparse microsculpture; with metapleural flange broader, spoon-shaped, sparsely pubescent lamella (Fig. 8); southeast Asia, Malaysia, Indonesia rothneyi Cameron

P. Bohart Pison agile (Smith) (Pison) agile: and Menke 1976:333 (listed). ^ Menke 1988:38 of Figs .3 4 6 10 15 agile. (member agile group, redescnbed).

Parapison agilis Smith 1869:300. Lectotype: female, "Ceylon" Lectotype Selection.—The natural History Mu- Sri , . . (now Lanka) (BMNH), present designation. , f_ ... ~ ... , , ... SeUm haS tW0 femaleS Wlth Smith S handwritten P. (Paraxon) agile: Kohl 1885:186 (listed). e labels. The first female has the museum white P (Ptsonoides) agilis: Turner 1916:616 (new combination, re- tyP described). round label (1) with "Type" printed inside a red 122 Journal of Hymenoptera Research

koreense

MAP 2 ^^ c?

of the Pison in the Old World. Fig. 1. Geographic distribution agile species group P. in the World. Fig. 2. Geographic distribution of koreense New Volume 3, 1 994

Figs. 3-19. Morphological features of Pison species. 3, head in frontal view of P. agile (a, female; b, male). 4-5, clypeus in frontal view P. P. (a, female; b, male): 4, agile; 5, pulawskii. 6-9, right metapleuron in lateral view (arrow shows metapleural flange): P. P. P. 6, agile; 7, koreense; 8, rothneyi; 9, P. pulawskii. 10-13, submarginal cell area of right forewing: 10, P. agile; 11, P. browni; P. P. 12, vechti; 13, pulawskii. 14, submarginal cell area of forewing of P. agiloides (a, right; b, left). 15-19, abdominal tergite I (a, lateral view; b, dorsal view): 15, P. agile; 16, P. koreense; 17, P. browni; 18, P. t'fdifi; 19, P. pulawskii. 124 Journal of Hymenoptera Research

—This is the darkest in the oval, a pale-blue round label (2) with "India" Discussion. species from other members the handwritten on the upper side and "56/150" hand- agile group, differing by written in two lines on the under side, a white narrow, brown apical bands on terga II-III, by the transverse of rectangular label (3) with "agilis Sm. Type" hand- strong, preapical depression tergum written in two lines, and the museum white rect- I and the shallow depressions that delimit the dorsum. The last character is in angular label (4) with "B.M. TYPE. HYM." printed propodeal unique in two lines and "21.538" handwritten below. the group. — Another female has a pale-blue round label (1) Range. Philippines. — with "Ceylon" handwritten on the upper side and Material Examined. 1 male, 2 females. "61 /36" on the under side, a pale-blue rectangular LUZON: Manila (USNM), Laguna (CAS). Tsuneki recorded the from label (2) with "P. agilis Smith. Type" handwritten (1983) species Bontoc, Luzon; in two lines. Only "Ceylon" was mentioned as the Mambucal, Negros; Cagayan de Oro, Mindanao. in the so the type locality original description Pison differens Turner "India" type may not be a true type. The Natural P. (Pisonoides) Turner 1916:617. female, History Museum also has more than a dozen agile differens Lectotype: India, Assam" (BMNH No. 21.540), females from "Ceylon" and three of them have the "Shillong, present designation. same round labels found on the second pale-blue P. (KrombeinieUum) differens: Bohart and Menke 1976:337 (new I believe them to be the members of the "type." combination, listed).

I the with P. 1988:38 of type series and have selected specimen differens: Menke (member agile group). Smith's type label as lectotype and the three other Lectohfpe Selection.—Turner described this spe- females as paralectotypes. I am discounting the cies from three females without selecting a holo- Indian as a specimen type. have two identical labels: — type. They rectangular Discussion. Pison agile is a member of a sub- (1) "Shillong. 5.03" handwritten in two lines and of species with mainly dark legs. Together group (2) "Assam. R.Turner. 1905-125" printed in three with n., it differs from other such agiloides sp. lines ("Shillong, Assam, 5000 ft. (Turner), May" in fine species having comparatively was mentioned in the original description). One microsculpture of abdominal tergum I in which female has in addition Turner's label "Pison are The the interspaces polished. remaining terga (Pisonoides) differens. Turn. Type." handwritten are similarly polished. Pison agile differs from in four lines and two museum labels, a round one and agiloides in having completely brown mid with "Type. H.T." printed in two lines inside a red hind legs, sparser and coarser scutal punctures ring and a white rectangular one with "B.M. and narrower whitish translucent apical tergal TYPE.HYM." printed in two lines and "21.540" bands. handwritten below. I have selected the last female —Known from southern India and as and the two others as Range. only lectotype — paralectotypes. Sri Lanka. Discussion. Pison differens is similar to rothneyi Material Examined.—2 males, 18 females. IN- in color of the body and of the translucent apical bands. The female differs from DIA: Anamalai Hills, Cinchona (BMNH, NMNH). tergal rothneyi by its small lateral ocelli and SRI LANKA: North Western Prov., Kurunegala comparatively elongate with its hindface Dist, Kurunegala; Central Prov., Kandy Dist., propodeum punctate (trans- carinate in The male also has Kandy, Udawattakele (BMNH, USNM, ZMUM). versely rothneyi). very small ocelli but the propodeum is not elon- In both sexes the dorsum is Pison browni (Ashmead) gate. propodeal pol- ished, finely sparsely punctate. The scutum is Figs. 11, 17 finely densely punctate in females. I think that Pisonoides browni Ashmead 1905:961. Holotype: male, Philip- differens may be a local form of rothneyi but more "Manila" no. examined. pines, (USNM 8332), material will be required to resolve this. P. (Pisonoides) browni: Turner 1916:617 (new combination, Range. —Known only from Assam State, In- listed). dia. P. (KrombeinieUum) browni: Bohart and Menke 1976:337 (new Material Examined.—1 3 females. IN- combination, listed). male, P. (KrombeinieUum) browni: Tsuneki 1983:81 (redescribed) DIA: Assam State, Shillong, IV. 1903, V.1903 P. browni: Menke 1988:38 (member of agile group). (BMNH). Volume 3, 1 994 125

Pison erythropus Kohl Pison hissaricum Gussakovskij

Smith 1869:299. "India" Parapison rufipes Lectotype: female, P. (Parapison) hissaricum Gussakovskij 1937:622. Holotype: (BMNH No. 21.539), present designation. Nee Pisomtus female, Tajikistan, "Stalinbad" (now Dushanbe) (ZIN), rufipes Shuckard 1838:79 (now in Pison). examined Parapison rufipes Home, in Home and Smith 1870:165 (biol- P. (Pison) hissaricum: Bohart and Menke 1976:336 (new com- ogy)- bination, listed). P. in and Smith 1870:188 P. hissaricum: rufipes Smith, Home (redesenbed). Menke 1988:38 (member of agile group). Pison (Parapison) erythropus Kohl 1885:183 (new name for Parapison rufipes Smith 1869, nee Ptsonitus rufipes — Discussion. Pison hissaricum also belongs to Shuckard 1838). the subgroup whose members have col- P. (Parapison) erythropus Kohl 1885:186 (listed). brightly ored This can be the P. erythropus: Bingham 1897:221 (listed). legs. species recognized by P. Pisonoides) erythropus: Turner 1916:616 (listed). following combination of features: legs except P. (Pison) Bohart and Menke 1976:335 (new com- erythropus: trochanters, tarsi and tibiae reddish apically; ter- bination, listed). gum I dull due to dense microsculpture of P. erythropus: Menke 1988:38 (member of agile group). interspaces, punctures of tergum I dense; and whitish translucent bands of I-III nar- Selection. —The Natural apical terga Lectotype History row. Museum has three females of this species bearing —Known from Uzbekistan and round labels with "India" handwritten on the Range. Tajikistan. upper side and "98.69" on the under side. One of Material Examined. —8 males, 10 females. them also has a rectangular label with "rufipes Sm. UZBEKISTAN: Aman-Kutan (ZMUM). TAJI- Type" handwritten in two lines. I believe them to KISTAN: Dushanbe; Varzob Valley, Kondara (ZIN, be the members of the series and I have type ZMUM). selected the last female as lectotype and the other two females as paralectotypes. — Pison koreense (Radoszkowski) Discussion. Pison erythropus belongs to the Figs. 7, 16 subgroup whose members have brightly colored legs. This species is easily recognized by its en- Paraceramius koreensis Radoszkowski 1887:433. Holotype: fe- tirely reddish-orange legs (including tarsi and male, "Koree" (Krakow, Poland. Lost). Neotype: fe- basal of the parts — trochanters). male, Russia, Primorskiy Kray (ZMUM), present desig- Biology. The nests observed by Home (1870) nation. in northwestern Ceramius koreensis: Morawitz in Dalla Torre 1894:3 (new India as Parapison rufipes (Sm.) combination, listed in Vespidae). included "a mass of loosely arranged cells of earth Pison (Parapison) koreense: Kohl in Dalla Torre 1897:712 (new attached to some such as a hanging objects, creeper, combination, listed in Crabroninae). or tendril, pendent straw, or even a curled dry Pison (Pisonoides) koreensis: Turner 1916:617 (new combina- leaf." Home mentioned that the cells were "very tion, listed). P. (Pisonoides) koreensis: Yasumatsu globular" and that their walls consisted of com- 1935:229 (listed). P. (Parapison) koreense: Gussakovskij 1937:622 (listed). paratively large "pellets. ..loosely attached to one P. (Pisonoides) koreensis: Yasumatsu 1939:83 (listed). another." He also "counted ("smallest eighteen P. (Paraceramius) koreense Krombein 1958a:166 (first record in in two spiders" Home) chambers." Nothing from United States). was reported about the number of cells per nest, P. (Paraceramius) koreense: Krombein 1958b:189 (listed). KrombemieUum koreense: Richards 1962:118 name for the species or even family of the prey nor was the (new cocoon described. Paraceramius Radoszkowski 1887, nee Saussure 1854). P. (Pisonoides) koreensis: Iwata 1964:1 —Known with from west- (biology). Range. certainty only P. (KrombemieUum) koreense: Krombein 1967:394 (new combi- ern India. nation, listed). — P. Material Examined. 2 males, 9 females. IN- (Krombemiellum) koreense: Menke 1968a:7 (listed). P. (KrombemieUum) koreense: Menke 1968b: 1102 DIA (BMNH, ZMUM); Maharashtra District, West- (redesenbed). P. koreense: Sheldon 1968:107 em Ghats, Lonavale (NMNH). (biology, larval morphology). P. (KrombemieUum) koreense: Hawkins 1974:279 (biology). P. (KrombemieUum) koreense: Bohart and Menke 1976:337 (listed). P. (KrombemieUum) agile: Krombein, et al 1979:1641 (synony- mized). 126 Journal of Hymenoptera Research

P. koreense: Kazenas 1980:92 (listed). plants, as in erythropns, is unknown. The cells were P. agile: Menke 1988:38 (listed). provisioned with six immature spiders of the ge- — nus Aranens or Neototype Selection. At my request, A. P. (Araneidae) (Iwata 1964) by 20-31 small of the Rasnitsyn recently searched for the type oikoreense spiders genus Dictyna (Dictynidae), in Radoszkowski's collection in Krakow, Poland including males and females of D. bellans Chamberlin and D. sublata Hentz but could not find it. I presume it was lost and I (Sheldon 1968). P. formise cocoons are have selected a specimen from the southern part of cylindrical (6.0-9.1 X 2.0-3.6 to Primorskiy Kray (32 km SE Ussurijsk, 25. VIII. 1986; mm) and, according Sheldon, were surrounded "delicate silken threads eastern Russia not far from Korea)as the neotype. by attaching cocoon to inside of Discussion. —Pison koreense is the best known cell." Sheldon also mentioned a large number of adult Melittobia member of the agile group because its biology has chalybii Ashmead been studied both in its native habitat in Japan (Eulophidae) which were reared from koreense (Iwata 1964) and in the United States (Sheldon cocoon "from a nest constructed in an unsealed cell of 1968) where it was introduced apparently after Trypoxylon politum." World War II (Krombein 1958a). I also observed nesting and hunting activities Krombein (1979) synonymized koreense with of koreense during 1983-1986 in the Primorskiy in the Far East of Russia. The agile and Menke (1988) accepted this synonymy. Kray females made their small cells in the After comparing the syntypes of agile with mate- clay between logs wall of an old rural rial from Sri Lanka, Japan, Russian Far East and shed, placing them in groups of up to 20 the United States, I have concluded that koreense is independent cells. Five to seven specimens of a It immature of the Araneus distinct species. differs from agile in having a spider genus were put into each cell. dull tergum I that contrasts with a shiny tergum II When hunting the female of koreense I looked for the on their webs (tergum shiny in agile). Pison koreense also differs prey and attacked them from the from agile in having bright golden translucent dorsum, embracing the carapace bands on II-III and a to the venter of the apical terga (whitish in agile), largely applying single sting After the the yellowish tibiae (brown in agile) and a transversely spider. paralyzing spider, female turned carinate propodeal hind surface (mostly punctate the prey venter up, clutched the base of its chelicerae with her mandibles it in agile). The metapleural flange of koreense is broad and flew with to and reddish it is in her cell. I observed once and it was (Fig. 7); narrow and dark agile egg-laying only Pison the last before the cell. (Fig. 6). koreense differs from differens and operation sealing Sheldon rothneyi in having a more compact abdomen and a (1968) also mentioned that the egg was attached to the of the weakly developed translucent apical band on ter- abdomen last spider in the cell. gum I. Thus, populations of koreense in Russian Far — the Biology. The wasps were studied in detail in East, Japan and United States have similar Japan by Iwata (1964) and in North America by biological features except prey. Spiders of the Sheldon (1968) who also provided a detailed de- family Dictynidae were preferred in North America and Araneidae in scription of the immature stages of koreense. Both — eastern Asia. Iwata and Sheldon observed that koreense con- Range. Russian Far East, Korea, Japan, east- structed small ern United States: (6.0-10.2 X 4.0-6.0 mm), fragile, China, Illinois, Kansas, Mary- mud cells with land, New Wisconsin. cylindrical finely cemented walls. Michigan, York,— Virginia, The cells are placed separately or in groups of up Material Examined. 28 males, 18 females. to 21 independent cells. The cells are attached in RUSSIA, Primorskiy Kray: Sichote-Alin, Vangou; 32 SE various protected places: inside a photographic km Ussurijsk; Partizansk; 70 km ENE tank (Krombein 1958a); in a culvert, in small de- Partizansk (CAS, USNM, ZMUM). CHINA, pressions and cracks and under a bridge (Sheldon Zhejiang: Hangchow (USNM). JAPAN, Fukui, Mt. 1968); in the nooks of a mud wall under eaves Haku: Ichinose; Chugu; Mie (USNM, ZMUM). (Iwata 1964); and even within the old empty cells USA: Kansas, Lawrence; Virginia, McLean of Trypoxylon (Trypargilmn) politum (Say) (Sheldon (USNM). 1968; Hawkins 1974). Construction of cells on 127 Volume 3. 1994

femora Pison rothneyi Cameron labrum, tibiae except apically, apically, lobe and Fig. 8 pronotal posteriorly, tegula metapleural of flange; translucent apicalbands abdominal terga P. (Parapison) rothneyi Cameron 1897a:81. Holotype: female, I-V and sterna II-V, and terga I-V laterally yellow- India, West Bengal State, "Barrackpore" (now Barakpur) ish-white. (OUM), examined. 1897a:25. male frons P. (Parapison) crassicorne Cameron Holotype: Clypeus, beneath, pronotum anteriorly, West (female in original description), India, Bengal State, scutellum, metanotum and propodeum with dense (now (OUM), examined. Syn- "Barrackpore" Barakpur) erect silvery pubescence, longest setae on onymy by Turner (1916:617). propodeum laterally and posteriorly; pronotal P. (Pisonoides) rothneyi: Turner 1916:617 (new combination, collar, scutum, and abdomen with listed). mesopleuron combi- setae P. (Krombeiniellum) erythropus: Tsuneki 1974:637 (new suberect, dense, short pubescence, longest nation, misidentified). on apical band of tergum I. 1976:336 P. (Pison) rothneyi: Bohart and Menke (listed). inner Labrum slightly emarginate apically, of P. rothneyi: Menke 1988:38 (member agile group). orbits almost parallel, OOD

. a carinate apically Also, rothneyi has more coarsely closed by sulcus or carina; metapleural flange sculptured scutum and deeper impressed broad, spoon-shaped, translucent, with long sil- differs from koreense parapsidal lines. Pison rothneyi very setae. abdominal with left forew- by the comparatively longer terga Right forewing (Fig. 14a) with two, translucent bands three outer well developed whitish-yellow ing (Fig. 14b) with submarginal cells, I. Pison also recur- on all terga, including tergum rothneyi submarginal cell of forewings open distally; has a broader, spoon-shaped reddish-orange rent vein II ending on tiny submarginal cell II (left metapleural flange (Fig. 8) and its tergum I is shiny forewing) and interstitial on right forewing. (semidull in differens and koreense). Tsuneki (1974) Abdomen simple; terga shiny, densely punc- misidentified specimens of rothneyi as erythropus. tate, diameter of punctures decreasing from basal tarsi of his con- I The dark femora and specimens to apical terga; tergum with microsculpture, firm this. other terga more polished between punctures; — distributed Range. This is the most widely sternum I similar to frons, other sterna densely, of the It is recorded of Oriental member agile group. finely punctate, shiny; translucent apical bands from eastern India, Thailand, Viet Nam, Malaysia abdominal segments 1 .5 to twice as broad as diam- and Indonesia. eter of first hind tarsomere. Material Examined.— 1 male, 13 females. IN- Male.— Unknown. DIA, West Bengal: Barrackpore (OUM). THAI- Discussion.—The new species differs from ag- LAND, Chieng Mai Province: Fang Horticultural ile in having the dense, fine punctation of the Exp. Station; Doi Inthanon N. P.: Huai Sai Luang scutum and abdomen, the longer propodeum, the (ZMK). Malaysia: Teluk Merban (ZMK); Kuala comparatively short setae of the thorax and abdo- de Lumpur (BMNH). VIET NAM: Tonkin, Poste men, the broad translucent apical bands of the Dong-Dang (ZMUM). INDONESIA, Sumatra: abdominal segments and the reddish tibiae and are Pakanbaru, Solok; Java: Bogor, Semarang metapleural flange. The last features shared reddish (NMNH). with rothneyi, but the latter has completely tibiae, tergum I is shiny and the propodeum is Pison new scutal agiloides Antropov, species elongate. Pison differens also has fine, dense 14 its Fig. punctures, but it differs from agiloides by larger OOD, by t its dark-brown tibiae and by the weakly Female.—Black Description of Holotype except shiny surface of tergum I. the following reddish: palpi, mandible largely, The open second submarginal cell of the unique 128 Journal of Hymenoptera Research

individual Recurrent vein I on cell I, specimen of agiloides may prove to be ending submarginal when more material is available. The presence of recurrent vein II ending on submarginal cell II near its base interstitial the true second submarginal cell in the left forew- (nearly ). that 7.6 mm. ing is likely to be atypical also, but it illustrates Length the loss of this cell is achieved via diminution of its Male.—Unknown. Discussion. —Pison differs from size in the agile group (see also Menke 1988). ningyuenfuense similar browni an undefined Etymology.—The ending of the species name the most by having dorsum sulcus or emphasizes the likeness of this species and agile. propodeal (no limiting carina), Range.—Sri Lanka. by the comparatively shallowly depressed apical of which is more Type.— Holotype female: SRI LANKA, part abdominal tergum I, coarsely Sabaragamuwa Province, Ratnapura District, punctate, with interspaces only weakly shiny, and Belihuloya Resthouse, 9.IV.1978, M. D. Hubbard, by the whitish translucent apical bands of terga I- T. Wijesinhe (USNM). V. Etymology. —The species name is derived from new the native name of the Pison ningyuenfuense Antropov, species — holotype locality. Range. Known only from the type locality in — China. Description of Holotype Female. Black except southeastern — female: the following reddish: palpi, mandible largely, Type. Holotype CHINA, Hunan alt. 600- labrum, foretibia anteriorly, midtibia basally, Province, Ningyuenfu, July 24-26-28, hindtibia basally and posteriorly and all spurs; 10,800, D. C. Graham (USNM). apical rim of clypeus, tarsi ventrally and tegula reddish-brown; metapleural flange brown; trans- Pison vechti Antropov, new species 18 lucent apical bands of abdominal terga I-V whit- Figs. 12, ish. Female. —Black Pubescence silvery; clypeus, frons, vertex, Description of Holotype except man- pronotum, scutum posteriorly, scutellum, met- the following bright reddish-orange: palpi, anotum and porpodeum with erect setae, longest dible largely, fore and mid tibiae and fore femur hind femora dorsum on pronotum laterally and on propodeum; gena, completely, mid and except mesopleuron and abdomen with suberect, dense, and hind tibia except posteriorly, tegula and short setae, those on gena and lateral angles of metapleural flange posteriorly, apical band on abdominal tergum I longest; scutum with ex- abdominal tergum I and medial spots on terga II- tremely short, erect setae as on femora. III before translucent apical bands; pronotal lobe Pronotum and scutum finely and densely posteriorly and spical tarsomeres beneath yellow- of I-V sterna punctate, shiny, punctures less than a diameter ish; translucent apicalbands terga and apart; mesopleuron, scutellum and metanotum II- VI bright golden. more sparsely punctate (punctures 1-2 diameters Pubescence of head and thorax silvery, mainly on translu- apart); propodeum coarsely punctate (punctures golden on abdominal terga (especially at least twice as large as those on scutum); abdomi- cent bands laterally); frons, vertex, gena, prono- nal tergum I moderately coarsely, sparsely punc- tum, scutum, scutellum, mesopleuron and met- tate (punctures similar in size to those on anotum with dense, short, erect setae (not longer propodeum), interspaces weakly shiny due to than mid ocellus diameter); propodeum dense microsculpture; sternum I similarly punc- posteriolaterally with longer dense, erect setae tate but shiny; following abdominal segments (almost twice as long as mid ocellus diameter); densely, finely punctate (like scutellum), shiny. abdomen with suberect or almost appressed dense, Translucent apical bands of abdominal terga short setae. II-V as broad as diameter of first hind tarsomere, Labrum truncate apically; median lobe of bands of tergum I and sterna II-V less than diam- clypeus obtusely angulate, with narrow, shiny eter of first hind tarsomere; apex of tergum I apical margin; inner orbits almost parallel; shallowly but distinctly depressed (depressed part OOD

of I v. d. ally as hindtibial diameter (those terga and V Pebaton, 28.VI.1932, J. Vecht; Buitenzorg, II-IV of first v. v. d. and sterna 1.5 times diameter Jnsl. Pin., 15.VI.1929, J. Vecht; Malang, hindtarsomere (Fig. 18). IV. 1933, Betrem; Bogor, 1955, Hamann; Mulie; W.

I cell Z. IV. v. d. Recurrent vein ending on submarginal I, Preanger, Soekaboemi, 1933, J. Vecht; recurrent vein II ending near base of submarginal Ambarawa, Lundeking (NMNH); Malang, IV.1933, cell II, the latter very narrow anteriorly, almost Betrem (ZMUM); Mt. Tijoeng, Djampang Tengah, triangular (Fig. 12). 1.1939, K. M. Walsh (BMNH); Kangean Isl.: Length 7.6 mm. Tembajangan, 11.1936, M. E. Walsh (NMNH). Variation in Females (18 specimens).—Femora and tibiae all bright reddish-orange in two females Pison chrysoptilum Antropov, new species from Buitenzorg, Java (28. VI. 1932) and — Tembajangan, Kangean Island all femora dark Description of Holotype Female. Black except posteriorly in a female from Mulie, Java, and sub- the following reddish-orange: palpi, mandible marginal cell II completely triangular in another except apically, labrum, pronotal lobe posteriorly, female from Buitenzorg (16.X.1941). Length 7.0- tegula mainly, widened part of metapleural flange, 7.8 mm. all femora tibiae — except posteriorly and except Male. As in female except: median lobe of apically; translucent apical bands of abdominal clypeus narrower, acutely prominent apically; terga I-V golden. punctation sparser on scutum (punctures 1-2 di- Pubescence silvery on clypeus, mesopleuron, ameters apart) and mesopleuron (punctures 2-4 propodeum laterally and abdominal sterna, golden diameters apart). Length 5.4 mm. on frons, vertex, gena, pronotum, scutum, scutel- Discussion. —Pison vechti is very similar to lum, metanotum, propodeal corsum and hindface hissaricum, erythropus and differens in having wid- and abdominal terga. Clypeus, pronotal collar, ened metapleural flange and largely bright red- propodeal dorsum and terga with suberect or dish-orange legs. It is easily separated from decumbent dense setae (shortest on abdomen, hissaricum and erythropus in having the broad, longest on pronotum); other areas with erect setae, golded, translucent apical bands of the terga. Fur- shortest on scutum and abdominal sterna, longest thermore, it differs from erythropus in having dark- on metanotum and propodeum posterolaterally. brown trochanters and tarsi and a densely punc- Median clypeal lobe rounded, with narrow tate, weakly shiny scutum. Pison vechti differs semitranslucent brown apical margin; inner orbits from differens in having the larger lateral ocelli and of eyes converging above; OD>POD>OOD. a comparatively short propodeum. The color of Clypeus, frons, vertex, pronotum, scutum, the abdomen of vechti is similar to the rothneyi, but scutellum and mesopleuron finely, densely punc- 130 Journal of Hymenoptera Research tate (punctures 0.5-1.5 diameters apart), shiny; tarsomeres III-V; antennal articles beneath, apical metapleural flange broad, spoon-shaped, widened margin of clypeus, tegula, mid and hind tarsomeres part translucent; propodeum more coarsely, III-V reddish-brown. sparsely punctate laterally and dorsally (at least 2 Pubescence silvery only; clypeus, frons, ver- diameters apart), hindface coarsely and densely tex, pronotal collar, scutum, scutellum, metano- punctate (0.5-1 .0 diameters apart), with transverse tum, metapleuron and propodeum with erect se- ridges ventrallv; propodeal dorsum not enclosed tae, longest on propodeum posterolaterally and by sulcus or carina, with narrow median furrow shortest on scutum (scarsely longer than on containing simple carina; tergum I punctate (l-# femora); gena, mesopleuron and abdomen with diameters apart), shiny in spite of microstriahon; suberect or appressed short setae (longest on api- sternum I with coarser punctures (as on propodeal cal bands of terga and shortest on sterna). hindface), shiny; other abdominal sclerites micro- Labrum truncate apically; clypeus compara- scopically but distinctly, densely and uniformly tively broadly rounded (Fig. 5A); inner orbits of punctate (approximately a diameter apart), shiny; eyes almost parallel; OOD=OD=POD. translucent apical margins of terga I-IV twice as Frons densely punctate, shiny; scutum very broad as diameter of first hind tarsomere (on densely, finely punctate, weakly shiny; tergum V and sterna III-V slightly less than diam- mesopleuron uniformly, sparsely punctate (1.0- eter). 1.5 diameters apart), polished; metapleural flange Recurrent vein I ending on submarginal cell I, not lamellate (Fig. 9); propodeal dorsum obliquely recurrent vein II ending near base of submarginal carinate basally and along median furrow, dor- cell II. sum not enclosed by sulcus or carina, sparselv Length 8.2 mm. — punctate (as on propodeal side) (1-3 diameters Variation in Females (2 specimens). Paratype apart); tergum I uniformly, densely ounctate (as female differs by its smaller size (6.2 mm) and on mesopleuron), shiny; tergum II more densely inner orbits of and almost parallel— eye. finely punctate, shiny; remaining terga weakly Male. Unknown.— shiny because of very dense microscopic punc- Discussion. Unlike all other members of the tures; sternum I dull, punctured as frons; sternum agile group, chrysopdlwn has golden pubescence II shiny, punctate as mesopleuron; other sterna on the frons, vertex, thorax and propodeum. It with dense micropunctures, weakly shiny; trans- resembles vechti because of its brightly colored lucent apical bands of terga I-V and sterna II-IV legs and golden translucent bands on the abdomi- whitish, as broad as diameter of hindtarsomere I nal terga. Pison chrysoptilum and vechti may be two (Fig. 19). forms of one species, but more material will be Recurrent vein I ending on submarginal cell I, necessary to resolve— this. recurrent vein II ending on submarginal cell II Etymology. This species name is derived from near its base (Fig. 13). the Greek words and 7.0 mm. din/sos (=gold) ptilon (=down, Length — fluff) emphasizing— the color of the pubescence. Variation in Females (11 specimens). Mid tarsi Range. Northern Borneo (Sarawak, Brunei). completely reddish-orange in three specimens Types. —Holotype female: MALAYSIA, from holotype locality and one from Sarawak: 4th div., Gn. Mulu, RGS Exp., 17.IX- Kurumbagaram; hind tibiae and tarsi brownish 23.X.1977, D. Hollis (BMNH). Paratype (1 female): posteriorly and antennal flagellum completely BRUNEI: Ulu Temburong, Base camp hut, 300 m, dark-brown in one specimen from Coimbatore. 115 16'E,4 26'N, 16.II-9.III.1982, M. C. Day (BMNH). Length 6.0-7.0 mm. Males (3 specimens).—As in females except: Pison pulawskii Antropov, new species clypeus with acute median prominence spicallv 19 Figs. 5, 9, 13, (Fig. 5b); labrum slightly emarginate; mandibles — without inner teeth. Length 4.8-5.2 mm. Description of Holotype Female. Black except Discussion. —The nonlamellate metapleural the following reddish-orange: palpi, mandible flange of pulawskii sets the species apart from other fore distad of coxal largely, labrum, legs apex, mid in the agile group. Other features of the species are and hind legs except basal parts of coxae and the form of the clypeus, the mostly reddish-orange Volume 3. 1994 131

legs and the narrow whitish translucent apical Hawkins, W. A. 1974. A western record for an introduced bands of the Pison the Kansas abdominal segments. Reddish legs (Hymenoptera, Sphecidae) journal of Entomological 47(2): 279. occur in erythropus and whitish abdominal bands Society Home, C and F. Smith. 1870. Notes on the occur habits of some in erythropus, agile and ningyuenfitense, but hymenopterous insect from the north-west provinces of these have a lamellate India. With an species metapleural flange. appendix, containing descriptions of — some new Etymology. This species is dedicated to species of Apidae and Vespidae collected by Mr. Home. Transactions the Wojciech J. Pulawski. of Zoological Society of London 7(3): 161-196. Range.—Western and southern India. Iwata. K 1964. notes on — Ethological four Japanese species of Types. Holotype female: INDIA, Rajasthan: Pison (Hymenoptera, Sphecidae). Mushi 38: 1-6. (24 35'N), 27.V.1989, W. J. Pulawski V. L. Uddaipur Kazenas, 1980. Materials to the fauna of digger wasps (CAS). Paratypes (3 males, 9 females): same place, (Hymenoptera, Sphecidae) of the Far East of the USSR. date and insects the Far East, 80-94. collector as holotype (CAS); same place Taxonomy of of Vladivostok, pp. Kohl, F. F. 1885. Die Gattungen und Arten der Larnden and collector as holotype, 21-25. V. 1989 (CAS, Auctorum. derk. k. Disa Verliandlungen zoologisch-botamschen ZMUM); Gujarat: (=Deesa), 4-6. VI. 1989, W. J. Gesellschaft in Wien 34: 171-268. Pulawski (CAS); Karnataka: 915 K. V. Bangalore, m, Krombein, 1958a. Pison tParaceramius) koreense (Rad.), a 26.V.1980, 30.V.1980, K. D. Ghorpade (ZMK); new adventive wasp in the eastern United States. Enfo- News 69: 166-167. Karikal Territory: Kurumbagaram, III.1947, P. S. mologtcal Nathan (USNM). Krombein, K. V. 1958b. Superfamily Sphecoidea. In: K. et Krombein, V., al., Hymenoptera of America North of Mexico - Synoptic Catalog. Agriculture Monograph 2, ACKNOWLEDGMENTS First Supplement, U. S. Department of Agriculture, Washington, DC. pp. 186-204.

I Krombein, K. V. 1967. In: express my gratitude to all colleagues who lent mate- Superfamily Sphecoidea. Krombein, K. B. D. rial for I V., Burks et al., America North study. am grateful to Dr. A. P. Rasnitsyn for his help Hymenoptera of of Mexico— in searches for the type specimen of koreense in Radoszkowski's Synoptic Catalog. Agriculture Monograph 2, Second U S. of collection. I am extremely grateful to Dr. A. S. Menke who Supplement, Department Agriculture, read earlier DC. 386-421. versions of the manuscript and suggested many Washington, pp. Krombein, K. V., P. D. Hurd, D. R. Smith and B. D. Burks. important improvements of the text and who helped me in 1979. the I thank Dr. K. V. Catalog ofHymenoptera in American North Mexico. English. Finally, Krombein and Dr. W. J. of Pulawski Vol. 2. Smithsonian Institution Press, DC. who critically and attentively reviewed the entire Washington, pp. 1199-2209. manuscript and who provided manv useful comments. A. Menke, S. 1968a. New genera and species of wasps of the tribe from the LITERATURE CITED Trypoxylonini neotropical region (Hy- menoptera: Sphecidae: Larnnae). Los Angeles County Museum Contributions in Science 135: 1-9. W. H. 1905. to Menke, A. S. 1968b. review of Ashmead, Additions the recorded Hymenopter- A the New World species of ous fauna of the Pison. I. The Philippine Islands, with descriptions of subgenus Krombeiniellum (Hymenoptera: new species. Proceedings of the United States National Sphecidae). The Canadian Entomologist 100(10): 1100- Museum 28(1413): 957-971. 1107. C. T. Bingham, 1987. The fauna of British India. Hymenoptera, Menke, A. S, 1988. Pison in the New World: a revision vol. 1. Taylor and Francis, London. 179 pp. (Hymenoptera: Sphecidae. Trypoxylonini). Contribu- Bohart, R. M. and A. S. tions the American Menke. 1976. Sphecid wasps of the of Entomological Institute 24(3): 1-171. World, a generic revision. University of California Press, Radoszkowski, O. 1887. Hymenopteres de Koree. Horae Los + Angeles, London, iv 695 pp. Scoitatis Entomologicae Rossicae 21(3-4): 428-436. P. 1897a. IV. O. W. 1962. Cameron, Hymenoptera Orientalia, or contribu- Richards, A revisional study of the masarid wasps. tions to a of the knowledge Hymenoptera of the Oriental Wm. Clowes and Sons, London. 294 pp. Part V. K. 1968. The zoological region. Memoirs and Proceedings of the Sheldon, J. nesting behavior and larval morphol- Manchester of Pison Literary & Philosophical Society 41(2): 1-144. ogy koreense (Radoszkowski). Psyche 75: 107- Dalla C. G. Torre, 1894. Catalogus Hymenopterorum hucusque 117. descriptorum systematica et synonymicus. Vol. IX. G. En- Shuckard, W. E. 1838. Descriptions of new exotic aculeate 181 Transactions gelmann, Lipsiae. pp. Hymenoptera. of the Entomological Society of Dalla C. G. 1897. London Torre, Catalogus Hymenopterorum hucusque 2(1): 68-82. F. 1869. descriptorum systematica et synonymicus. Vol. VIII. G. Smith, Descriptions of new species of the genus Pison, 750 and a list Engelmann, Lipsiae. pp. synonymic of those previously described. V. V. 1937. Transactions Gussakovskij, Especes palearctictiqques des genres of the Entomological Society of London. 9: 289- Didmeis Wesm., Pison Latr. et Psen Latr. (Hymenoptera, 300. Travaux de ' K. 1974. Sphecodeal). I'Institut Zoohgie de I Academie Tsuneki, A contribution to the Knowledge of des Sciences de I'URSS 4(3-4): 599-695. Sphecidae occuring in Southeast Asia (Hym.). Polskie 132 Journal of Hymenoptera Research

Pismo Entomologiczne 44(3): 585-660. Yasumatsu, K. 1935. The genus Pison Spinola of the Japanese Tsuneki, K. 1983. Further studies on the Larrinae of the Empire (Hymenoptera, Trypoxylonidae). Annotationes Philippine Islands, with remarks on the Indian species zoologicae Japonenses 15(2): 227-238. of the genus Lyroda (Hymenoptera, Sphecidae). Special Yasumatsu, K. 1939. Notes supplementaires sur le genre Publications of the japan Hymenopterists Association 24: 1- Pison Spinola du Japon (Hymenoptera, Trypoxylonidae). 116. Festschrift zum 60 Geburtstage von Professor Dr. Embnk Turner, R. E. 1916. Notes on the wasps of the genus Pison and Strand, Riga 5: 81-84. some allied genera. Proceedings of the Zoological Society of London 1916: 591-629. 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- The study was conducted in cerrado vegeta- social wasps on many insects (Gobbi et al. 1984, tion during a five week survey (August, October Gobbi and Machado 1985, Machado et al. 1987, through December 1988 and June 1989) at the Rio Raw 1988) confers on them a fundamental impor- Manso Hydroelectric Power Station (HPS-Rio tance as biological control agents. Nonetheless, Manso) (14°52' S and 55°50' W), in the Chapada the study of the nests of these insects has focused dos Guimaraes county, southern Mato Grosso mainly on their architecture (Jeanne 1975, Kojima State in central Brazil. 1982) and on the numbers of individuals per colony Cerrado sensu-lato is a semideciduous (Richards 1978), with few reports on pairs of spe- xeromorphic vegetation dominant in Central Bra- cies nesting close together (Windsor 1972, Starr zil, occupying about 20% of the whole country. It 1988). occurs in various structures from closed forest- Information on preferences for nest habitats like forms to pure grasslands (Eiten 1972). "Habi- (Richards 1978, Reed and Vinson 1979, Forsyth tat" is used here to refer to a vegetation subtype 1980) and on colony densities (Rau 1942, Kitayama within the cerrado sensu-lato. et al. 1989) of neotropical social wasps is still very The censuses of social wasp colonies (nests scarce. In central Brazil, 130 species of wasps have with resident adults) were conducted in campo been collected (Richards 1978, Raw pers. comm. umido, vereda, campo sujo, cerrado sensu-stricto, and our own collections), but studies on their gallery forest and surrounding dirt roads. Charac- ecology are very fragmentary. terizing these habitats very briefly it could be said The aim of this report is to add new informa- that campo umido is a wet grassland without tion on colony densities and preference for nest visible woody plants (Fig. 1); vereda is a broad habitats of social wasps in the cerrado sensu-lato of marshy valley bottom grassland with buriti palm Central Brazil. (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 <

forest Dr. Roberto Fig. 5. Area of gallery (Photo by Cavalcante).

includes forms with the total cerrado sensu-stricto 1972), gallery forest and dry forest (deciduous and with a woody plant cover of about 30-40% and semi deciduous mesophytic forest, Eiten 1972). meters canopy generally less than seven (Fig. 4) For the study on colony density, known areas and forest is a narrow in forest 2 and in gallery usually evergreen were sampled, gallery (7500 m ) forest the streams 2 lo- mesophytic following (Fig. 5) cerrado sensu-stricto (10000 m ). Nests were (Eiten 1972). cated among leaves, on branches, on the trunks of Searches for nests were conducted in four trees and in holes in tree-trunks and in the ground. different months, covering two seasons: dry sea- Binoculars were used to locate nests in the canopy. 2 son (June and August) and wet season (October In addition, five squares of 100 m each were habitats. Each habitat re- and December) in six surveyed in campo timido, where nests of wasps each ceived the same intensity of searching during were sought among tufts of grasses and sedges of these seasons. (Fig.l). were collected to From each nest, adult wasps The survey for estimates of the density of guarantee identification of the species. All the colonies, within the study area, was done during nests and insects sampled are deposited in the the wet season. Three people spent approximately de Laboratorio de Zoologia, Departamento 360 hours searching for nests. In each habitat the de Brasilia. Zoologia in the Universidade time spent per square meter was about one per- The data on Xavantina and Serra do Cachimbo son-minute. (10°50'S and 51°47'W), in northern Mato Grosso For the estimates of density of adult wasps/ state were compiled from Professor Richards' ha, the number of adults within some of the nests of nest habitats (1978) species descriptions using were counted and also, for some species, data doubt on the only the data where there was no from Richards (1978) were compiled. location of the nests). Richards' censuses were In each table the habitats, when it was pos- conducted in dirt roads and campo sujo, clearings, sible, were arranged in order of increasing com- cerrado Cerradao medium tall sensu-stricto, (the plexity of structure because this makes it easier to arboreal form of cerrado with a closed or semi- see general patterns in species preferences for open canopy with 30-40% tree crown cover, Eiten habitats. Volume 3. 1 994 137

RESULTS AND DISCUSSION seven in four and three in five habitats (Table 3). The area surveyed was different in each habitat so 1. Nesting Habitat Preference of Wasps the number of colonies found do not represent one absolute abundance but the relative habitat pref- A total of 100 colonies, comprising 30 species erence for nesting. in 15 genera of wasps were encountered in six Of the total number of colonies recorded by different habitats in Chapada dos Guimaraes (Table Richards and us from Mato Grosso, the species 18 in with the of colonies 1). Among them, species nested only one largest number was Polybia nine in and three followed occidentalis habitat, two,— (Polybia (Myrapetra) ruficeps (14%), by Polybia (9%), ruficeps Schrottky a very common species of the Polybia jurinei (6%), Polybia (Myrapetra) erythrothorax region, Chartergus chartarius (Olivier) and Richards (5%), Polybia ignobilis (5%), Paraclmrtergus Clwrtegelhis communis Richards) in three habitats fratemus (Gribodo) (4%) and Epipona tatua (4%). (Table 1). Of the species that nested in only one The genus Polybia represents 60% of the colonies habitat, most were found in cerrado sensu-stricto reported from the two regions of Mato Grosso (seven species), in gallery forest (five species) and State (Table 3), ranging from 67% in Xavantina and in campo umido (four species) (Table 1). Of the 30 Serra do Cachimbo (Table 2) to 45% in Rio Manso species collected in this study, only Polistes (Table 1). Among all the colonies of the 19 species (Epicnemius) pacificus liliaciosus de Saussure was of Polybia collected, 58% were found in cerrado not collected by Richards (Table 1 and 2). This sensu-stricto (Table 3) ranging from 31% in Rio subspecies has been collected only rarely in cen- Manso and 55% in Xavantina and Serra do tral Brazil and, until Richards' 1978 record, its Cachimbo (Tables 1 and 2). distribution was thought to be restricted in Brazil Richards (1978) suggested that the choice of to the North (Amazonas, Amapa and Para states). habitat for nesting is very characteristic and less Polybia ruficeps was the commonest species of wasp diverse than that used for foraging. His opinion is at Rio-Manso with 23 colonies, of which 70% were confirmed in our survey at Rio Manso. For ex- found in cerrado sensu-stricto. Three other species, ample, both Synoeca suririama (L), which nests in Epipona tatua (Cuvier), Synoeca surinama (L) and gallery forest, and Polybia (Trichothorax) sericea Oiartergus chartarius, with eight colonies each, (Olivier) in cerrado sensu-stricto, were collected were also common (Table 1). when they hunted in gallery forest, cerrado sensu- Richards (1978) recorded 199 colonies com- stricto, campo sujo and campo umido. Of the 30 prising 51 species in 14 genera nesting in six habi- species of wasps at Rio Manso, 21 (70%) were tats (Table 2). Most of the species he collected (25 encountered foraging in more than one habitat (in of the 51), were found nesting only in one habitat preparation). (eight of them in dry forest), 12 species nested in A water source is important for wasps to nest two habitats, eight nested in three habitats, three successfully, (Rau 1942, Forsyth 1980). The latter nested in four, and one species Polybia (Myrapetra) author concluded that the greater colony density platycephala Richards nested in five habitats. In of wasps in gallery forest was due mainly to the Xavantina and Serra do Cachimbo, Polybia presence of water. Wet habitats at Rio Manso (Myrapetra) occidentalis (Olivier) with 24 colonies (gallery forest, eight/ha and campo umido 60/ha) (12%), was the most abundant, followed by Polybia also had higher colony densities than did dry ruficeps (10%), Polybia (Apopolybia) jurinei de habitats (cerrado sensu-stricto five/ha). Polistes Saussure (8%), Polybia (Trichothorax) ignobilis (Epicnemius) subsericeus de Saussure and (Haliday) (7%) and Polybia (Formicicola) rejecta (F) (Mischocyttarus) drewseni de (5%) (Table 2). Saussure which nest in campo umido, had higher The data collected by us and by Richards were colony density compared to other species in drier combined and for the same type of habitat the data habitats, such as cerrado sensu-stricto. Availability were lumped (Table 3 representing the sum of of water all year round and nest protection by tufts Tables 1 and 2). At the nine habitats 299 colonies, of grasses and sedges, could be the cause of that comprising 61 species and 16 genera, were col- high density. lected. Twenty eight species (46%) nested in only Most have preferences for particular one 13 habitat, species nested in two, 10 in three, 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 Volume 3, 1994 139

Table 2. Wasp nesting habitats of Xavantina and Serra do Cachimbo: Mato Grosso, Brazil (Data from Richards, 1978)

Species 140 Journal of Hymenoptera Research

Species CS DC CE CD GF DF TT

Stelopolybia angulata Volume 3. 1 994 141

Species CU VE CS OF DC CE CD GF DF TT

Polistes cinerascens 142 Journal of Hymenoptera Research

- Table 4. Wasp colonies density in three different habitats in Rio Manso Chapada dos Guimaraes: Mato Grosso, Brazil.

Campo Cerrado Gallery umido sensu stricto forest

2 Area (m ) Volume 3. 1 994 143

ACKNOWLEDGMENTS

We acknowledge the contribution made by the late Professor the Universidade de Brasilia and the anonymous reviewer O.W. Richards to studies of social wasps in general. It is not for useful comments on the manuscript. Dr. Raw also iden- unusual for subsequent workers to discover interesting in- tified several of the species of wasps. Dr. Paul M. Marsh formation in the data he published. Financial support was made some editorial changes on the manuscript. Sra. Beatnz provided by the North Brazilian Electricity Company Mossri kindly assisted in the collection of field data. We also (Eletronorte) and the University of Brasilia. We are espe- thank Dr. Antonio Jose E. Brussi, for his help and encourage- cially grateful to Dr. Anthony Raw and Dr. Linda Caldas of ment

LITERATURE CITED

Eiten, G. 1972. The Cerrado vegetation of Brazil. The Botanical Machado, V. L. L. Gobbi, N. & Simoes, D. 1987. Material Review 38:201-341. capturado naalimentai;aodeSrp/opo/i/l)i(!fi«//if)fs (Olivier Forsyth, A. 1980. Nest site and Habitat selection by the social 1791) (Hymenoptera: Vespidae) Anais da Soctedade wasp, Metapolybia azteca Araujo (Hymenoptera: Entomologica do Brasil 16 (1): 73-79. Vespidae) Brenesia 17: 157-162. Partridge, L. 1978. Habitat selection In Behavioural Ecology Gobbi, N. & Machado, V. L. L. 1985. Material capturado e (Eds. JR. Krebs & N.B. Davies) Blackwell Scientific utihzado na alimentacao de Polybia (Myrapetra) paulista Publications, Oxford, 351-376. Ihering 1896 (Hymenoptera-VespidaeJ./lnais da Sociedade Rau, P. 1942. Habitat preferences of Polistes wasps (Hy- Entomologica do Brasil 14(2): 189-195. menoptera: Vespidae) Entomological Nezus 53: 293-295. A. Gobbi, N. Machado, V. L. L. & Tavares Filho, J. A. 1984. Raw, 1988. Social wasps (Hymenoptera: Vespidae) and Sazonalidade das presas utilizadas na alimentacao de insect pests of crops of the Surui and Cinta Larga Polybia occidentalis (Olivier 1791) (Hymenoptera: Indians in Rondonia, Brazil. The Entomologist 107: 104- Vespidae). Anais da Sociedade Entomologica do Brasil 13(1): 109. 63-69. Reed, H. C. & Vinson, S. B. 1979. Nesting ecology of paper Henriques, R. P. B. Rocha, I. R. D. & Kitayama, K. 1992. Nest wasps (Polistes) in a Texas urban area (Hymenoptera: density of some social wasps species in Cerrado vegeta- Vespidae). Journal of the Kansas Entomological Society 52 tion of Central Brazil (Hymenoptera: Vespidae) (4): 673-689. Entomologia Generalis 17(4): 265-268. Richards, O. W. 1978. The social wasps of the Americas. Jeanne, R. L. 1975. The adaptiveness of social wasp architec- British Museum of Natural History. London 580 pp. ture. Quarterly Review of Biology 50: 267-287. Starr, C. K. 1988. The nesting association of the social wasps Kitayama, K. Rocha, I. R. D. & Bulhoes B. M. 1989. Densidade Mischocyttarus immarginatus and Polybia spp. in Costa de ninhos de vespideos em diversos habitats de Mato Rica. Biotropica 20(2): 171-173.

Grosso. I . Shnpdsio Latino-americano sobre Inseto Socials Windsor, D. M. 1972. Nesting association between Neotropicais. Rio Claro SP. R-67 P 44. twoNeotropical polybiine wasps (Hymenoptera: Kojima, J. 1. 1982. Nest architecture of three Ropalidia species Vespidade)Biofropic«. 4: 1-3. (Hymenoptera: Vespidade) on Leyte Island, The Philip- pines. Biotropica 14(4): 272-280.

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 catalog 19 species for the world and Azevedo (1991) described B. dentata from Sao Paulo State, Bakeriella Kieffer is a neotropical genus, rang- Brazil. ing from Southern United States to Northern Ar- The present study focuses on four species of gentina (Azevedo 1991). Kieffer (1910a) described this genus, B. montivaga (Kieffer), B. reclusa (Evans) this genus based on one species from Para, Brazil, and two new species. Ideally what is needed is a B. flavicornis, which has the pronotum carinate complete revsion of the genus. Unfortunately, laterally and medially as well as anteriorly. This specimens are rare in collections and until more genus is related to Epyris Westwood by having the material is available for study, I have chosen to scutellar groove divided (Evans 1964). Kieffer present these notes and new species descriptions. (1910b) described B. depressa from Peru which has Specimens for this study belong to the follow- a carinate pronotum also. Evans (1964) expanded ing collections (abbreviations according to Arnett the definition of Bakeriella when he described ten & Samuelson (1986)): IBSP, Instituto Biologico de species, in some of which the median or lateral Sao Paulo, SP, Brasil; MZSP, Museu de Zoologia carinae of the pronotum are weak or absent. Evans da Universidade de Sao Paulo, SP, Brasil; UFES, (1969), in his revision of Epyris, pointed out the Colecao de Entomologia da Universidade Federal relationship of the monthwgiis-group of Epyris to do Espirito Santo, Vitoria, ES, Brasil. The last col- Bakeriella by the fact that both groups have similar lection is not included in Arnett & Samuelson scutellar pits. Evans (1979) reviewed Bakeriella, (1986) because it is new. providing identification keys for both males and Abbreviations used follow Evans (1964) and females and arranging the genus into four species- are as follows: LFW, length of forewing; WH, groups. In this revision, Evans described one spe- maximun width of head, including eyes; LH, length cies, B. erythrogaster, which has the scutellar groove of head, measured from median apical margin of similar to the genus Rliabdepyris Kieffer. Evans clypeus to median point of vertex; WF, minimun also transfered members of the montivagus-group width of frons; HE, maximun height (or length) of from to Bakeriella Epyris simply by the fact that eye; WOT, maximun width of ocellar triangle, they closely resemble each other. Curiously, he including posterior ocelli; OOL, shortest ocello- did not include B. subcarinata Evans 1965 in his ocullar line. The nomenclature of the integument revision. Gordh & Moczar (1990) cited in their follows Eady (1968). 146 Journal of Hymenoptera Research

Bakeriella incompleta, sp. n. (fig. 1). Scutellar pits tranverse, much wider than Figs. 1-2 long, separated by a thin septum which is lower than of foveae. disc as as Holotypefemale.—Brazil: SP,Ibitinga;05.X.1988; margin Propodeal wide with three discal median one com- yellow pan trap, rubber-tree crop, no collector long, carinae, the others given. (UFES). plete and incomplete, occupying about — two thirds of disc and but Description ofholotypefemale. Length of body converging posteriorly 6.0 mm; LFW 2.85 mm; head and thorax black, not reaching median carina; space among discal carinae with tranverse disc abdomen black except tip dark brownish; palpi striae; without sublateral inner of lateral light brownish; mandible brownish, black at base carinae; side carinae a series of small and with a light brownish spot outside, teeth paralelled by irregular depres- reddish; antennal scape dark brownish, apex sions; anterior part of space between lateral and brownish, pedicel and flagellum somewhat discal carinae occupied by a rather large depres- infuscated; tegula brownish; fore and hindcoxae sion; posterior angle of the disc foveolate; decliv- black, midcoxa darkbrownish; trochanters, femora, ity of propodeum with median carina and perpen- tibiae and tarsi dark brownish, the anterior being dicular striae. Metapleuron with longitudinally sinuous striae. lighter; wings subhyaline somewhat brownish, Mesopleuron granulate, with large fovea veins, prostigma and stigma dark brownish. anteriorly and small, deep pit at anterior with small elon- Head (Fig. 1). Mandible with 5 teeth, apical margin, pentagonal fovea and fovea with tooth largest (Fig. 1). Clypeus with subangulate gate upper (fig. 2). Mesosternum small median anterior median lobe, with median carina straight in lat- pit, margin of acetabular carina striae. eral profile, clypeus forming a small excavated paralelled by Midtibia spinose. Forefemur reentrance in the region of contact with frons. WH 2.0 X as long as thick, midfemur 1.65 X as long as 1.03 X LH; WF 0.5 XWH;WF 1.34 X HE; WOT 1.33 thick. Claw trifid, first tooth rounded off, the oth- X OOL. Antennal scrobe not carinate. Posterior ers sharp. ocelli about 3 X their diameter from the vertex. Metasoma. Sterna setose on posterior of each sternum. Vertex nearly straight and with lateral angles Terga with few setae on the posterior rounded off. Distance from the top of eyes to crest half of each tergum. Terga I and II with lateral of vertex about 0.7 X HE. Length of first four setae. Tranverse section of metasoma circular. antennal segments in a ratio of about 11:5:4:4. Length of mesosoma 1.35 X length of metasoma. Valvae III dark Segments 5 to 12 about as wide as long. Temple not brownish with only its end setose, carinate. Frons somewhat alutaceous and bril- setae range between one third and a half length of liant. the valvae. — Mesosoma(Fig. 1). Somewhat more alutaceous Paratype female. BRAZIL: SP, Balsamo; 21. IX. white than frons, mesoscutum polished anteriorly. 1988, pan trap, no collector given. Pronotal disc slightly convex anteriorly, tranverse (IBSP). anterior carina without teeth, median carina weak, The paratype differs from holotype as fol- lows: situated in a slight depression and absent anteri- WH 1.00 XLH;WF0.57XWH; WF 0.76 X HE; orly, not intersecting transverse carina. Posterior OOL 1.62 X WOT; propodeal disc 1.08 as large as margin of pronotal disc paralelled by a series of long. — small foveae; lateral margin of disc somewhat Remarks. This species is known only from the female. sharpened, not carinate, and slightly diverging Bakeriella incompleta differs from other in posteriorly; median width of the disc slightly species the genus in having the median longitu- dinal greater than its length; disc slightly shorter than pronotal carina incomplete anteriorly, not mesoscutum and scutellum together. Notauli not reaching the anterior transerve carina, but reach- reaching the anterior margin of mesoscutum, con- ing the posterior margin. In B. cristata Evans this widened carina is verging posteriorly, posteriorly, and incomplete— posteriorly. slightly curved anteriorly. Parapsidal furrows Etymology. The specific name refers to the straight, somewhat more enlarged posteriorly, fact that the median longitudinal carina of the occupying only the posterior half of mesocurum pronotal disc is incomplete. Volume 3. 1 994 147

1-2. Bakeriella n. = Figs mcompleta, sp 1, head and mesosoma, dorsal view; 2, mesopleurum, lateral view. Scale 0.5 mm.

Bakeriella n. grossensis, sp. ened, anterior tooth largest, the second of median — size. Clypeus with median lobe angulate, without Holotype female. Brazil: Barra do Tapirape, median carina, clypeus forming a small excavated MT; XII.1960; unknown trap; B. Malkin, col. reentrance in the region of contact with frons. Eyes (MZSP). with hairs. WH 1.08 X 0.58 X — sparse long LH; WF Description ofholotypefemale. Length of body WH; WF 1.16 X HE; WOT 1.62 X OOL. Antennal 3.75 2.4 mm; LFW mm; head and mesosoma black, scrobe not carinate. Distance posterior ocelli from metasoma black with apex somewhat brownish; vertex about 2.8 X diameter of ocelli, angle of mandible palpi brownish; yellowish with teeth ocellar triangule less than a right angle. Vertex antenna with darker; yellowish scape darker; rounded, very slightly angled laterally. Distance tegula yellowish; coxae and femora brown, tro- from the top of eyes to crest of vertex about 0.57 X chanters, tibiae and tarsi yellowish; wings HE. Length of first four antennal segments in a subhyaline, veins brown. ratio about 20:8:7:7. Temple not carinate. Frons Head. Mandibles with 5 teeth, the basal teeth coriaceous with punctures which are separated by small and rounded off, the 2 apical teeth sharp- 1-3 X their own diameters. 148 Journal of Hymenoptera Research

Mesosoma. Pronotal disc coriaceous, punc- the occipital carina being obscured in full frontal tures similar to frons, with straight anterior trans- view of head. In addition, the mandible of B. verse carina, without median longitudinal carina, grossensis has two big apical teeth, while B. polita posterior margin paralelled by series of small has one and B. —inconspicua has three. foveae; sides of disc subparalell; median width of Etymology. The specific name is in reference disc 1.17 X its length. Pronotal disc 0.77 longer to the Mato Grosso State, region of origin of the than mesoscutum and scutellum together. Ante- holotype. rior half of mesoscutum alutaceous, posterior half coriaceous. Notauli straight and very thin, vesti- Bakeriella montivaga (Kieffer) 1910b gial anteriorly, widened posteriorly, reaching both anterior and posterior margin of mesoscutum. Epyris montivagus Kieffer 1910b:31; Evans 1969:324 Bakeriella Evans 1979:261 Parapsidal furrows straight, occupying only pos- montivaga: terior half of mesoscutum and reaching its poste- This species is recorded from Peru, Colombia rior margin. Scutellar pits transverse, 2.0 X as wide (Kieffer 1910b), Bolivia, Ecuador, Venezuela, as long, separated by thin septum. Propodeal disc Panama, Costa Rica, Honduras, Mexico (Evans 1.05 X as wide as long, with three discal carinae, 1969), and now from Brazil. median one complete and lateral ones converging Four females have been studied: Balsamo, SP, behind, not reaching median carina; space be- Brazil, one female, 28. IV. 1988, tween discal carinae with small transverse striae, yellow pan trap, rubber-tree no collector (UFES). The space between discal carinae and lateral carina crop, given others were from Pichincha, Centro Cientifico R. polished; disc without sublateral carinae; lateral Ecuador, 21.XII.1980, 1.1.1981, 4.1.1981, carinae boardered by a depression with small Palangue, forest, S. Sandoval, col. (MZSP). transverse striae; posterior angle of disc foveolate; These are similar to the declivity of propodeum with median carina and specimens type; disc with the sides perpendicular striae. Metapleuron alutaceous, pronotal diverging posteriorly; WH 0.99 to 1.17 X LH; WF 0.53 to 0.59 X WH; WF without striae. Mesopleuron coriaceous, with large 1.18 to 1.46 X HE; OOL 1.63 to 1.94 X ventral fovea occupying the lower half of WOT; propodeal disc slightly wider than long. The line mesopleuron, with deep pit in upper margin of of small foveae near the posterior of the lower fovea, with a fovea in the anterior part of margin pronotal disc is weakly in the mesopleuron and with long fovea in upper part of angled anteriorly middle. This character is common in males. mesopleuron opened anteriorly. Mesosternum only Evans (1969) considered the possibility that B. coriaceous, with pit in the middle of posterior quinquepartita (Kieffer) and B. region. Midtibia spinose. Forefemur 2.5 X as long montivaga represent variations of a both as thick, midfemur 1.7 X. Claws trifid, first tooth single species. Indeed, species are similar and the differences between rounded off, the others sharpened. very only them were out in Evans'(1979) to Metasoma. Wholly polished, terga and sterna pointed key females. with few setae. Setae of terga on posterior and lateral margins. Transverse section of metasoma Bakeriella reclusa (Evans) 1969 subcircular. Length of mesosoma slightly longer than the metasoma. Valvae III setae brown with on Epyris reclusus Evans 1969:329 apex. — Bakeriella reclusa: Evans 1979:261 Remarks. This species is known only from This species was described from Costa Rica the holotype. It runs to B. inconspicua and B. polita and El Salvador. Evans (1979) studied a female in the key written by Evans (1979). B. grossensis from Ecuador. I have examined a male from Ecua- differs from B. inconspicua in having the scutellar dor: Morona Santiago, Los Tayos, 3. VIII. 1976, pits tran verse, the claws trifid, the eyes not strongly DeVries, col. (MZSP). Measurements for this spe- setose, the metasoma black with reddish and apex cies are: LFW = 2.13 mm, WH 1.2 X LH; WF 0.55 X the mandible 5-toothed. B. grossensis differs from WH; WF 1 .06 X HE; WOT 1 .44 X WOT. The pronotal B. polita in lacking a median carina in the clypeus, disc is slightly wider than long, and the propodeal in lacking a lateral carina on the pronotal disc, the disc is as long as wide. head being not well developed above the eyes and Volume 3. 1994 149

The following key to females with an anterior transverse carina on the pronotal disc is modified B. dentata B. from Evans (1979:258) and includes Azevedo 1991, 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 of 5(3) Sides 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 Sides of 6(5') pronotal disc sharpened; mandible with one large apical tooth; scutellar pits slightly wider than long polita Evans 1964 6' Sides of disc pronotal 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. Internal 9(7') 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 Evans, H. E. 1969. A revision of the genus Epyris in the spider collections of the world. Gainesville, E.J. Brill /Fauna Americas (Hymenoptera: Bethylidae). Transactions of & Flora Publications, 220p. the American Entomological Society 95(2): 181-352. Azevedo, C. O. 1991. Especie nova e notas sobre Bakeriella Evans, H. E. 1979. A reconsideration of the genus Bakeriella Kieffer (Hymenoptera, Bethylidae) da regiao de Sao (Hymenoptera: Bethylidae). Journal of the Nexv York En- Carlos, SP. Revista Brasileira de Entomologia 35(3): 535-8 tomological Society 87(3): 256-66. Eady, R. D. 1968. Some illustrations of microculpture in the Gordh, G. and L. Moczar. 1990 A catalog of the world Hymenoptera. Procedings of the Royal Entomological Soci- Bethylidae (Hymenoptera). Memoirs of the American En- ety of London, series A General Entomology 43(3-6): 66-72. tomological Institute 46: 1-364. E. of the 1910a. Evans, H. 1964. A Synopsis American Bethvlidae Kieffer, J. J Description de Nouveaux Micro-

(Hymenoptera : Aculeata). Bulletin of the Museum of Com- hymenopteres du Bresil. Annates de la Societe de parative Zoology, Harvard University 132(1): 1-222. Entomologique de France 78: 287-348. 1910b. Evans, H. E. 1965. Further studies on Neotropical Epynni Kieffer, J. J. Description de Nouveaux Microhy- (Hymenoptera: Bethylidae). Psychef Cambngde) 72(4): menopteres du Bresil. Annates de la Societe de 265-78. 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 study combined behavioural tests with micro- been demonstrated in a number of species of para- scopic techniques in order to identify the position sitic wasps (Weseloh 1976, 1980; Tagawa 1977, of the sex pheromone gland in C. rubecula. 1983). Although unable to localize the sex phero- mone source more precisely than the abdominal MATERIALS AND METHODS region, Vinson (1978) suggested that Dufour's gland produced a sex identification pheromone in Behavioural Experiments. —C. rubecula was Cardiochiles nigriceps. Obara and Kitano (1974) first reared in the laboratory on larvae of Pieris rapae L. demonstrated that the source of the female sex using the methods of Keller (1990). Females be- in Cotesia is pheromone glomerata the tip of the tween one and four days of age were frozen at -15° abdomen, near the ovipositor. Tagawa (1977) later C for approximately 30 minutes before being dis- described a pair of secretory glands located at the sected. of the base second valvifer on the ninth abdominal Using the same experimental protocol, we tergite of this species, which he suggested was performed two experiments to test the respon- for sex responsible pheromone production. A sub- siveness of males to different female body parts. In sequent study (Tagawa 1983) revealed the exist- the first, the head and mesosoma (thorax plus first ence of similar glands in a corresponding position abdominal segment) were tested against the in all seven braconid species examined. Upon metasoma (remaining abdominal segments), and his initial revising findings (Weseloh 1976), in the second the genital capsule (terminal ab- Weseloh (1980) likewise concluded that paired dominal segment plus ovipositor) was tested epidermal glands on the last abdominal tergite against the remaining anterior portion of the were the sex pheromone source in Cotesia metasoma. melanoscelus. A glass microscope slide was divided into two Previous field observations (Keller unpub- halves and the two body parts being tested were established lished) that courtship in male Cotesia placed randomly in the middle of either half, rubecula (Marshall) (Hymenoptera: Braconidae) is presenting a choice to the males. The slide was elicited a female sex by pheromone. The present clamped at its midpoint and held horizontally at a 152 Journal of Hymenoptera Research

in an 750 and mounted on height of 30 cm by a small metal stand. Tests were dried EMSCOPE CPD aluminium stubs Acheson conducted in a flight tunnel (Keller 1990) at a wind TAAB using Electrodag coated vacuum speed of 30 cm/s. Males were kept in a separate 915. They were by evaporation and and examined in cage and removed for testing individually with a with carbon gold palladium minimum of disturbance. They were released 30 an ETEC Autoscan scanning electron microscope cm downwind of the slide and observed to deter- at an accelerating voltage of 20 kV. —Genital were mine (1) which half of the slide they landed on, (2) Light Microscopy (LM). capsules for 4 hours in fixative whether subsequently they oriented to either body removed and immersed (3% made in 0.1 segment and performed typical elements of court- glutaraldehyde/3% formaldehyde up to which had been ship (Field and Keller 1993), and (3) whether they M phosphate buffer, pH 7.4, were attempted to copulate with either body segment. added 2.5% polyvinyl pyrrolidone). They The same males were then tested again, but washed in 0.1 M phosphate buffer overnight and this time the female body segments were washed then dehydrated by passing through a series of re- alcohols. After in in diethyl ether prior to testing. This solvent washing propylene oxide, they moved all contaminating pheromone from the were infiltrated with increasing concentrations of surface of the body segment (Golub and TAAB epoxy embedding resin over 48 hours and Weatherston 1984). Therefore any subsequent then embedded in resin by curing at 60° C for a emission of pheromone could be due only to con- further 48 hours. tinued leaking of the pheromone onto the surface Using glass knives in a Sorvall MT2-B "Porter- transverse sections from an internal glandular reservoir. The prefer- Blum" ultramicrotome, serial the ence of courting males for a particular segment of 0.5 urn thickness were cut, starting from near after washing would therefore localize the phero- base of the ovipositor and proceeding anteriorly. mone gland to somewhere within that segment. These were stained using 0.025% Toluidine Blue Most males were successful in navigating to in 0.5% Borate buffer and photographed with a the slide; those that did not were discarded after a Wild MPS 45 photoautomat on Ilford Pan-f 50 58 few attempts. To lessen the effects of pheromone ASA film. Agfa x 3 light green and Wratten contamination and depletion of pheromone re- filters were used to achieve suitable contrast. Males serves, the stand was washed with diethyl ether were also subject to the same processing proce- and body segments and slides were replaced at dures and serial sections were cut from the poste- of the regular intervals throughout the course of the rior tip abdomen. experiments. Variation in attractiveness of females Transmission Electron Microscopy (TEM). -Tis- was assumed to be negligible. Fifty males were sue samples were prepared in the same mariner as tested in each experiment and data were analysed for LM. Gold to silver sections (0.08 mm) were cut using two-tailed binomial tests (Zar 1984). with a Diatome diamond knife using a Reichert- Scanning Electron Microscopy (SEM).—Two Jung Ultracut and stained with saturated uranyl methods were used to prepare specimens for SEM acetate in 70% alcohol for 20 minutes, followed by observations. In the first, newly-dissected speci- lead citrate for 12 minutes (Reynolds, 1963). They mens were immersed in Peterson's KAA (kero- were examined in a JEOL 100S transmission elec- sene (10%), 95% alcohol (75%), glacial acetic acid tron microscope at an accelerating voltage of 80 (15%)) prior to washing in 0.1 M phosphate buffer kV. and dehydration through an alcohol series. This technique is commonly used to preserve larval RESULTS insects because it distends structures that may — otherwise lose shape (Smithers 1981). Behavioural Experiments. In the first experi- The second method was to "clear" the speci- ment, neither the head/mesosoma nor the mens by gently warming in 10% KOH for 30-60 metasoma was found to be more attractive when minutes. This removed most tissue and allowed the segments had not been washed in ether. How- dissection of the specimens so that both internal ever, when surface contamination had been re- and external cuticular structures could be exam- moved with ether, males responded preferentially ined under SEM. Following dehydration, the KAA- toward the metasoma, indicating that it was the treated and cleared specimens were critical-point source of the sex pheromone (Table 1). Volume 3. 1994 153

Fig. 1. SEM of lateral view of in Cotesia rubecula. Arrow indicates of genital capsule position hair plate and pores (see Fig. 2); - - n ninth tergite; o ovipositor. Scale bar = 100 urn 2. Fig. High magnification SEM of hair plate and pores. Scale bar = 10 urn 3. Fig. Transverse section of under LM two sex - genital capsule, showing possible pheromone-producing glands: gp glands associated with surface - at base of - = pores; gb gland genital capsule; hp hair plate. Scale bar 30 urn 4. TEM of basal section of - - Fig. 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).

Number Responding

Treatment Response Metasoma Head+Mesosoma

Unwashed Land 23 27 Court 23 27 Copulate 10 22 Washed Land 36 l" Court 36 r Copulate 19 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).

Number Responding

Treatment Response Anterior Metasoma Genital Capsule

Unwashed Land

Washed Volume 3, 1994 155 may have been a result of the small proportion of ence of a particular threshold amount of chemical sections cut (approx. 1/10) that were actually re- may "prime" males so that they are ready to trieved, stained and mounted for examination. mount and copulate after a certain amount of The results of ultrastructural examination of this courtship, but an object of appropriate size, shape gland were inconclusive and failed to offer any and colour in the vicinity of a pheromone source indication of the organelles present or whether it may be required to release copulation behavior. was secretory in nature. One male did attempt copulation with the tiny TEM examination revealed the presence of a genital capsule (0.5 mm in length), so the phero- of second pair secretory glands (Fig. 4), located at mone alone may be sufficient to elicit attempts at thebase of the genital capsule and adjacent to folds copulation in some cases. of membrane which are presumably stretched and The histology and electron microscopy failed compressed during movements of the ovipositor. to resolve the exact location of the sex pheromone The ultrastructure of these glands was also indis- gland within the genital capsule, but narrowed the tinct, but pore canals leading to the surface and an possibilities to two structures. One of these is a abundance of secretory bodies were visible on or group of cells located directly underneath surface just under the surface of the cuticle. 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 DISCUSSION valvifer, which places this gland in approximately the same position as the putative pheromone The behavioural experiments clearly demon- glands indicated by Tagawa (1977, 1983) for a strate that the genital capsule is the source of the number of other closely-related braconids. Al- sex pheromone in C. rubeculn. With all surface though this constitutes circumstantial evidence pheromone removed, males were attracted to the that this is the sex pheromone gland in C. rubecula, in to the metasoma preference head/mesosoma the absence of any clearly defined secretory prod- and, when given a further choice, made their ucts in the ultrastructural examinations leaves initial approach more often to the genital capsule some uncertainty. The pores associated with these rather than to the anterior portion of the metasoma. glands required thorough washing before they This is indicative of the continuous emission of became visible by SEM. from internal pheromone an reservoir located in Another possible source of sex pheromone is the genital capsule, which provides an olfactory a gland located further toward the base of the cue to the searching wasp. genital capsule. The most striking feature of this There are, however, two anomalous results gland is an abundance of associated secretory which require explanation. Firstly, a single male bodies amassed on and under the surface of the chose the washed head/mesosoma over the cuticle which are visible under TEM (Fig. 4). One in the first metasoma experiment. This could have possibility is that these secretions act as a lubricant been to due incomplete washing of the body parts. for the ovipositor as it is extended and retracted

Secondly, although significantly more males from the metasoma . However, if it is the sex phero- landed near and courted the genital capsule as mone gland, then such movements would un- opposed to the anterior portion of the metasoma, doubtedly smear the secretions over the cuticle, with the they attempted copulation genital cap- which would be ideal for pheromone dispersal. sule less This is frequently. not considered to be An additional possibility is that the one secretion evidence that the sex pheromone gland is in the has the dual function of lubrication and sex attrac- anterior portion of the metasoma. Rather, it sug- tant. In the absence of isolation and identification that a gests visual cue is involved in triggering of the pheromone, the question of which of the two at the male. attempts copulation by Upwind flight, glands is the source of sex pheromone in C. rubecula landing, orienting and courting all appear to be will remain unresolved. stimulated entirely by chemical cues. The pres- 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 Snodgrass, RE. 1935. Principles of Insect Morphology. McGraw- signalling in the parasitic wasp Cotesia rubecula. journal Hill Book Co., New York. 667 pp. Insect 6: 1977. Localization of Behaviour 737-750 Tagawa, J. and histology of the sex phero- Golub, MA. and Weatherston, I. 1984. Techniques for ex- mone-producing gland in the parasitic wasp, Apanteles tracting and collecting sex pheromones from live insects glomeratus. journal of Insect Physiology, 23: 49-56 and from artificial sources. In H. and in Hummel, Miller, Tagawa, J. 1983 Female sex pheromone glands the para- T., eds. .Techniques in pheromone research. Springer- Verlag, sitic wasps, genus Apanteles. Applied Entomology and New York. 464 pp. Zoology, 18: 416-427.

Keller, MA. 1991 . Responses of the parasitoid Cotesia rubecula Vinson, SB. 1978. Courtship behavior and source of a sexual to its host Pieris rapae in a flight tunnel. Entomologia pheromone from Cardiochiles nigriceps. Annals of the Experiementalis et Applicata, 57: 243-249. Entomological Society of America, 71: 832-837. Obara, M. and Kitano, H. 1974. Studies on the courtship Weseloh, R.M. 1976. Dufour's gland: Source of sex phero- behaviour of Apanteles glomeratus L. -1. Experimental mone in a hymenopterous parasitoid. Science, 193: 695- studies on releaser of wing-vibrating behaviour in the 697. male. Kontyu.il: 208-214. Weseloh, R.M. 1980. Sex pheromone gland of the gypsy moth

Reynolds, E.S. 1963. The use of lead citrate at high pH as an parasitoid, Apanteles melanoscelus : revaluation and ul- electron-opaque stain in electron microscopy. Journal of trastructural survey. Annals of the Entomological Society Cell Biology, 17: 208-212. Of America, 73: 576-580. Smithers, C. 1981 Handbook of Insect Collecting. A.H. and A.W. Zar, J.H. 1984 Biostatistical Analysis. 2nd edition. Prentice- Reed, Sydney. 120 pp. 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 six under Orthomiscus in the Holarctic O. eridolius medusae , leaving species Region: Kasparyan, unicinctus Kasparyan, pectoralis (Hellen), -platyura Mason, simplex (Mason) and (Holmgren). Diagnostic characters, species relationships and a key to the species are given. Diagnostic characters are illustrated by diagrams and photographs.

INTRODUCTION O. medusae, O. amurensis, and O. eridolius. Gupta (1990) transferred the Japanese Kristotomus sim- The genus Orthomiscus was described by Ma- plex Mason, 1962 to this genus. In the present son (1955) to accommodate two new Nearctic study, O. amurensis Kasparyan is transferred to the species, O. platyura and O. leptura. It belongs to the genus Kristotomus as it exhibits most of the charac- Kristotomus-Complex of genera in the tribe ters of that genus, like the tapered and slender Exenterini (Gupta 1990), including Kristotomus ovipositor and sheaths, lower valve of ovipositor Mason, Kerrichia Mason and Orthomiscus Mason. without teeth and egg with a simple stalk and This group is characterized by having the apical anchor. Orthomiscus thus is recognized as a Hol- rim of the hind tibia with a fringe of long, close arctic genus—with six species. bristles on the inner side and with a flat polished Biology. Not much is known about the biol- area on the lower and inner sides between the ogy of Orthomiscus except that O. pectoralis has apical fringe of bristles and the tarsal socket; by been reared from late larval cimbicids (Trichiosoma having the hind tibia widest at apex and truncate; lucorum L. and T. tibiale Stephens) that feed on and by its short ovipositor which hardly surpasses birch and ash (Betula and Crataegus) in Europe the tip of abdomen. These genera also exhibit the (Jussila 1975). usual exenterine characters such as the presence of Material.—This paper is based on the speci- only one spur on the middle tibia and the absence mens present in the collection of the American of spurs on the hind tibia. Entomological Institute, Gainesville (GAINES- Mason (1962) commented on the relationships VILLE). The types of the Russian species were of the genus and included two European species: borrowed from the collections of the Zoological Exenterus unicinctus Holmgren, 1858 (with which Institute, Russian Academy of Science, St. Peters- he synonymized O. leptura) and Cteniscus pectoralis burg (ST. PETERSBURG) through the kindness of Hellen, 1951. Kasparyan (1976, 1986, 1990) de- Dr. D. R. Kasparyan. Under synonymy of various scribed three additional new species from Russia: species, the location of the type is indicated by the

'Florida Agricultural Experiment Station Journal Series R-03128. 158 Journal of Hymenoptera Research

where the is city type housed (see Townes, Momoi decurved, not tapered apically and not extending Townes 1965 for to & details). The key the species beyond apex of abdomen. Upper valve of oviposi- is based on the females which are more diagnostic. tor flattened dorsally in basal half. Lower valve of ovipositor with an obliquely serrated edge (Figs. Genus ORTHOMISCUS Mason 10, 12, 18, 29). Ovipositor sheaths short, broadly triangular or sometimes a little slender. Female Orthomiscus Mason, 1955. Canadian of 33: Journal Zoology, subgenital plate thin, creased medially and pinched 63. apically. Egg stalk short, anchor simple, button- Type: Orthomiscus platyura Mason; original designation. like or flattened and enclosing part of the Taxonomy: Mason 1962: 1273, Gupta 1990: 7; Kasparyan egg 1990: 36. (Figs. 4-14).

Diagnostic characters. —Head subcuboidal, SPECIES RELATIONSHIPS swollen behind eyes. Mandibular teeth equal or the of subequal. Occipital and hypostomal carinae com- On basis egg structure, the six species of Orthomiscus form two distinct plete. Occipital carina bent inwards, meeting species groups. hypostomal carina at a right angle above base of Group 1 comprises platyura Mason (type-species), medusae and eridolius mandible by a distance equal to the basal width of Kasparyan Kasparyan (Figs. mandible. Epomia distinct. Epicnemial carina ex- 4, 7, 9), where the egg stalk is short and wide and the anchor is tending at least to the middle of hind margin of four-pronged and encloses part of pronotum. Basal area of propodeum confluent the egg. Group 2 comprises pectoralis (Hellen), unicinctus and with areola or only partly separated from it. Legs (Holmgren) simplex (Mason) (Figs. slender, elongate. Tarsal claws with 2-6 basal teeth. 8, 12, 30) where the egg has a simple short stalk and a short Areolet present, oblique, sometimes second button-like anchor. In the latter group, the stalk intercubitus partly weak. Nervellus inclivous, in- arises apicoventrally in pectoralis and unicinctus while in it is tercepted in its lower 0.1 5 to 0.3. Tergitel elongate, simplex central and the anchor is flat and oval. usually about 2.0x as long as its apical width, its dorsomedian and dorsolateral carinae sharp and Gupta (1990) stated that the plesiomorphic distinct; dorsolateral carina usually extending to state of the egg in Cteniscini was probably an egg the entire with a short stalk and anchor. This of is length of the tergite and passing just type egg seen in some dorsad of spiracle and not broken there. Tergite 2 generalized members of the genus Kristotomus usually with a pair of strong to weak oblique (like claviventris, santoshae and basolateral grooves. Ovipositor short, stout and ctenonyx). Kristotomus also exhibits more general-

eridolius pectoralis

medusae body unicinctus punctate platyura

wide mandible (reversal) second segment -of maxillary palp swollen T2 with deep grooves simplex

modified egg mandible anchor/stalk widened simple apically Kristotomus egg anchor/stalk stalk

the Diagram depicting 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 2 4a Tergite 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 2 Tergite without oblique grooves. Abdominal tergites black, with margins of tergites 2-3 (sometimes 4-5 of female black. at its also) yellow. Mesosternum Nervellus intercepted 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 creased subgenital plate sharply medioventrally. Ovipositor slender (Fig. 29). Egg with a simple, oval anchor (Fig. 30). japan 3. simplex (Mason) 5b black. Mesopleurum Tarsal claws with a pair of weak teeth. Female subgenital plate convex ventrally, not creased. thick strongly Ovipositor (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) and pedicel pale brownish. Thorax and with Figs. 8, 12 propodeum black, pronotum, scutellum, metascutellum, a mark above epicnemial carina, Ctemscus pectoralis Hellen 1951. Notulae Entomologicae mark above middle coxa, tegula and subtegular 31: 31. F. des. Female, Finland: Esbo Helsingfors, Type: ridge, yellowish brown. Propleurum yellow. Fore (HELSINKI). — Kernch 1952. Bulletin of the British and middle coxae and trochanters yellow. Fore Museum (Natural History) Entomology Series, 2 (6): and middle femora yellowish brown, their tibiae 432. F. des., distr. and tarsi Hind coxa reddish brown. Hind Orthomiscus pectoralis: Mason, 1962. Canadian Entomologist, paler. tibia and tarsus more brownish to reddish 94: 1 274. n. comb. Japan: Nagano Prefecture: Kamikochi. femur, — Townes, Momoi & Townes, 1965. Memoirs of the brown. Abdomen black above, yellow below. American Institute, 5: 109. cat. — Jussila, Entomological Tergites 1-6 with narrow, pale apical margins and 1975. Acta Entomologica Fennica,41 (2): 53. des. of male, tergite 7 with an obscure, triangular yellow mark. host records. — 1976. Entomologicheskoe Kasparyan, sheaths infuscate above and Obozrenie, 55: 137. — Kasparyan, 1977. Nasekomye Ovipositor yellow, rufous towards Mongoliaca, 5: 463. Mongolia. — Kasparyan, 1990. Fauna apex. 3 43. — with a short of USSR Insecta, Hymenoptera, (2): key, des., fig. Egg. (Fig. 8) Egg apicoventral Russia. stalk and a —button-like anchor. Length. 5.0-10.0 mm. Fore wing 3.8-8.5 mm.; No of this were available specimens species Ovipositor about 0.5 mm. for study. Kerrich (1952) and Kasparyan (1990) Distribution. —Europe (Finland), Russia, have detailed of this provided descriptions spe- Mongolia, Japan. that follows is from cies. The description adapted Mason (1962) reported this species from Japan Kerrich (1952).— and included two females from Nagano Prefec- Diagnosis. This species appears different ture: Kamikochi, collected by Townes in 1954. from the others by having the face and thorax These specimens could not be located in the AEI extensively punctate and the propodeum rugoso- collections. punctate (cf. Kerrich, 1952 description). All other species have only weak to indistinct body puncta- 2. Orthomiscus unicinctus (Holmgren) tion. 4-5 as as wide. The has a Tergites long egg Figs. 6, 11, 19-23,36,39 short apicoventral stalk and a small button-like anchor (Fig. 8). Exen terusumanct us Holmgren, 1858. SvenskaVetensk.-Akad. Handlirsch 1: 234. F. des. Male and female. —Face moderately punctate. (n. f), M, key, Lectotype (selected Roman, 1914), Female, Sweden: Clypeus with fine punctures. Epomia short and by Lapland: Tarna (STOCKHOLM). weak. Mandibular teeth equal in length. Notauli Exenterus macrocephalus Holmgren, 1858. Svenska Vetensk.- Scutellum convex, sharply impressed. strongly Akad. Handlirsch (n. f), 1: 243. M. des. Type: Male, carina finely punctate. Epicnemial incomplete Sweden (STOCKHOLM). Syn. by Roman, 1914. medially. Sternaulus superficially impressed. Ctemscus macrocephalus:]acob &Tosquinet, 1896. Annalesde la Societe 34: 108. new comb. Propodeum rugose to punctate in places. Areola Entomologique Beige, Cteniscus unicinctus: Roman, 1914. Arkiv for Zoologie, 9 (2): about as long as wide, confluent with basal area. — 17. syn., des. Kerrich, 1942. Transactions of the Soci- Abdomen elongate and slender. Abdomen elon- — ety for British Entomology, 8 (2): 63-64. Britain. and slender. 1 about 1.6x as as its gate Tergite long Kerrich, 1952. Bulletin of the British Museum (Natural its dorsomedian and dorsolateral apical width, History) Entomology Series, 2 (6): 434. key, des., fig. Localities in Swe- carinae strong. Postpetiole with a pitlike depres- Britain, France, Belgium, Germany, den. sion bounded by carinae. Tergite 2 with small but Orthomiscus leptura Mason, 1955. Canadian Journal of Zool- distinct thyridia. Tergites 4-5 as long as their basal ogy, 33: 64. M, F. key, des., fig. Type: Female, U.S.A.: width. abdominal in Apical segments compressed New Hampshire: Pinkham Notch (AEI, Gainesville) female. Female less folded subgenital plate sharply Examined. Syn. by Mason, 1962. than in O. unicinctus (with which Kerrich com- Orthomiscus unicinctus: Mason, 1962. Canadian Entomolo- 94: 1274. distr. — Townes, pared this species). Ovipositor sheaths slender. gist, syn., Japan: Sapporo. Momoi & Townes, 1965. Memoirs of the American En- Ovipositor comparatively stout as is usual in the tomological Institute, 5: 109. cat. — Mason, 1966. Cana- genus (Fig. 12). — dian Entomologist, 98: 48. key. Japan. Carlson, 1979. Color. —Ground color black. Mouthparts, Catalog of Hymenoptera in America North of Mexico, 1 : clypeus, face, and temples broadly, yellow. Scape 377. cat. U.S.A., Canada. — Kasparyan, 1990. Fauna of 161 Volume 3. 1994

45. malar and on lower half, USSR Insecta, Hymenoptera, 3 (2): key, des., fig. eyes, space, temples lower half of its Russia. yellow. Propleurum, pronotum, of upper margin, tegula, subtegular ridge, tip Second of max- lower half Diagnosticfeatures.— segment scutellum, epicnemium, mesosternum, and flat, about 2.0x as wide as abdominal venter, and sides of illary palp widened of mesopleurum, on 2 brown. . to the rest (Fig. 39). Oblique impressions tergite abdomen, yellow Legs yellow yellowish more slender than faint. Ovipositor comparatively Fore and middle coxae and trochanters yellow. with a or with in the previous species (Figs. 20, 21). Egg Hind coxa largely brownish yellow large a button-like anchor Hind short apicoventral stalk and black or brown patches. Hind femur brown. on its basal half. in from Ontario (Figs. 6, 11). Temple yellow only coxa largely yellow specimens Hind coxa with 2 Scutellum black except at apex. and blackish in specimens from Alaska. Tergites black or with blackish brown patches. Abdomen largely and 3 often with triangular yellow patches on 2 and 3. Tho- In males the and with yellow patches tergites their apical margins yellow. yellow that of female to the level of rax of male largely black, generally along the inner orbits extending only variable. its lower yellow in lower half, but color antennal sockets, temple yellow on 0.3, Male and female—Flagellum with 23-24 seg- and thorax almost wholly black. The color is quite ments in male and 23-26 segments in female. First variable. Second segment about 1.4x as long as the second. The European specimens were not examined widened and flat, about segment of maxillary palp to ascertain the range of variation. Kerrich (1952: rest of the 39). described this 2.0x as wide as the segments (Fig. 434) and Kasparyan (1990: 45) have teeth in but lower tooth Mandibular equal length species in detail. than the ca- — oval, with a more prominent upper. Hypostomal Egg (Fig. 6, 11, 21). Egg small, not raised at a button-like anchor. rina of normal shape, conspicuously short apicoventral stalk and with carina. Vertex widened — fore 5.0-7.0 its junction occipital Length. 6.0-8.5 mm.; wing mm.; Interocellar distance 0.25x the Kerrich stated that a posteriorly. ovipositor 0.4-0.6 mm. (1952) ocellocular distance. female measuring 4.0 mm. was seen from Lapland. Mesoscutum rather — Thorax subpolished. Specimens examined. CANADA: Quebec: M. strongly convex. Notauli distinctly impressed up Stoneham, 1 male, 1 female, 21.VI.1938, H. & to middle of mesoscutum. Lateral carinae of scutel- Townes. U.S.A.: New Hampshire: Pinkham Notch, of scutellum. lum distinct only at base Epomia 1 female, 25.VI.1938, H. & M. Townes; Mt. Madi- Sternaulus and Ver- sharp across pronotal collar. son, 3 males, 24.VI.1938, H. & M. Townes. epicnemial carina moderately impressed. mont: L. Willoughby, 2 male, 17, 25.VI.1945, C.P. Propodeum convex, subpolished. Areola (Fig. 23) Alexander. (Types and paratypes in AEI, with confluent widened at its junction costulae, Gainesville). Non-type material: U.S.A.: Oregon: with basal area. Basal vein of fore wing uniformly Corvallis, 2 males, 1 female, V-VII. 1978, H. & M. Nervellus 6- but weakly arched (Fig. 36). intercepted Townes. Alaska: Anchorage, 1 female, in its lower 0.25-0.33. 16.VII.1976, Petre Rush. CANADA: Ontario: 1 slender Abdomen subpolished. Tergite (Fig. Cumberland, 1 male, 3 females, VI.1975, L. Ling. and 19), its dorsomedian carinae strong extending Distribution.—-U.S.A., Canada, Europe (Brit- of the its dorsolateral to 0.7-0.8 the length tergite, ain, France, Belgium, Germany, Sweden), Russia, carina and to the entire of sharp extending length Japan. 2 distinct and the tergite. Tergite with thyridia Female faint oblique grooves. subgenital plate 3. Orthomiscus simplex (Mason) and folded medially pointed apically. Ovipositor Figs. 26-30, 42 slender, of uniform diameter (Fig. 21). Ovipositor 1966. Canadian 98: narrow. Kristotomus simplex Mason, Entomologist, sheaths (Figs. 20, 21) comparatively Egg 46. M, F. key, des., fig. Type: Female, Japan: Nagano situated — anchor small and apicoventrally (Fig. 6). Prefecture: Kamikochi (GAINESVILLE. Examined. — black. Thorax almost Obozrenie, 55: Color. Ground color Kasparyan, 1976. Entomologicheskoe wholly black in male and some females. Abdomi- 150. (Entomological Review, 55: 108). key. Contrib. Amer. Ent. Inst., below, Orthomiscus simplex: Gupta, 1990. nal tergites largely black. Scape, pedicel 8. n. comb. — 1990. Fauna of USSR of 25 (6): Kasparyan, mouthparts, clypeus, face, inner orbits up to top Insecta, Hymenoptera, 3 (2): 44. key, des., fig. Japan. 162 Journal of Hymenoptera Research

Diagnostic features. —Mandible widened femur, tibia and tarsus blackish brown (male) to apically, its lower tooth wider and longer than the yellowish brown (female). Pronotum of female more and upper (Fig. 27). Hypostomal carina slightly raised extensively yellow mesosternum and lower of also at its junction with occipital carina (Fig. 42). Basal part mesopleurum may be yellow. vein of fore wing curved. Egg reniform-ovate, Abdominal tergites black with yellow apical on with a short central stalk and a simple anchor. Egg stripes, particularly— tergites 2-3. surface coarsely granular. Head and thorax in Egg. Reniform-ovate (Fig. 30). Stalk shorter female yellow on lower half and black on upper than its own diameter, arising near the middle of the Anchor about 0.33 as as flat half, in male largely black. Abdominal tergites egg. long the egg, and oval. black, their apical margins— yellow. Egg— surface coarsely granular. Male and female. Scape about 1.5x as long as Length. 4.5-6.5 mm.; fore wing 4.5-6.0 mm.; about 0.5 mm. wide. Flagellum with 24 segments. First flagellar ovipositor — long. segment 1.4x as long as the second. Face (and rest Specimens examined. JAPAN: Nagano Prefec- ture: 6 2 of head also) smooth and shiny. Mandible wid- Kamikochi, males and females (holotype, and 23-31. VII. ened apically, the lower tooth wider and slightly allotype paratypes), 1954, Townes longer than the upper (Fig. 27). Malar space 0.4x family (AEI, Gainesville). Distribution. — the basal width of mandible. Maxillary palp seg- Japan. Discussion. —Mason this ments normal, not swollen (Fig. 42). Hypostomal (1966) placed species under it carina slightly raised at its junction with occipital Kristotomus, stating that was the most of the carina (Fig. 42). Vertex widened behind eyes (Fig. primitive member genus and closely related 26). Interocellar distance 0.4x the ocellocular dis- to Ortlnmuscus, particularly in the structure of the tance. egg and development of the dorsolateral carina of the Thorax largely smooth and shiny. Notauli petiole. Apart from the widened mandible, distinct anteriorly. Lateral carinae of scutellum this species shares most characters of Orthomiscus, distinct. Epomia normal, not very sharp. particularly the structure of the ovipositor, in which Epicnemial carina extending to half the height of the lower valve has teeth-like indentations. The mesopleurum. Sternaulus indistinct. Propodeum dorsolateral carina of the petiole and the egg struc- normally areola ted. Petiolar area small and round. ture also relate it to the present genus. Apical transverse carina uniformly curved be- tween its junction with lateral longitudinal cari- 4. Orthomiscus medusae Kasparyan nae. Basal vein in fore wing medially curved, more Figs. 7, 10, 31,38, 41 strongly so than in the preceding species. Second intercubitus of areolet sometimes absent. Orthomiscus medusae Kasparyan, 1976. Entomologicheskoe Obozrenie, 55: 140 (Entomological Review, 55 (1): 99). Abdomen smooth and shiny. Tergite 1 about F. M, des., fig. Type: Female, Russia: Sakhalin: Kunashir 2.25x as long as its apical width, its dorsomedian — Island (ST. PETERSBURG). Examined Kasparyan, carinae to about 0.75 its its 1990. extending up length, Fauna of USSR Insecta, Hymenoptera, 3 (2): 42. dorsolateral carina Russia. complete, sinuate and passing key, des., fig. above the just spiracle (Fig. 28). Tergite 2 with — weak basolateral oblique grooves. Female Diagnostic features. Mandible widened subgenital plate strongly creased and pointed apically, the lower tooth longer than the upper apically. Ovipositor (Fig. 29) somewhat slender (Fig. 31). Hypostomal carina raised at its junction and cylindrical, its teeth weak but present at the with occipital carina. Scutellum subconvex. Basal vein of fore tip. Ovipositor sheaths small and slender. Male wing strongly curved medially (Fig. 2 with faint claspers broader— than in other species. 38). Tergite very to indistinct oblique Color. Ground color black. Male compara- grooves. Female subgenital plate convex ventrally, not tively darker than the female. Face, clypeus, man- sharply creased. Egg (Figs. 7, 10) with a short dibles, lower half of temples, pronotal collar, hind medioventral stalk and a four-pronged anchor corner of pronotum, whole of fore and middle enclosing the egg. Thorax and abdomen black; legs, hind coxae and trochanters, and abdominal sometimes pronotal collar and presternum yellow venter, yellow. Ventral side of scape and pedicel and tergites 2-3 with yellow marks, particularly in and apex of scutellum yellowish-brown. Hind females. Volume 3. 1 994 163 — Male and female. Flagellum with 23-24 seg- enclosing part of the egg, somewhat similar to ments in male and in female. First 25-27 segments that seen in—O. platyura. flagellar segment about 1 .3x as long as the second. Length. 7.0-9.0 mm.; fore wing 5.5-7.0 mm.; Face and clypeus subpolished, with sparse punc- Ovipositor about 0.5 mm. tures. Malar space about 0.33-0.4.X (male) to 0.4- Specimens examined. —RUSSIA: SAKHALIN: the 0.5x (female) basal width of mandible. Man- Kunashir Island: Sernovodsk, 1 female (paratype), dible widened its slightly apically (Fig. 31), lower 26.VIII.1973, D. R. Kasparyan. JAPAN: Nagano tooth a little longer than the upper. Maxillary palp Prefecture: Kamikochi, 1 male, 22. VII.1954, 1 male, slender, normal (Fig. 41). Vertex widened posteri- 24. VII. 1954 (labeled O. unicinctus Tow.), Townes smooth and In terocellar orly, shiny. distance 0.25x family (AEI, Gainesville).— the ocellocular distance. Distribution. —Russia, Japan. Epomia present but not very strong. Notauli Relationships. Structurally this species is close distinct to middle of mesoscutum. Scutellum to O. simplex (Mason) from which it differs mainly subconvex, its lateral carinae confined to base. in the complex structure of the egg anchor, black carina Epicnemial strong, ending in middle of mesopleurum, male tarsal claws with a pair of hind of in- margin pronotum. Sternaulus weakly weak teeth at base, female subgenital plate convex dicated areolated its anteriorly. Propodeum but but not strongly creased ventrally and by possess- dorsal and sublateral carinae longitudinal quite ing a comparatively thicker ovipositor. The egg often erased the costulae or beyond absent; com- structure relates it to O. plati/ura. bined length of basal area and areola approxi- mately 1 .5x the length of apical area (according to 5. Orthomiscus platyura Mason the In original description). one paratype and in Figs. 1-5, 15-18,37,40 the Japanese specimens examined in the AEI col- lection, the areola is narrow, 2. Ox as long as wide, Orthomiscus platyura Mason, 1955. Canadian Journal of Zool- 33: 66. F. confluent with the basal area, and a little widened ogy, M, key, des., fig. Type: Female, U.S.A.: Rhode Island: Examined. — at the costulae. Westerly (GAINESVILLE). Basal vein curved medially (Fig. Carlson, 1979. Catalog of Hymenoptera in America 38). Tarsal claws with a pair of weak teeth at base. North of Mexico, 1:377. cat. U.S.A. New Hampshire, Tergite 1 2.0x as long as its apical width; its Vermont, California. Canada: Quebec, Ontario. dorsomedian carinae extending to 0.75 its length and its dorsolateral carinae and — strong complete. Diagnosticfeatures. Areola narrow, parallel- with a median oval Postpetiole depression. Terg- sided and confluent or partly separated from the ite 2 with rather weak oblique basolateral impres- basal area. Tergite 2 with rather deep oblique sions. Female subgenital plate convex, not strongly basolateral grooves. Ovipositor stout, its sheaths creased ventrally, apically narrow and pinched. broadly triangular. Egg with a four-lobed anchor, stout Ovipositor (Fig. 10). Ovipositor sheaths as large as the egg. Scutellum mostly yellow. Hind broad. coxa Abdominal — yellow. tergites beyond first red- Color. Ground colorblack. Scape and pedicel dish brown to reddish yellow. reddish brown and — dorsally lighter ventrally. Fla- Male and female. Scape about 1.5x as long as gellum yellow. Face, clypeus, mouthparts, and wide. Flagellum with 24 segments in male and 26 lower half of temple, yellow. Thorax black with segments in female. First segment 1.32 to 1.35x as collar and in female. as the pronotal prosternum yellow long second segment. Face (Fig. 15) pol- of scutellum brown. Coxae and trochanters Apex ished, depressed laterally. Clypeus convex and Fore and middle yellow. legs otherwise yellowish polished. Mandibular teeth subequal, upper tooth brown. Hind tibia and tarsus femur, blackish slightly longer than the lower. Malar space 0.45x brown. Abdomen black with tergites 2 and 3 with the basal width of mandible. Second segment of marks. 4 with a faint triangular yellow Tergite maxillary palp slender and similar to other seg- mark in the yellow paratype examined. In one ments (Fig 40). Vertex polished, quadrate, wid- male studied only yellow apical lines are present ened behind eyes, 1.75x as wide as eye in dorsal on the abdominal tergites. view. Interocellar distance 0.22x the ocellocular Oval in Egg-— shape (Fig. 7, 10), with a distance. Temple in lateral view 1.7x to 1.75x as medioventral short stalk, its anchor four-pronged, wide as eye. 164 Journal of Hymenoptera Research

Thorax mostly smooth and shiny. H. & M. Townes. South Carolina: Cleveland, 3 Mesoscutum convex, pubescent. Notauli deep females, V-VI.1971, G. Townes. Alabama: Gulf and distinct to middle of mesoscutum. Scutellum Shores, 1 male, 22.IV.1968, H. & M. Townes. All flat, with lateral carina confined to base. Prono- these specimens are in the AEI collection at tum somewhat dull, epomia strong and extend- Gainesville, and those collected prior to 1955 are close to of ing upper edge pronotum. Epicnemium paratypes (see Mason, 1955: 67). Additional dull, epicnemial carina strong and extending to paratypes were described by Mason from Maine, 0.75 the height of mesopleurum. Sternaulus indi- Vermont, New Hampshire, Massachusetts, Con- cated on anterior 0.4 of mesopleurum. Propodeum necticut, and Quebec. convex, shiny. Areola narrow, parallel-sided, Distribution. —U.S. A., Canada. about 2.0x as long as wide (Fig. 3), confluent with basal area or only partly separated from it. 6 Orthomiscus eridolius Kasparyan Abdomen shiny. Tergite 1 2. Ox as wide at Figs. 9, 14, 24-25 as apex at base, about 2. Ox as long as its apical width, its dorsomedian carinae extending up to Orthomiscus eridolius Kasparyan, 1990. Fauna of USSR Insecta, F. Hymenoptera, 3(2):44. M, key, des., fig. Type: Female, about 0.8 its length (carinae weaker apicaUy, wid- Russia: Far East: Khabozavodsk Region (ST. ened, and a area between enclosing depressed PETERSBURG). Paratypes examined. them), its dorsolateral carina complete and strong This (Fig. 17). Tergite 2 with deep, oblique, basolateral species has been described in detail by grooves. Ovipositor thick, decurved (Fig. 18). Kasparyan (1990). Structurally it resembles O. sim- but Ovipositor sheaths short, broadly triangular. plex (Mason) shows differences in the curva- Color. —Black and yellow. LowerO. 66 of head, ture of the basal vein, shape of the areola, shape of scape, flagellum beyond 4th segment, pronotum, tergite 2, and in the egg structure. The diagnostic scutellum, metascutellum, presternum, mesoster- features are given— below. num, lower part of mesopleurum, fore and middle Diagnosis. Temple more buccate than in O. coxae and trochanters, sternites, ovipositor sheaths, simplex. Mandibular teeth equal. Epomia present, and subgenital plate, yellow. Legs otherwise and moderately strong. Areola slightly widened at its abdomen 1 with costulae and confluent with basal (except tergite ), yellowish brown. Terg- junction ite 1 black. Brown to blackish patches present on area. Apical transverse carina of propodeum at its with lateral abdominal tergites and pronotum. Wings hyaline. angulate junction and median carinae. Egg— (Figs. 4, 5, 18). Short, oval, ventrally flat longitudinal Petiolar area appearing pen- Basal and bearing on nearly half its length a short but tagonal. vein of fore wing only slightly curved near its wide stalk; egg anchor large and enclosing most of junction with the medius vein. Nervellus the egg. Anchor bearing two long posterior and intercepted at its lower 0.15-0.2. Tergite 2 without two long anterolateral arms that extend toward oblique grooves. Ovipositor short and thick (Fig. the dorsal sheaths surface— of the egg. 25). Ovipositor short, tapered. Egg (Figs. Length. 6.0-8.0 mm.; fore wing 5.0-6.5 mm.; 9, 14, 25) oval, with a ventral flattened anchor that ovipositor about 0.4 mm. encloses almost half of the egg. Specimens examined. —CANADA: Ontario: Coloration essentially similar to that of O. Cumberland, 1 female, ll.VI.1975,L. Ling. U.S.A.: simplex. Head and thorax in female yellow in lower Rhode Island: Westerly, 1 male, 2 females, 9- half and black in upper half. Mesosternum of 1 1 .VI. female black or brownish-black. 1936, M. Chapman. New York: McLean Bogs, Face of male partly 1 Abdomen 1 Tompkins Co., female, 30. VI. 1939, J. G. yellow. beyond tergite brownish black, Franclemont. New Jersey: Tabernacle, 1 male, with yellow spots. Mesosternum black in female. 11.VI.1939, H. K. Townes; Millwood, 1 female, Face of male largely black. 21.VI.1936, H. K. Townes. Maryland: Takoma Park, Length. —4.5-6.0 mm.; fore wing 3.8-4.8 mm.; 1 H. & M. about 0.4 mm. male, 2.VII.1944, Townes. Pennsylvania: ovipositor — Trout Run, 1 female, 29.V.1938, H. K. Townes; Distribution. Russia: Far Eastern Region. Spring Brook, 1 female, 27.VI.1945, H. K. Townes; Michigan: Huron Mrs., 2 males, 2 females, VI. 1961, Orthomiscus amurensis Kasparyan, 1986 is hereby H. & M. Townes; Iron River, 1 female, 23. VI. 1969, transferred to Kristotomus. Volume 3, 1994 165

Kristotomus amurensis (Kasparyan), of the American Entomological Institute 25 (6): 1-88. new combination Hellen, W. 1951. Neue Ichneumoniden aus Ostfennoskandien. I. Notulae Entomologicae 31: 25-31. Figs. 13, 32-35 Holmgren, A. E. 1858. Forsok till upstallning och beskrifning av de i Sverige fauna Tryphonider [Monographia Orthomiscus amurensis Kasparyan, 1986: 55. Kasparyan, 1990. Tryphonidum Sueciae]. Svenska Vetensk.-Akad. of 3 40. Fauna USSR Insecta, Hymenoptera, (2): key. Handlirsch (N.F.) 1: 93-246, 305-394. des., Female, Russia: Far East: Khabozavodsk fig. Type: Jacob, J. C. and J. Tosquinet. 1890. Catalogue des Examined. Region (ST. PETERSBURG). Ichneumonides de la Belgique appartenant au groupe des Tryphonides. Annales de la Societe Entomologique The holotype of this species was examined in Beige 34: 44-135. Jussila, R. 1975. Ichneumonological (Hym.) from 1992. The ovipositor and its sheaths are tapered reports Finland III. Annales Entomologici Fennici 41: 49-55. and slender and both extend beyond the apex of Kasparyan, D. R. 1976. [New species of ichneumonids of the abdomen. The lower valve of the is ovipositor tribe Cteniscini (Hymenoptera, Ichneumonidae) from without teeth. The egg has a simple stalk and eastern Asia. The genera Cycasis Townes, Orthomiscus anchor. The interocellar distance is about 0.8x the Mason and Kristotomus Mason.] Entomologicheskoe Obozrenie 55(1): 137-150. translation in Entomo- ocellocular distance. The malar space is 0.5x the [English logical Review 55 (1): 99-108.] basal width of mandible. The temple is moder- Kasparyan, D. R. 1977. [Ichneumonids of the subfamilies swollen, and is about as wide as the ately eye Pimphnae and Tryphoninae (Hymenoptera, Ichneu- width. The dorsolateral carina of tergite 1 is inter- monidae) new for Mongolia and Transbaikalia.] 5: 456-470. rupted just above the spiracle. Nasekomye Mongolica [Insects of Mongolia] D. R. 1986. [Two new of the tribe Exentenni The above mentioned characters and the gen- Kasparyan, species (Hymenoptera, Ichneumonidae) of the Far East] In: eral body form would place this species in Ler, P. A. (Editor) [Systematics and ecology of insects from Kristotomus, to which this is genus species hereby the Far East.] Academiya Nauk SSSR, Vladivostock, 155 It transferred. appears related to Kristotomus pages. (Pp. 54-57). bitccatus Kasparyan. Kasparyan, D. R. 1990. Fauna of USSR Insecta, Hymenoptera Vol. Ill, No. 2. Ichneumonidae, Subfamily Tryphoninae: ACKNOWLEDGMENTS Tribe Exentenni, Subfamily Adelognathinae. Leningrad: Nauka Publishing House. 340 pages. of Kerrich, G. J. 1942. Second review literature concerning thanks are to Dr. D. R. Peters- My due Kasparyan (St. British Ichneumonidae (Hym.), with notes on palaearctic burg) for the loan of the type specimens and for his com- species. Transactions of the Society for British Entomology ments. The collections of the American Entomological Insti- 8 (2): 43-77. tute available for for I a revision in (AEI) were readily my studies, which Kerrich, G. J. 1952. A review, and greater part, of am thankful to the President of the Institute. The draft of the the Cteniscini of the Old World. Bulletin of the British paper was kindly read by and commented upon by Drs. John Museum (Natural History) Entomology Series 2 (6): 305- Barron and Charles Porter and their comments have been 460. incorporated. I am also thankful to Mr. Andrei Sourakov for Mason, W. R. M. 1955. A revision of the Nearctic Cteniscini

I. Kerrich translating Kasparyan's key to the Russian species of the (Hymenoptera: Ichneumonidae) Eudiaborus genus for my use. Subsequently I received a similar transla- and a new genus. Canadian Journal of Zoology 18-73. tion from Dr. Kasparyan. The use of published figures from Mason, W. R. M. 1962. Some new Asiatic species of Exenterini the works of Henry Townes, D. R. Kasparyan and W. R. M. (Hymenoptera: Ichneumonidae) with remarks on ge- Mason is duly acknowledged and the source is cited under neric limits. Canadian Entomologist 94: 1273-1296. figure legends. Mason, W.R.M. 1966. A primitive new species of Kristotomus Mason (Hymenoptera: Ichneumonidae). Canadian Ento- LITERATURE CITED mologist 98: 46-49. Roman, A. 1914. Beitrage zur schwedischen Ichneu- monidenfauna. Arkivfor Zoologie 9 (2): 1-40. Carlson, R. W. 1979. Family Ichneumonidae. In: Krombein, Townes, H. 1969. The genera of the Ichneumonidae Part I. Smith & others, Catalog of Hymenoptera in America North Memoirs of the American Entomological Institute 11: 1-300. of Mexico 1: 315-741 (Ichneumonidae pages). Washing- Townes, H., S. Momoi and M. Townes. 1965. A Catalog and ton, DC. reclassification of the Eastern Palearctic Ichneumonidae. Gupta, Virendra K. 1990. The taxonomy of the Kristotomus- Memoirs of the American Entomological Institute 5:1-671. Complex of genera and a revision of Kristotomus (Hy- menoptera: Ichneumonidae: Tryphoninae). Contributions 166 Journal of Hymenoptera Research

lateral view of ventral Figs. 1-14. 1-5. Orthomiscus platyura Mason: 1, habitus, 2, head; 3, propodeum and tergite 1; 4, egg; 5, medusae Hellen; 9, view of egg. Figs. 6-9. Lateral view of egg of: 6, O unicinctus (Holmgren); 7, O Kasparyan; 8, O pectoralis and sheath of: O. O. eridolius Kasparyan. Figs. 10-14. Apex of abdomen showing subgemtal plate, ovipositor ovipositor 10, O. eridolius. 1-3 after medusae; 11, O. unicinctus; 12, 0. pectoralis; 13, 0. [=Kristotomus] amurensis (Kasparyan); 14, [Figs. Townes, 1969; 4-6 after Mason, 1955; 7-14 after Kasparyan, 1990.] Volume 3, 1 994 167

17

15-18. O. face Figs. platyura: 15, 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

26 31. 26-30. Orthomiscus face and of Figs. : simplex (Mason): 26, vertex; 27, mandible; 28, tergite 1; 29, tip abdomen showing and ovipositor 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

36-38. Fore the curvature of the basal vein in: Figs. wing showing 36, Orthomiscus unicinctus; 37, O. platyura; 38, O. medusae. Volume 3. 1 994 173

39-42. Figs. 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; V

INTRODUCTION tures, supporting its monophyly. Pediaspidini in- cludes only one genus of palaearctic distribution: The Cynipidae or gall wasps is one of the Pediaspis, Tischbein gall maker on Acer (Aceraceae). better known families of Cynipoidea. It is esti- Some authors (Weld 1952, Weidner 1968, Askew mated that there are around 1,800 species distrib- 1984) include this genus within the Aylacini but uted mainly in temperate areas of the northern since it exhibits a specialized scutellar morphol- hemisphere (Dalla Torre and Kieffer 1910, ogy and has a heterogenic life-cycle, it is better Nordlander 1984). More than 250 species have placed in a separate tribe. Rhoditini is composed been recorded in Europe (Nieves-Aldrey 1987). of two genera: Diplolepis Fourcroy and Liebelia Galls of this of are Kieffer group parasitic wasps counted (Ronquist, "in press") which induce galls the among most specialized and morphologically on Rosa (Rosaceae). Cynipini comprises of oak gall complex of all zoocecidia. About 75% of all known inducers and they are associated exclusively with are cynipids gall inducers associated with , mainly Qitercus. The inquiline group is almost with an- Fagaceae, exclusively Querais; composed of genera which do not produce galls, other is associated large group with Rosaceae and living as inquilines in the galls of other Cynipidae. Asteraceae, while a few species are linked to Some authors, especially in North America, have Lamiaceae, Papaveraceae, Valerianaceae, given subfamily status to this group (Ashmead Aceraceae, Fabaceae, and Apiaceae. 1903, Burks 1979); more frequently it has been With to the classification of regard higher the included together with the gall inducers in a large Cynipidae, the family is divided here into six group without subdivisions (Weld 1952, Eady and tribes, on the basis of the five tribes proposed by Quinlan 1963). Here it is regarded as a tribe. Ashmead see also (1903); Kinsey (1920). All the Recent studies on phylogenetic relationships tribes but viz one, Eschatocerini, are present in within Cynipidae (Ritchie 1984, Ronquist in press) west-Europe. Eschatocerini is only represented by support the monophyly of as well as the Neotropical Eschatocerus Mayr, gall maker on that of Cynipini and Rhoditini. On the other hand, species of Acacia and Prosopis. This genus presents Aylacini appears to be based on symplesiomorphy an axilla large, triangular and situated in the same and would be an unnatural paraphyletic group. plane as scutellum; scutellar foveae and notauli Nevertheless, for historical reasons and consider- absent reduced and wing venation; all of them ing biological characters, I prefer to treat them very distinctive apomorphic morphological fea- here as a tribe . 176 Journal of Hymenoptera Research

KEY TO THE TRIBES OF CYNIPIDAE

1 Pronotum dorsally longer, in median dorsal line at least one-sixth, usually one-third as long as greatest length on outer lateral margin, and frequently with truncation and pits forming pronotal plate (Figs. 1, 2); projection of ventral spine of gaster of female always short. Species usually fully alate, except male of Plwnacis centaureae and some forms of Synergus ttmumacerus, which are brachypterous 2 — Pronotum dorsally very short, in medial dorsal line one-seventh or less as long as outer lateral margin, and without truncation or pits (Fig. 3); projection of ventral spine of gaster of female usually longer than broad, viewed laterally. Agamic forms of some species apterous 4 2 Scutellum dorsally flat and trapezoidal with a large, heart-shaped impression on the disk; scutellar foveae absent (Fig. 4). One species with heterogonic life cycle, associated with Aceraceae (Acer spp) Pediaspidini — Scutellum oval or rounded, always dorsally convex; scutellar foveae usually present, sometimes reduced to an arched depression. Inquilines or gall makers associated with plants other than Acer 3 sulci 3 Gula (Fig. 5) reduced to a long, narrow median strip; gular united well before reaching hypostomata; metasomal tergites 2+3 (third and fourth abdominal terga )(Fig. 8) fused, at least in the female, into one large segment occupying nearly the whole of the gaster in the female; if showing only a false suture in either sex (Fig. 6), then the face usually has two vertical and parallel carinae (Fig. 7); face and thorax sculptured, sometimes very lightly, but never smooth and shining; Maxillary palps with 5 palpomeres 1 in and Rosa — with and 2 fused. Inquilines galls of Quercus Synergini Gular sulci (Fig. 9) usually free or meeting near the hypostomata; metasomal tergites 2-7 (Fig. 10) free in most cases; if metasomal tergites 2+3 fused in the female into one large segment then head and thorax almost smooth and shining; pits of pronotum usually present and sometimes associated with a light pronotal plate; maxillary palps 5-segmented. Gall makers on plants other than Quercus, Rosa, and Acer Aylacini 4 Axilla situated in the as scutellar foveae 1 1 notauli large, triangular and same plane scutellum; (Fig. ) absent; absent; wing venation (Fig. 12) reduced; Rs+M and Rl not visible; radial cell of forewing with radius quite separate from anterior margin of wing; frons (Fig. 13) with a strong medial keel. One genus from South America, gall maker on Acacia Eschatocerini — Axilla much smaller; not situated at the same level as the scutellum; scutellar foveae present or forming a shallow arcuate transverse depression; (Rs+M) of fore wing present and Rs reaching or almost reaching to anterior margin of wing; frons without medial keel or with a much weaker keel 5 5 Mesopleuron (Fig. 14) with longitudinal furrow; hypopygium of female (Fig. 15) ploughshare-shaped; radial cell of wing usually closed anteriorly. Gall makers on Rosa Rhoditini — Mesopleuron without longitudinal furrow; hypopygium of female never shaped as above; radial cell open anteriorly. Gall makers on Quercus, Castanea and Lithocarpus Cynipini

The tribe Aylacini comprises a group of primi- (1963), Quinlan (1968) and Kovalev (1982), but as five genera of cynipid gall-inducers associated far as I know no entire revision of the group has mainly with herbaceous plants. Most of their rep- been accomplished. resentatives are linked with species of Asteraceae The group appears to be particularly abun- producing milky juice, but some genera and spe- dant and species-rich in the far east of Europe cies are associated with Papaveraceae, Lamiaceae, where many taxa have been described recently Valerianaceae and Apiaceae. Galls of the species (Zerova et al. 1988) (see also Belizin 1959). Unfor- of this tribe are not are I have failed to obtain material of complex ; they produced tunately type mainly in twigs or fruits and are not detachable these taxa to be included in this revision, from the host plant tissues. The life cycles are uni- or bivoltine but without heterogony or alternation MATERIAL AND METHODS of generations (Folliot 1964, Askew 1984). There is no previous full revision of the genera of this tribe. This study is supported by the examination of SomepartialrevisionofgeneraincludedinAylaciru the types of almost all the genera represented in were made by Weld (1952), Eady and Quinlan western Europe. The initial purpose was to in- Volume 3, 1994 177 elude all the genera described from the western AYLACINI (Ashmead 1903) Palaearctic region but since I failed to borrow the Aulacini Ashmead 1903:147 type material of some genera from eastern Europe, Quinlan 1968:275 mainly Russia, Ukraine and Rumania, the work Aylaxini (Ashmead): Aylacinae (Ashmead): Kovalev 1982:85 was finally restricted to west-Europe. It was also Aulacideini (=Aylaxini) Fergusson 1988:143 in order to the taxonomic status necessary, clarify Type genus: Aylax Hartig 1840 (=/WflxHartig, 1843 (unjusti- of to and relationships some genera, examine some fied emendation)), orig. desig. Nearctic material. These were: Antistrophus, Gillettea and the type species oiAulacidea Ashmead. Insects of relatively small size; 1-5 mm. Col- Type species of genera of Aylacini are cited ac- oration usually black, brown or red-brown. Head cording to Rohwer and Fagan (1917). The list of viewed dorsally transverse; usually two times collections from which material have been exam- broader than long; in frontal view usually slightly ined for this study is as follows: broader than high or as broad as high; temples not behind the ocelli small; face with MCNM Museo Nacional de Ciencias Naturales, Madrid, expanded eyes; striae from to antennal sockets Spain radiating clypeus MNHP Museum National d'Histoire Naturelle, Pans, weak to moderately developed, medially usually France (Mme Casevitz-Weulersse) absent; subocular sulcus absent. Frons and vertex MZLU Museum of Zoology and Entomology Lund Uni- usually with dull sculpture, coriaceous, alutaceous versity, Lund, Sweden (R. Danielsson) or reticulate; rarely smooth and shining; frontal NHML The Natural History Museum (British Museum), carina not area raised or London, England (N. Fergusson) present; supraclypeal NHMW Naturhistorisches Museum, Vienna, Austria (M. slightly protuberant; antenna filiform with 12 to Fischer) 14 antennomeres in females, 13 to 15 in males. USNM National Museum of Natural Smithsonian History, Pronotum dorsally long; in median dorsal line Institution, Washington,DC, USA (A. Menke) one-sixth at least, usually one-third as long as ZMHB Zoologisches Museum, Humboldt-Universitat, on outer lateral Berlin, Germany (F. Koch) greatest length margin, frequently ZMH Zoologisches Museum Hamburg, Hamburg, Ger- with truncation and submedial pits forming a many (R Abraham) light pronotal plate; sides of pronotum usually ZSBS Zoologische Sammlungen des Staates, Bayerischen pubescent; mesoscutum usually with coriaceous, Munich, Germany (E. Diller) alutaceous or reticulate sculpture, less frequently MORPHOLOGICAL TERMS transversally rugulose or almost smooth; notauli complete to almost absent; median mesoscutal The terminology mostly follows that used by impression, if present, usually impressed only in Eady and Quinlan (1963), Ritchie (1984) and posterior third or less of mesoscutum; scutellar Ronquist and Nordlander (1989). The term "tho- foveae usually present, sometimes confluent and rax" as used here includes the propodeum and not well defined posteriorly; mesopleuron usually thus is an equivalent to the "mesosoma" of the longitudinally striated, reticulate or rugose-reticu- mainly American literature. Measurements and late. Wings well developed except the brachypter- abbreviations used here are as follows: A3 and A4, ous males of Phanacis centaureae Forster; radial cell third and fourth antennomeres; POL (post-ocellar of forewing open or closed in the margin. Abdomi- is the distance distance) between the inner mar- nal tergites 3 to 7 free; third abdominal tergum of the gins posterior ocelli; OOL (ocell-ocular dis- covering 1/3 to 1/8 of the whole length of gaster, is the distance from the of tance) outer edge a laterally with or without a pubescent patch; terg- posterior ocellus to the inner margin of the com- ites with or without punctures. Ventral spine of line is the shortest pound eye; transfacial distance hypopygium very short. Tarsal claws simple, rarely across the face between the antennal sockets. Ab- with acute basal lobe or tooth. breviations of forewing venation are used as Eady and Quinlan (1963). Supraclypeal area is the me- dial area between the clypeus and the antennal sockets. Length and width of the radial cell of the forewing are measured following Wiebes-Rijks (1979). 178 Journal of Hymenoptera Research

KEY TO WEST-PALAEARCTIC GENERA OF AYLACINI (INCLUDING NEARCTIC GENUS ANTISTROPHUS WALSH)

and vertex and 1 All claws with distinct basal lobe or tooth (Figs. 16, 18). Mesopleuron, mesoscutum smooth shining. Associated exclusively with Rubus and Potentilla (Rosaceae) 2 rather striate or — All claws simple (Figs. 19, 20). Mesoscutum and vertex dull, sculptured, reticulate; mesopleuron striate or reticulate. Genera not associated with Rosaceae 3 to anterior of radial cell 2 Claws weakly lobed (Figs. 17, 18). Subcosta and radius reaching margin wing; sometimes partly closed anteriorly (Fig. 21). Third and fourth abdominal terga completely fused in females, separated in males. Gall-maker on Potentilla Xestophanes Foerster — Claws strongly lobed (Fig. 16). Subcosta and radius not reaching to anterior margin of wing (Fig. 22); radial cell open anteriorly. Third and fourth abdominal terga completely free in both sexes. One genus associated with Rubus and Potentilla Hartig 3 Mesopleuron longitudinally striated. Pronotum always with two distinct submedial pits 7 or — Mesopleuron reticulate (Figs. 24, 28, 41), rugulose-reticulate (Fig. 116) rugulose-striate. Pronotum with or without submedial pits 4 4 Rl and Rs not quite reaching to anterior margin of wing (Fig. 42); wing fringe absent (Fig. 42). Third antennal segment shorter than fourth (Fig. 37) Antistrophus Walsh — Rl and Rs reaching to anterior margin of wing and hair fringe distinct, usually long. Third antennal segment as long as or longer than fourth 5 5 Pronotum (Fig. 23) with two distinct submedial pits. Pronotum, mesoscutum and mesopleuron strongly striae face reticulate (Fig. 24). Radial cell (Fig. 25) clearly open anteriorly. Radiating on absent (Fig. 26). Third antennal segment as long as fourth (Fig. 27). Gall-maker on Papaver somniferum L — Iraella gen. n. Pronotum usually without submedial pits, rarely with two rather small and indistinct pits. Mesopleuron reticulate or ruguloso-reticulate. Radial cell usually closed, at least partially. Radiating striae on face present, sometimes weak or faint. Third antennal segment longer than fourth 6 6 Mesopleuron reticulate (Fig. 28). Notauli usually faint or absent anteriorly. Gall inducer on Asteraceae.. Phanacis Foerster — or Notauli Gall inducer on Asteraceae Mesopleuron rugulose-reticulate ( Fig. 116) rugulose-striated. complete. Timaspis Mayr Head, in frontal view, as high as broad, slightly higher than broad in male (Fig. 29); malar space at least as long as the height of an eye in the female, longer in male. Scutellar foveae shallow and indistinct. Hair fringe on apical margin of forewing long (Fig. 30). Gall-maker on Valenanella (Valerianaceae) — Cecconia Kieffer Head in frontal view broader than high (Figs. 44, 57); malar space usually shorter than height of an eye. Scutellar foveae distinct. Wing fringe on apical margin, if present, usually not as long 8 8 Rl and Rs of forewing not quite reaching to anterior margin of wing; radial cell clearly open at base and along margin (Figs. 62, 79, 110) 9 — Rs almost reaching to anterior margin of wing and Rl continuing more or less along margin of wing (sometimes Rl only touching the margin); radial cell closed, partially closed or obsoletely closed {Figs. 34,50,55, 103) 12 9 Pronotum shorter medially (Fig. 60). Mesopleuron (Fig. 62) striate with some reticulation between striae. Third abdominal tergum without lateral pubescent patch. Gall maker on Papaver (Papaveraceae) — Barbotinia gen. n. Pronotum longer medially (Figs. 75, 107). Mesopleuron (Fig. 77) finely or strongly striate. Third abdominal tergum usually with lateral pubescent patch 10 10 Pronotum longitudinally striate and strongly pubescent (Fig. 31). Mesoscutum finely reticulate (Fig. 32). Third female antennal segment longer than fourth (Fig. 33). Third abdominal tergum with a distinct lateral pubescent patch (Fig. 31). Hair fringe on margin of forewing moderately long. Associated with — Glechoma (=Nepeta) (Papaveraceae) Liposthenus Forster Pronotum without longitudinal striae (Fig. 77), coriaceous or reticulate, only moderately pubescent. Mesoscutum coriaceous or rugulose-coriaceous. Third female antennal segment shorter than fourth (Fig. 73). Third abdominal tergum with or without lateral pubescent patch. Hair fringe on margin of forewing absent or very short 11 Volume 3, 1994 179

11 Hair fringe on wing margin completely absent (Fig. 110). Mesoscutum coriaceous, with scattered piliferous punctures (Fig. 108); notauli distinct only in posterior half of mesoscutum. Gaster laterally smooth, without Gall inducer on Phlomis Rhodus — punctures (Fig. 109). (Lamiaceae) Quinlan Hair fringe on wing margin very short or absent. Mesoscutum without punctures (Fig. 76). Notauli almost complete. Fourth to seventh abdominal tergites finely punctate (Fig. 78). Associated with Centaurea (Asteraceae) Isocolus Foerster 12 Rl of forewing continuing along margin of wing almost completely closing radial cell; sometimes vein is very lightly pigmented on the margin and radial cell appears obsoletely closed (Figs. 50, 55, 69). — Pronotum always dorsally quite long 13 Rl usually not continuing along margin of wing, occasionally slightly so; radial cell appearing open, at least anteriorly (Figs. 34, 103) 14 13 Mesopleuron clearly striate (Figs. 49, 54). Third abdominal tergum with lateral pubescent patch. Antenna of female 13-segmented. Notauli usually complete. Hair fringe on margin of forewing moderately long. Associated with Asteraceae and Lamiaceae Aulacidea Ashmead — not Mesopleuron clearly striate (Fig. 68), with some striae on alutaceous sculpture. Third abdominal tergum without lateral pubescent patch (Fig. 70). Antenna of female 12-segmented (Fig. 65). Hair fringe on of margin forewing absent (Fig. 69). Gall inducer on Salvia (Lamiaceae) Hedickiana gen. n. 14 Notauli weak or obsolete even in posterior half of mesoscutum. Third abdominal tergum without lateral — pubescent patch. Gall maker on Phlomis (Lamiaceae) Panteliella Kieffer Notauli clearly impressed, at least in posterior half of mesoscutum. Third abdominal tergum with or without lateral pubescent patch 15 1 5 Pronotum short and more transverse 1 dorsally relatively pits narrowly separated, strongly (Fig. ). Scutellar foveae antenna smaller, sculptured. Female 13-segmented. Rl partially closing the radial cell (Fig. 34). Gall maker on Papaver (Papaveraceae) Aylax Hartig — Pronotum dorsally longer and pits rounder and more broadly separated (Fig. 100). Scutellar foveae larger, rounded and smooth (Fig. 101). Female antenna 12-13 segmented. Rl not partially closing the radial cell (Fig. 103). Associated with Salvia (Lamiaceae) Neaylax gen. n.

ANTISTROPHUS Walsh ing tergites not punctate. Comments. —The structure of the antenna, Antistrophus Walsh 1869:74. Type species: Antistroplius pronotum and radial cell relate it to the ligodesmiae-pisum Walsh. Monotypic (examined) European Isocolus, but it differs from Isocolus in the Asclepiadiphila Ashmead 1897b:263. Type species: genus reticulate of Asclepiadiplula steplianotidis Ashmead. Monotypic. strongly sculpture the mesopleuron — and mesoscutum. This feature relates it to Phanacis Diagnosis. Head, viewed dorsally two times but it differs from that genus in important traits broader than long; viewed frontally, slightly such as the presence of submedial pits in the broader than high; temples slightly expanded be- pronotum, scutellar foveae distinct, third anten- hind the eyes. Frons and vertex strongly reticulate; nal segment shorter than fourth, radial cell open, radiating striae present; supraclypeal area reticu- and hair fringe on distal margin of forewing ab- late, slightly protuberant; antenna 13-14 segmented sent. in in — female, 14-segmented male; A3 clearly shorter Distribution and Biology. Antistrophus, in spite than A4. Pronotum strongly reticulate, in dorsal of being cited in Weld (1952) as a genus confined view, long; submedial pits conspicuous. to North America, is mentioned in the generic Mesoscutum reticulate; notauli slightly faint ante- keys of the same work as also present in Greece, riorly; median mesoscutal impression weakly im- galling plants of Salvia (Lamiaceae). Later on, scutellar foveae pressed posteriorly; large, Quinlan (1968) reiterates, without questioning it, rounded, confluent; scutellum strongly reticulate; the record. But the truth is that there is no correct mesopleuron reticulate. Wings hyaline; Rl not distribution record of this species for that country, the of quite reaching margin wing; radial cell therefore Quinlan's citation is quite clearly errone- open, nearly three times as long as broad, without ous and the occurrence of the genus in Europe hair on distal fringe margin. Third abdominal quite unlikely. All data suggest that it is a genus of tergum without lateral pubescent patch; follow- exclusively Nearctic distribution. 180 Journal of Hymenoptera Research

The genus comprises nearly 10 species in North pressed in posterior fourth of mesoscutum; scutel- America, associated with species of Silphium and lar foveae large, confluent, rounded; scutellum Chrysotlwmnus (Asteraceae), in the stems of which rounded, strongly reticulated; mesopleuron (Fig. reticulate they produce galls. 41) reticulate; propodeum and laterally pubescent; propodeal carinae slightly divergent. REVISED SPECIES Wings hyaline with pale venation; Rl and Rs of forewing not quite extending to wing margin, radial cell Rs rather curved Antistrophus pisum Ashmead open (Fig. 42); ; length of the radial cell about 2.8 times the width; areolet Antistrophus pisum Ashmead 1885:294 indistinct; cubitalis (Rs + M) joining the basalis Walsh 1869:74 name). Antistrophus ligodesmm-pisum (invalid (Rs+M) at a point close to junction of basalis with Asclepiadiphila stephanotidis Ashmead 1897b:263 Cu^ fore-wing without distinct hair fringe on wing Examined material. —Walsh's type series of A. margins. Gaster, viewed laterally, slightly longer + pisum in the USNM consists of 34 syntypes: 11 6\ than head thorax; third abdominal tergum (Fig. 23 9 of them mounted on and smooth without or , part micro-pins part 41) basally sculpture pubes- cards. 4 with red labels not nor on 9 , printed "Type No cence, following segments punctate pu- 1523 U.S.N.M." and white handwritten bescent; ventral spine of hypopygium very short. "Antistrophus pisum Walsh"; the female in better MALE, differs from female in darker coloration, condition, only missing the last segments of the mainly rufo-brown; antenna longer (Fig. 38), with is 7 14 antennomeres. right antenna, designated lectotype; $ , glued on a single card, with labels number 208, red "Type No 1523 USNM"; white "Lygodesmia gall", AULACIDEA Ashmead are designated paralectotypes. The type series of

A ulacidea Ashmead 1 897a :68. Aulax A. stephanotidis Ashmead comprises 46" and 79; Type species: mulgedncola Ashmead Aulax Ashmead 1 set card with labels: white "4780 4-0- (= hamngtoni 1877) Original 9 , on point, designation 91" and red "Type No 3737 USNM", is designated — viewed about two lectotype. — Diagnosis. Head, dorsally, Redescription. FEMALE. (Figs. 35-42). Length times broader than long; viewed frontally slightly 3 mm. Yellow-red or amber with darkened parts broader than high; temples not expanded or slightly mainly in vertex, mesoscutum, scutellum and expanded behind eyes; Frons and vertex coria- upper part of gaster; antenna and legs concolorous ceous; radiating striae present laterally on face; with the body. Head scarcely pubescent and supraclypeal area raised, coriaceous; antenna 13- strongly reticulate, viewed dorsally (Fig. 35) two segmented in female, 14-segmented in male, A3 times broader than long, viewed frontally (Fig. 36) slightly shorter than A4. Pronotum long dorsally 1.1 times broader than high. POL slightly shorter and submedial pits present, broadly separated; than OOL and about three times greatest diameter mesoscutum coriaceous, granulate or finely rugu- of lateral ocellus; transfacial line around 2.4 times lose; notauli usually complete; median mesoscutal height of eye; temples moderately expanded be- impression usually present posteriorly; scutellar hind eyes; malar space 1.2 times height of eye; face foveae distinct, usually large and rounded; laterally with radiating striae; supraclypeal area mesopleuron longitudinally striated. Radial cell reticulate, only slightly prominent; antenna (Fig. of forewing entirely closed on margin; hair fringe 37) relatively long, slightly shorter than body, present, moderately long. Third abdominal ter- filiform, with 13-14 antennomeres; if only 13, then gum, in females, with lateral pubescent patch; top segment is very long; third segment clearly following tergites— usually punctate. shorter than fourth, slightly excavated in middle. Comments. Morphologically, the genus is Pronotum (Fig. 39) reticulate and pubescent; dor- well distinguished by its closed radial cell and sally long, with distinct submedial pits associated striate mesopleuron. The type species was de- to clearly visible pronotal plate; mesoscutum (Fig. scribed from North America and is redescribed 40) strongly reticulate; notauli faint in anterior below. — third; moderately convergent posteriorly; median Distribution and Biology. Genus of holarctic mesoscutal impression indistinct, weakly im- distribution. The genus is represented in West- Volume 3, 1 994 181

Europe by 12 species. Most of European species SPECIES FROM WESTERN EUROPE induce galls on stems and leafs of species of Asteraceae specially on Hieracium; one species The species of this genus recorded in Western are listed in the check list produce galls in achenes of Nepeta (Lamiaceae). Europe (see Appendix). The novelties are represented by some species Aulacidea harringtoni (Ashmead) described by Thomson which are considered ei- ther good species or represent new synonymies 1887:146 Aulax harringtoni Ashmead (Nieves-Aldrey, in press). The following are other Aulax Ashmead 1896:133 (examined) mulgediicola new combinations or redescriptions of poorly — known Type material. Series composed of 75 speci- species: set card 1 9 with mens, individually on points. , Aulacidea kerneri comb. n. labels: red printed "Type No 3092 USNM"; white (Wachtl) "Aulax is handwritten mulgediicola" , designated Aylax kerneri Wachtl 1891:277 lectotype. — 43-50. Redescription. FEMALE. Figs. Length: Examined material. — I have been not able to 2.4 mm. Reddish or darker on yellow orange, examine the type material of this species. Observa- antenna and Head, dorsum; legs entirely yellow. tions come from specimens reared from galls on in dorsal view a little less than two times (Fig. 43), Nepeta spp collected in Spain. broader than in frontal view 1.2 long; (Fig. 44) The following characters advise the transfer of times broader than POL: OOL as 10:9; OOL high; this species to the genus Aulacidea Ashmead: prono- more than three times diameter of slightly greatest tum relatively long dorsally (Fig. 52); 13-segmented lateral ocellus; trans facial line about 1 .4 times height antenna, with A3 slightly shorter than A4 (Fig. 51); of behind eye; temples slightly expanded eyes; mesoscutum with complete notauli (Fig. 53); lon- frons and vertex coriaceous; face striated finely laterally gitudinally mesopleuron (Fig. 54) ; closed with area incomplete radiating striae; supraclypeal radial cell, although Rl is slightly pigmented on not antenna filiform with 13 prominent; (Fig. 45) margin (Fig. 55). antennomeres, third segment shorter than fourth. coriaceous Pronotum (Fig. 47) and scarcely pubes- Aulacidea scorzonerae (Giraud) cent; dorsally long, with distinct submedial pits broadly separated; mesoscutum (Fig. 48) finely Aulax scorzonerae Giraud 1859:370 rugulose or granulate; notauli complete; median Aulacidea scorzonerae (Giraud): Kieffer 1902:96 meoscutal impression visible in posterior 1/3 of — mesoscutum; scutellar foveae ovate, well defined, Examined material. The type material, sentby the 66 set on micro- separated by a septum relatively broad posteri- MNHP, comprises specimens isolated or in Most are com- orly; scutellum weakly rugose; mesopleuron (Fig. pins pairs. specimens striated and in condition. All the 49) finely longitudinally. Wings hyaline ; plete good specimens a white label "Museum Paris, coll. Giraud" radial cell of forewing (Fig. 50) closed on the carry and some labels handwritten Giraud. margin; length of the radial cell about 2.4 times carry by 1_, set next to a is width; areolet indistinct; hair fringe on wing api- complete, male, hereby designated cal margin distinct. Gaster (Fig. 49), viewed later- lectotype. —FEMALE. 88-95. ally, as long as head + thorax; third abdominal Redescription. Figs. Length: 2 mm. Coloration from almost tergum basally with lateral pubescent patch, fol- variable, entirely ventral of red-brown with darker thorax, antenna or lowing segments finely punctate ; spine orange hypopygium very short. MALE. Differs from fe- dark-yellow and legs entirely red-yellow except coxae and base of femora darkened to dark-brown male in antenna (Fig. 46), with 14 antennomeres. or black with darker too. Head Comments. —The morphological features of the legs scarcely pu- bescent and viewed type species are congruent with those exhibited by finely coriaceous; dorsally two times broader than viewed f ron- the species represented in Europe. (Fig. 88) long; — 1 .2 timesbroader than POL:OOL Distribution and Biology. Nearctic. The spe- tally (Fig. 89) high. as OOL about 3.5 times the diameter cies makes galls in stems of Lactuca spp, mainly L. 10:7; greatest of lateral ocellus: villosa (Burks, 1979). transfacial line around 1.8 times 182 Journal of Hymenoptera Research height of eye; temples not expanded behind eyes; ally on face; supraclypeal area slightly prominent; face laterally with incomplete radiating striae; antenna of female with 13 antennomeres; A3 as supraclypeal area only slightly prominent; An- long as A4; antenna of male with 14 segments. tenna (Fig. 90) filiform, relatively short, about 0.6 Pronorum relatively short dorsally and submedial times length of the body; with 13 antennomeres, pits strongly transverse and slightly separated; A3 clearly longer than A4. Thorax scarcely pubes- mesoscutum coriaceous-granulate; notauli com- cent; pronorum (Fig. 91) dorsally long; submedial plete or faint anteriorly; median mesoscutal im- pits clearly visible, slightly transverse; pronotum pression present in posterior 1/2 or more of with coriaceous sculpture; mesoscutum (Fig. 92) mesoscutum; scutellar foveae small, slightly trans- finely coriaceous-alutaceous sculptured; notauli verse and confluent; mesopleuron longitudinally faint anteriorly, posteriorly not broad; median striated. Radial cell of forewing only partially or mesoscutal impression weakly impressed in pos- obsoletely closed on the margin; Rl reaching to terior 1/6 of mesoscutum; scutellar foveae shal- anterior margin of wing; hair fringe present, mod- low, shining and almost smooth, ovate, separated erately long. Third abdominal tergum of females by septum broader posteriorly; scutellum rounded with lateral pubescent patch; following tergites with coriaceous-alutaceous sculpture; mesopleu- without punctures.— ron (Fig. 93) with fine longitudinal striae; Comments. The genus was originally very propodeum with subparallel lateral carinae; me- heterogeneous. Later on, some species were set dian area scarcely pubescent, almost smooth. apart and transferred to such new genera as Forewings slightly longer than body, hyaline; ra- Aulacidea Ashmead and Isocolus Forster. My present dial cell (Fig. 94) closed on front margin, about 2.3 conception of the genus is very restricted and I times longer than broad; hair fringe relatively limit it to the species which are gall makers on seed short; areolet inconspicuous. Gaster as long as capsules of Papaver and maybe— also of Hypecoum. head+thorax; third abdominal tergum with a slight Distribution and Biology. The genus is repre- lateral pubescent patch (Fig. 93); following terg- sented in Europe by two species: A. papaveris (Perris ites without visible punctures; ventral spine of 1839) and A. minor Hartig 1840. Aylax hipecoi Trot- hypopygium quite short. Tarsal claws simple. ter 1912, described from Northern Africa (Tripoli), MALE. Differs from female in 14- segmented with galls in Hypecoum fruits, also found in Greece antenna A3 broadened and also to this (Fig. 95),— slightly apically. Algeria, could be assigned genus. Comments. The species had been very sel- However, the species has not been recorded again dom recorded and, consequently, was poorly and I am ignorant as to the location of the types, so known having been incompletely described. After I have not been able to study the species. examination, I confirm that it belongs to Aulacidea. It is a species which is closely related to A. BARBOTINIA gen. n. tragopogonis Thomson, from which it differs in Figs. 56-62 host plant as well as in its incomplete notauli, less — oramensisBarbotm pronounced scutellar foveae, relatively shorter Type species. Aylax 1964:152. Herein des- Derivation: Feminine Dedicated to the radial cell and invisible gastral punctuation. ignated. gender. — French cynipidologist Mr. Barbotin, discoverer of the Distribution and Biology. Only recorded from type species. Austria and Reared from in the Hungary. galls — stems of Scorzonera austriaca Willd and S. humilis L. Description. Head transverse; viewed dor- sally (Fig. 56) more than two times as broad as not behind AYLAX Hartig long; temples expanded eyes; viewed frontally (Fig. 57) trapezoid-shaped; transfacial line times of Ay/fl.v Hartig 1840(1839):186, 195. Typespecies: rhoeadis almost two height eye; malar space Bouche. Ashmead 1903 Desig. by (not seen) longer than height of eye; radiating striae on face Aulax Hartig 1843:412. Emendation. strong, complete, laterally reaching the antennal — face without subocular Diagnosis. Head, viewed dorsally, about two sockets, medially absent; mandibles frons and vertex times as broad as long; viewed frontally slightly sulcus; large; minutely antenna females with 14 broader than high; temples not or slightly ex- alutaceous; filiform, antennomeres male with 15 antennomeres panded behind eyes; radiating striae present later- (Fig. 58), Volume 3, 1994 183

(Fig. 59); pedicel as long as broad; A3 as long as A4. CECCONIA Kieffer Pronotum (Fig. 60) dorsally relatively short; in Cecconia Kieffer 1902:7, 93. Type species: Aulax valerianellae median dorsal line, not more than 1/4 as as long Thomson 1877. Monobasic and original designation greatest length on outer lateral margin; submedial (examined) We/iiid/n Ionescu and Roman 1962:551. pits clearly visible slightly transverse and moder- Type species: Weldiella aequalis Ionescu and Roman 1962. Monobasic and ately separated; mesoscutum (Fig. 61) granulate- origi- nal designation (not seen) puncruate but punctures very obsolete or indis- — tinct; notauli complete; median mesoscutal im- Diagnosis. Head slightly less than two times pression impressed in about the posterior half of broader than long viewed dorsally; in frontal view mesoscutum; scutellar foveae large, rounded, con- as broad as high or slightly higher than broad; tiguous, shining and almost smooth; scutellum malar space as long as height of eye; clypeus granulate-rugulose; mesopleuron (Fig. 62) irregu- subquadrate; radiating striae well impressed; larly longitudinally striated but some reticulation supraclypeal area slightly prominent; female an- between the striae. R not tenna with 13 A3 shorter than A4 or as present Wings hyaline; t segments; quite reaching the margin and radial cell of forew- long as A4; male antenna with 14 antennomeres. ing open (Fig. 62); distal wing margins with short Pronotum dorsally not very long with distinct hair fringe. Gaster as long as head+thorax; third submedial pits; mesoscutum with alutaceous abdominal of tergum covering about 1/3 the gaster, sculpture; notauli very lightly impressed, absent basally smooth without punctures or pubescence; in anterior half of mesoscutum; median mesoscutal ventral spine of hypopygium short. Legs with impression absent; scutellar foveae confluent, tarsal claws. simple rounded, not well defined posteriorly; scutellum Comments. — The proposed genus is closely dorsally convex, with reticulate sculpture; related to these two in Aylax Htg. genera having mesopleuron finely striated longitudinally. Wings common a pronotum which is relatively short hyaline; radial cell of forewing open on the mar- and submedial which are medially pits transverse gin; about three times longer than broad; wing and Barbotinia scarcely separated. However, dif- margins with long hair fringe. Third abdominal fers from in that Aylax the head has a trapezoidal tergum with a lateral pubescent patch; following shape with a relatively long malar space in rela- segments not punctate. tion to the of an the — height eye; conformation of the Comments. The genus is easily distinguished scutellar foveae, very large and rounded; the stri- morphologically: the head shape higher than broad ate-reticulated of the sculpture mesopleuron; the viewed frontally; presence of long hairs in the radial cell of the forewing completely open on the distal wing margin; the shape of the scutellar margin, with the Rl vein hardly reaching the wing foveae and notauli conformation. Weldiella was and the absence of a lateral margin pubescent erected by Ionescu and Roman (1962) for the spe- in the third abdominal patch, tergum. cies W. aequalis found in Rumania, also gall-maker Distribution and — Biology. The new genus com- on fruits of species of Vrt/t'n'one//rt.(Valerianaceae). a known Barbotinia prises single species: oraniensis On the basis of the original description (since I comb, (Barbotin 1964) n., gall maker on capsules of have not had the chance to examine the type Papaver species (Papaveraceae). Its known distri- species) I consider the differences used to differen- bution area includes the mediterranean region tiate that genus from Ceconia insufficient to sup- and Northern Africa (Barbotin 1964). In Spain it is port the description of a separate genus. At most, relatively frequent (Nieves-Aldrey 1984). In the the mentioned features could indicate the exist- there are six NHML specimens and galls of this ence of a second species, different from C. collected in species Italy, Moladi Baxi, 1974, iden- valerianellae, within the genus Cecconia. tified as — by Quinlan Aylax sp. Distribution and Biology. The genus com- prises two known species which produce galls in the fruits of of Valerianella species ( Valerianaceae). C. valerianellae (Thomson) is widely distributed in Europe from Sweden to Rumania, through the Iberian Peninsula; C. aequalis (Ionescu & Roman), is only recorded from Rumania. 184 Journal of Hymenoptera Research

DIASTROPHUS Hartig Europe, the genus comprises many species in North America, most of them also associated with spe- Diastrophus Hartig 1840 (1839): 186. Type species: Diastrophus cies of Rubus, the others associated with Potentilla rubi Hartig. Monotypic and Fragaria. Type material of D rubi Hartig, is housed in Munich (ZSBS)

It is of 1 and 5 9 . 9 on composed i Lectotype , mounted HEDICKIANA n. micro-pin, designated by Weld (1931) (examined). gen. — Figs. 63-70 Diagnosis. Head slightly more than two times Aulaadea levantma broader than long viewed dorsally; in frontal view Type species: Hedicke 1928:81. By present designation Derivation: Feminine Dedicated to slightly broader than height; malar space slightly gender the German entomologist Hedicke, discoverer of the shorter than height of eye; transfacial line 1.5 to 2 type species. times height of eye; supraclypeal area rugose- Examined material. —4 9 punctate with some striae; radiating striae incom- (paratypes) deposited in the NHML. The stated the author plete; female antenna with 13 segments, male an- holotype by to be in his has not been found tenna with 14 antennomeres; A3 slightly longer deposited collection, and is lost. than A4. Pronotum dorsally long with distinct probably Consequently, we designate a the studied submedial pits asociated with a distinct pronotal lectotype amongst paratypes: 1_, mounted on on card red plate; mesoscutum mostly smooth and shining; micro-pin point, label handwritten " notauli well impressed and complete; median "paratypus"; labels Aulacidea levantina 1928 Hedicke det" and "Tel mesoscutal impression variable as to its length; Aviv Pales- tine Bodenheimer". scutellar foveae deep, rounded, smooth and shin- — ing; scutellum with a longitudinal furrow; Description. FEMALE. Head transverse; viewed mesopleuron almost entirely smooth and shining, dorsally (Fig. 63) more than two times broader than medially with some fine longitudinal striae. Wings long; POL almost equal to OOL; behind weakly infuscate; radial cell of forewing open on temples slightly expanded eyes; viewed broader the margin; wing apical margin with hair fringe frontally (Fig. 64) clearly than high; transfacial line than 1.5 moderately long. Third abdominal tergum with- slightly more times height

of ; malar shorter than out lateral pubescent patch; following segments eye space height of eye; striae on face well punctate. Tarsal claws of the legs with conspicu- radiating impressed laterally; ous acute basal lobe or tooth. supraclypeal area slightly protuberant, without — frons and vertex with Comments. The genus is quite distinct mor- striae; coriaceous-alutaceous antenna 12 phologically and together with Xestophanes Forster sculpture; (Fig. 65) segmented, A3 more slender at the as 1869 and Gonaspis Ashmead 1897 it forms a sepa- slightly base; long as A4. rate group within Aylacini. In fact, morphologi- Pronotum (Fig. 66) dorsally relatively long submedial cally, they are distinguished from the rest of the pits clearly visible, broadly separated mesoscutum genera of this tribe because they exhibit claws (Fig. 67) coriaceous-punctate; notauli broad but in having a basal lobe or tooth; in the biological posteriorly impressed only posterior aspect, both genera are also well characterized by 3/4 of mesoscutum; median submedial impres- sion visible in 1 their association with species of Ritbus and Poten- posterior /4 of mesoscutum; scutel- lar foveae tilla (Rosaceae). Their special position within large, rounded, contiguous, shining and Aylacini has been pointed out by recent phylog- finely coriaceous; scutellum rugulose; eny studies of some genera of cynipids, mainly mesopleuron (Fig. 68) longitudinaly striated but some reticulation between the striae. inquilines, by Ritchie (1984) and Ronquist (in press). present Wings radial cell of The latter study indicates that Diastrophus is the slightly infuscated; forewing (Fig. 69) closed on front genus of gall-wasps most closely related to the margin; distal wing margins with hair inquilines, which form a monophyletic group. fringe quite short. Gaster as long as — third abdominal Distribution and Biology. In Europe head+thorax; tergum covering about 1 of Diastrophus is represented by only two species: D. /3 gaster; basally smooth without punc- tures or rubi Hartig, gall-inducer in species of Rubus and D. pubescence; following segments punc- ventral of mayri Reinhart, 1876, which forms galls in the tate; spine hypopygium short. Tarsal claws stems of Potentilla argentea L. In contrast with simple. Volume 3, 1 994 185 — Comments. The genus is erected for the single acters, like the general body shape, face and species: H. lez>antina (Hedicke), described within mesopleuron sculpture and shape of the scutellar Aulacidea Ashmead. In common with Aulacidea, foveae, approach the new genus to the Timaspis- the new genus exhibits a closed radial cell but it Phanacis complex, but the presence of submedian can be distinguished from Aulacidea as follows: pits in the pronotum is a differentiating trait; other antenna with 1 2 segments; mesopleuron not clearly features that distinguish them are: the entirely longitudinally striated, in part irregularly reticu- reticulate face with total absence of radiating striae; late; incomplete notauli and mesoscutum sculp- clypeus projecting anteriorly; third antennal seg- tured with piliferous punctures; base of third ab- ment being of almost the same length as fourth dominal tergum lacking lateral— pubescent patch. and radial cell of the forewing open on front Distribution and Biology. Eastern Mediterra- margin. nean; recorded from Crete, Syria and Israel The genus is erected for a single species: Iraella (Hedicke 1928). The only known species, Hedickiana luteipes (Thomson) comb. n. (= Timaspis papaveris levantina (Hedicke), induces conspicuous juicy galls Kieffer). Redescription of Aylax luteipes Thomson in stems of Salvia triloba L. (Lamiaceae). including statement of the synonymy indicated above was recently published (Nieves-Aldrey, in IRAELLA n. gen. press). — Figs. 23-27 Distribution and Biology. Genus distributed from Sweden to Spain and Iran. The only known Aulax Thomson 1877:807. Type species: luteipes (= Timaspis species of the genus is associated with Papaver papavens Kieffer in Goury et desig- Gignon). By present somniferum L. (Papaveraceae), on the stems of nation (examined) which it produces inconspicuous galls. I have had Derivation: Feminine gender. Dedicated to my wife Ira the chance to examine in the Natural History — Museum of London some from Iran Description. Head viewed dorsally scarcely specimens obtained from stems of less than two times broader than long; temples not Papaver pseudorientale and P. bracteatum which to expanded behind eyes; viewed frontally (Fig. 26) Lindley, might correspond a different within the same slightly broader than high; transfacial line slightly species genus. more than of striae height eye ; radiating on face absent; clypeus trapezoid-shaped, moderately ISOCOLUS Forster projecting anteriorly, supraclypeal area slightly Isocolus Forster 334 = Eubothrus prominent, reticulate; frons and vertex with re- 1869:330, ( Forster). Type spe- cies: Diastrophus scabiosae Giraud 1859. Monobasic and ticulate-coriaceous sculpture; female antenna, 13- original designation (examined) 14-segmented, male antenna 14-15-segmented; A3 — viewed almost of equal length to A4 (Fig. 27). Pronotum Diagnosis. Head, dorsally, about two times broader than at most (Fig. 23) dorsally relatively short; submedial pits long; temples only present; mesoscutum reticulate-coriaceous; notauli slightly expanded behind eyes; viewed frontally faint in anterior one third of mesoscutum; median 1.3 times broader than high; transfacial line around 1.5 times the of malar shorter mesoscutal impression visible in posterior 1/3 of height eye ; space mesoscutum; scutellar foveae transverse, than height of eye; radiating striae visible later- sometimes even confluent; scutellum reticulate; mesopleuron (Fig. ally, usually incomplete, present in the 24) reticulate. Wings hyaline; radial cell of forew- supraclypeal area; antenna of female with 1 3 shorter than ing (Fig. 25) open; three times longer than wide; antennomeres, A3 A4; antenna of male with 14 distal wing margins with hair fringe long. Third segments. Pronotum dorsally long and submedial abdominal tergum basally with lateral pubescent pits conspicuous, broadly sepa- mesoscutum with patch; following— segments without punctures. rated; coriaceous-rugulose Comments. The following morphological fea- sculpture, sometimes with transverse rugae; notauli tures define this new genus: face lacking visible usually complete; median mesoscutal im- visible at least in radiating striae; clypeus projecting anteriorly; pression usually posterior one third of pronotum relatively short dorsally but with vis- mesoscutum; scutellar foveae distinct, ible well scutellum submedian pits; mesopleuron entirely reticu- defined; rugulose; mesopleuron late; radial cell open on front margin. Some char- with conspicuous longitudinal striae; wings hya- 186 Journal of Hymenoptera Research

radial of line; cell forewing open; Rl and Rs not face laterally with radiating striae; supraclypeal quite reaching anterior margin of wing; distal area slightly raised, with some striae; antenna of hair short. margin wing with fringe very Third (Fig. 73) about 2/3 the length of the body, with 13 abdominal tergum usually without lateral pubes- antennomeres, A3 slightly shorter than A4. Prono- cent with patch, rarely pubescence; abdominal tum (Fig. 75) coriaceous, laterally pubescent; dor- sometimes from with distinct submedial tergites usually punctate, weakly, sally long, pits ; fourth ventral of short. tergite; spine hypopygium mesoscutum (Fig. 76) with coriaceous-rugulose Tarsal claws simple.— sculpture posteriorly with some conspicuous trans- Comments. The genus is defined mainly on verse rugae; notauli complete, well impressed the basis of wing characters as follows: Rl and Rs posteriorly; median mesoscutal impression dis- not quite reaching to anterior margin of wing, tinct in posterior half of mesoscutum; scutellar radial cell clearly open and wing fringe absent or foveae large, rounded; shining and almost smooth; very short. separated by broad posteriorly septum; scutellum Distribution and — is Biologi/. The distribution ovate, rugulose with the posterior half more or less Palaearctic: five are known in Western flat and species sloping backwards; mesopleuron (Fig. 77) Europe. From the biological point of view the longitudinally striated. Wings hyaline or very is associated genus with Asteraceae; the species lightly infuscated; Rl and Rs of forewing not quite inducing galls in stems, achenes or Lnvolucral extending to wing margin and radial cell open bracteae of of Centaurea species and Serrahda (Fig. 79); Rs slightly curved apically; length of the (Asteraceae). radial cell about 3 times the width; areolet indis- tinct; fore wing without distinct hair fringe on Of Some Redescription Species wing margins. Third abdominal tergum (Fig. 78) basally smooth without sculpture or pubescence; Isocolus scabiosae (Giraud) posterior third of third tergite and following seg- scabiosae Giraud 1859:368 Diastrophus ments conspicuously punctate; ventral spine of Diastrophus areolatus Giraud 1859:369. Syn. n. hypopygium very short; tarsal claws simple. Isocolus scabiosae (Giraud): Forster 1869:334 differs MALE, from female in antenna (Fig. 74) Isocolus rogenhoferi Wachtl 1880:542. Syn. n. with 14 A3 curved in the Aulax scabiosae (Giraud): Mayr 1882: 10 antennomeres; slightly Aylax scabiosae (Giraud): Dalla Torre and Kieffer 1910:671 middle. Isocolus scabiosae Rohwer and 1917:369 (Giraud): Fagan Comments. —The complete series from the Isocolus scabiosae (Giraud): Weld 1952:271 Giraud Collection, sent by the MNHP, comprises 63 individuals. In Material examined.—The type series of D. contrast with the specimens black described other in- scabiosae Giraud, in the MNHP, comprises 63 speci- predominantely above, dividuals of a chestnut-reddish mens: 30<5, 339, mounted on micro-pins, as well lighter, shade and as some galls on achenes of Centaurea scabiosa. 1 9 some completely reddish individuals are included set next to a in the series. The latter also exhibit other , male, with handwritten label "3 mai" morpho- differences such as a and "Museum Paris, coll. GIRAUD", is hereby logical stronger mesoscutal and less notauli designated lectotype. Paralectotypes, 18 6 and 12 9 sculpture impressed and medial mesoscutal impression. Also included in the series are some of transformed achenes Redescription.—FEMALE. (Figs. 71-79). Length galls consisting from 2.8-3.3 mm. Black; gaster red-brown to black; an- heads of Centaurea scabiosa L. and in the tenna and legs red brown; coxa, base of femora compliance with handwritten labels from and first two segments of antenna darker. Head Giraud: "caput centaureae scabiosae" and the date, at least of the of the series viewed dorsally (Fig. 71) 2.2 times broader than part specimens should have been obtained from these long; viewed frontally (Fig. 72) 1.3 times broader galls. Giraud origi- described the than high; temples only slightly expanded behind nally species Diastrophus scabiosae as obtained eyes; frons and vertex with sculpture coriaceous- from conspicuous galls in the stems of Centaurea scabiosa L. but these alutaceous; POL slightly shorter than OOL , OOL (Giraud 1859), galls are about four times greatest diameter of lateral ocel- not included in the series. The specimens ob- tained from in achenes lus; transfacial line around 1.6 times height of eye; galls could belong to Isocolus Wachtl. the status of the malar space slightly shorter than height of eye; rogenhoferi However, Volume 3, 1994 187

in latter has not been clear, since only differences area slightly raised, without striae; antenna (Fig. the coloration of the antennal flagellum together 82) with 13 antennomeres, A3 clearly shorter than with the different position of the galls have been A4. Pronotum (Fig. 83) reticulate-coriaceous, lat- pointed out in the literature (Kieffer 1901, Eady erally pubescent; dorsally long, with distinct and to it from /. scabiosae. submedial Quinlan 1963) separate pits ; mesoscutum (Fig. 84) with reticu- For these reasons, I consider that there is only one late-coriaceous sculpture; notauli complete, well species with a wide morphological variability be- impressed posteriorly; median mesoscutal impres- ing capable of inducing galls on both stems and sion distinct in posterior 1/5 of mesoscutum; scutel- flower heads of species of Centaurea, mainly C. lar foveae large, subtriangle-shaped; not smooth; scabiosa L. It is possible that the two types of galls separated by a posteriorly broader septum; scutel- correspond to two different generations of the lum reticulate-coriaceous, with longitudinal fur- In same species. the same work, Giraud described row variably impressed; mesopleuron (Fig. 85) another species, Diastrophus areolatus, from 2 fe- longitudinally striated. Wings hyaline, slightly males caught by netting near Vienna. The differ- longer than the body; Rl and Rs of forewing not ences between the latter and I. scabiosae were stated quite reaching to wing margin and radial cell open to of be reddish coloration the antennal flagellum (Fig. 86); length of the radial cell about 2.6 times and presence of areolet in the fore-wings. In the the width; areolet indistinct; fore wing without examined series, I found 2 specimens which could distinct hair fringe on wing margins. Third ab- to D. areolatus labeled correspond "captur 13 juill, dominal tergum (Fig. 85) basally with a lateral

Turksch" . The presence or absence of the areola, as pubescent patch; posterior third of third tergite well the general coloration, is quite variable in the and following segments conspicuously punctate; whole of the series, so I consider D. areolatus as ventral spine of hypopygium very short; tarsal conspecific with J. scabiosae and, therefore, I estab- claws simple. MALE. Differs from female in an- lish the corresponding new synonymy. tenna with 14 antennomeres (Fig. 87), A3 not modified only —slightly curved. Isocolus serratulae Mayr comb. n. Comments. This species was described by Mayr from material collected by Rogenhofer on Aulax serratulae 1882:9 Mayr Serratula heterophylla auct ross non (L.) Desf. The Aylax serratulae (Mayr): Kieffer 1901:313 gall is probably produced on this plant, but it has not been described. Since the description, the spe- material. — Series of 31 Type composed speci- cies has been recorded once from Ukrania, includ- mens: 126", 19 9, most of them set on micro-pins ing galls Zerova et al. (1988), so it is only known 1 (NHMW). 9 , with handwritten labels "serratulae from Austria and that country. The features of this ydetRogenhofer" and "Luxenburg,Rogenh 1881", species fit well in the genus Isocolus, except for the hereby designated lectotype. Other specimens third abdominal tergum with basal pubescent carry handwritten labels "Aid. serratulae det. G. patch, so I transfer it to this genus. Mayr" and "Serratula heterophylla"; designated paralectotypes. Isocolus lichtensteini (Mayr) comb. n. Redescription.—(Figs. 80-87). FEMALE. Length 2.5 mm. Black; red-brown to dark-brown; gaster Aulax lichtensteini Mayr 1882:7 antenna and base of femora legs light brown; coxa, Aylax lichtensteini (Mayr): Kieffer 1901:297 and two first segments of antenna darker. Head Isocolus lavaresi Nieves-Aldrey 1984:243. Syn. n. viewed dorsally (Fig. 80) about 2 times broader than material. — series of 2 9 long; viewed frontally (Fig. 81) 1.2 times Type Type composed ; broader than high; not behind 19, mounted on micro-pin, with handwritten la- temples expanded " bels Aulax det. G. the eyes; frons and vertex with sculpture minutely lichtensteini, Mayr" and "Cen- " coriaceous-reticulate; POL:OOLas 11:9; OOL about taur, gall lich designated lectotype. Paralectotype, female three times the greatest diameter of lateral ocellus one with label "Centaur, salmant. gall Lich". Comments. —This transfacial line around 1 .5 times the height of eye species was described from malar material collected in space around 0.6 times the height of eye, Montpellier (France), pro- in the stems of Centaurea L. face laterally with radiating striae; supraclypeal ducing galls salmantica 188 Journal of Hymenoptera Research

(=Microlonchus salmanticus) (Mayr, 1882). Isocolus presence of wing fringe on distal margin; third tavaresi Nieves-Aldrey, described from material antennal segment being longer than fourth; third reared from galls on Centaurea aspera L. in Spain abdominal tergum having a lateral pubescent patch (Nieves-Aldrey, 1984), later also found on Centau- and in general sculpture and—biology. rea melitensis L., (Nieves-Aldrey, unpublished), is Distribution and Biology. The genus com- conspecific with the two type specimens studied prises a single known species from Western Eu- and, therefore, I establish the corresponding syn- rope which causes the formation of galls in the onymy. runners, stems and leaves of Nepeta (=Glechoma) The type specimens differ from individuals hederacea L. (Lamiaceae). collected in Spain (Nieves-Aldrey 1984) only in their lighter chestnut-reddish coloration, also ex- NEAYLAX gen. n. tended for antenna and legs, including the coxae. salviae Therefore, the types exhibit only a slight colour Type species: Aylax (Giraud 1859):369. By present contrast between the two first antennal segments designation, (examined) Derivation: from Neos and Aylax (new Aylax) and the flagellum, and the longitudinal furrow of scutellum is more pronounced compared to the — Description. Head, viewed dorsally, two Spanish specimens collected. times broader than long; temples not expanded behind the striae on face well im- LIPOSTHENES Forster eyes; radiating pressed laterally; supraclypeal area raised, with- out striae; frons and vertex with coriaceous or Liposthenus (!) Forster 1869:332. alutaceous antenna 12-13- Liposthenes Forster 1869:336. Type species: Aulax glechomae sculpture; segmented in in Hartig. Monotvpic and orig. desig. (examined). The females; 14-15-segmented males; A3 as long of Aulax in the consist of 4 types glechomae Hartig, ZSM, as A4 or shorter than A4. Pronotum dorsally long; 9 on of label of , mounted One them carry lectotype pins. submedial pits clearly visible, not broadly sepa- ? designated by Weld in 1931 (Weld 1952). — rated; mesoscutum coriaceous or alutaceous, some- Diagnosis. Head, viewed dorsally, slightly times punctate; notauli complete or faint anteri- less than two times broader than long; in frontal orly; scutellar foveae large, rounded; mesopleuron view slightly broader than high; transfacial line longitudinaly striated. Wings hyaline; radial cell 1.5 times height of eye; malar space shorter than of forewing open on front margin but Rj reaching height of eye; frons and vertex finely reticulate; margin; distal wing margins with hair fringe mod- radiating striae incomplete, weakly impressed; erately long. Third abdominal tergum basally antenna of female with 13 segments, A3 slightly smooth or with a lateral pubescent patch; follow- longer than A4; male antenna 14-segmented. ing segments finely punctate but punctures not Pronotum dorsally long, laterally strongly pubes- conspicuous; ventral spine of hypopygium short. cent and with some longitudinal striae or rugae; Legs with tarsal—claws simple. submedial pits conspicuous, asociated with a dis- Comments. Neai/lax is erected to comprise tinct pronotal plate; mesoscurum regularly coria- some species formerly included in Aylax Htg, but ceous and shining; notauli well impressed and which do not fit well in that genus. The new complete; median mesoscutal impression very proposed genus is closely related to Isocolus Forster, short scutellar foveae from it differs cell ; deep, rounded, smooth and which as follows: radial of shining; mesopleuron finely and clearly striated forewing not completely open along the margin longitudinally . Wings hyaline; radial cell of forew- and at the base; first abscissa of radius reaching the ing open on front margin; length of radial cell wing margin and presence of hair fringe on the slightly less than three times width. Third ab- distal margin of the wing (absent in Isocolus). dominal tergum with conspicuous lateral pubes- Biologically Neai/lax differs from Isocolus in its cent patch; following segments not punctate. Tar- association with Lamiaceae rather than with sal claws simple. Asteraceae. The dorsally long pronotum, the Comments. —The genus is morphologically rounded scutellar foveae, and the association with similar to Isocolus Forster, but can be separated by Lamiaceae and not with Papaveraceae separate it the pronotum being strongly pubescent laterally; from Aylax Htg. Volume 3, 1 994 189 — Distribution and Biology. The genus com- Rj reaching margin; length of radial cell about prises two species distributed in centre and south three times the width; hair fringe on distal margin of Europe which produce galls on species of Salvia of wing long. FEMALE, differs from male as fol- (Lamiaceae) and a third species, only recorded lows: antenna with 13 antennomeres (Fig. 98); from the Iberian Peninsula whose host plant and third abdominal tergum without lateral pubes- gall are unknown. cent patch; following segments with fine and in- conspicuous punctures; ventral spine of REVISED SPECIES hypopygium short. — Distribution and Biology. The species was Neaylax salviae (Giraud), comb. n. described from specimens collected in Dalmatia, reared from galls in the fruits of Salvia officinalis L. Aulax salviae Giraud 1859:369 Species of apparently Eastern Mediterranean dis- hocolus salviae (Giraud): Nieves-Aldrey 1988:224. Syn. n. tribution. — Type material. The type series of A. salviae, in SPECIES TRANSFERRED TO the sent to me M. consists of NHMW, by Fischer, NEAYLAX GEN. N. 12 1 mounted on a card specimens. 6 , rectangular with labels: red "Type" and white handwritten Neaylax verbenacus (Nieves-Aldrey), comb. n. "Aulax salviae Giraud", is designated lectotype.

1 8 1 9 , mounted on with labels "Col- , micro-pin, Isocolus verbe>\acus Nieves-Aldrey 1988:221. Syn. n. lect. Mayr" and "Aulax salviae Gir, det. G. Mayr" do not to the series, clearly belong type although This species is closely related to N. salviae are with A. salviae Giraud. In the they conspecific (Giraud). It was described within Isocolus from same series 16, 29, collected deter- by Mayr, Spain. Its galls are produced in fruits of Salvia mined as Aulax salviae Giraud, are cynipids but do verbenaca L. not correspond to that species. Finally 2 speci- mens, identified as A. salviaeby Fahringer, belong Neaylax versicolor (Nieves-Aldrey), comb. n. to the family Eucoilidae. Redescription.—MALE. (Figs. 96-103). Length Aylax versicolor Nieves-Aldrey 1985:122. Syn. n. 1 .8 mm. Red-brown; antenna red-brown; legs with This species was described under Aylax coxae red-brown, tibiae and tarsi paler. Head (Nieves-Aldrey 1985). It is known only from Spain. viewed dorsally (Fig. 96) about 2 times broader Gall unknown. than long; temples not expanded behind the eyes; frons and vertex with sculpture coriaceous- PANTELIELLA Kieffer alutaceous; POL as long as OOL; OOL about 4 times the greatest diameter of lateral ocellus; face Pantelia Kieffer 1901:248 (non Bolivar 1887, Orthopf.) laterally with radiating striae; area supraclypeal Panteliella Kieffer 1901:324. Type species: Aulax fedtschenkoi almost raised, without striae; antenna (Fig. 98) (Rubsaamen 1896). Monobasic with 14 antennomeres, A3 clearly shorter than A4. Endocaulonia Ionescu and Roman 1960:222. Type species: Endocaulonia bicolor Ionescu and Roman 1960. Pronotum (Fig. 100) coriaceous; dorsally long, By origi- nal with distinct designation. submedial pits; mesoscutum (Fig. alutaceous 101) finely and with some obsolete Weld (1952), stated that the types, consisting punctures medially; notauli relatively expanded of slide-mounts of two adults in balsam, were in posteriorly, complete but slightly faint anteriorly; the Berlin Museum. Later on, Quinlan (1968) stated median mesoscutal impression indistinct in poste- that the type material consisted of two specimens rior of 1/4 mesoscutum; scutellar foveae large, deposited in the Vienna Museum but Dr. Fischer rounded, smooth and shining, separated by a informed me that the only material of A. fedtschenkoi narrow scutellum with in septum; weakly rugulose, housed that museum consisted in galls. I have a weak longitudinal furrow; mesopleuron (Fig. had the chance to examine one of the preparations 102) longitudinally striated. Wings hyaline; radial of the Berlin Museum, consisting of a head in bad cell of forewing (Fig. 103) open on front margin but condition, with one of the two antennae being of Hymenoptera Research 190 Journal

The has PHANACIS Forster complete, a leg and the gaster. preparation a red label and is printed "Aulaxfedtschenkoi; type centaureae on Phlomis Plmnacis Forster 1860:145. Type species: Plianacis Onlar, juni 1893; Blattgallen Bijuk Forster 1860. Monobasic. tuberosa". The was redescribed by type species Gillettea Ashmead, 1897:69. Type species: Gillettea taraxaci Weld (1930) and by Quinlan (1968). Ashmead 1897. Original designation. Endocanlonia was described by Ionescu and a discov- in Roman in 1960 based on E. bicolor, species Weld (1952) pointed out that Forster's types ered by these authors in Rumania producing galls the Berlin Museum were destroyed by dermestids I have in the stems of Phlomis tuberosa L. Although but he could observe good specimens of P. centau- clear not been able to examine the types, it seems rea in Vienna. However, I have abundant Iberian that this is from from the detailed description genus material of the type species at my disposal to Panteliella Kieffer of the very similar only differing which I have produced the diagnosis genus. relative of and third with slightly in the lengths pedicel Gillettea Ashmead was synonymized of the two I antennal segment. The descriptions Phanacis Forster by Eady and Quinlan (1963). feature, are and I genera, except for the mentioned fully have examined the type species of Gillettea is and the host which coincident, as the gall plant, confirm the synonymy. consider that Endocaulonia Ionescu and — two times makes me Diagnosis. Head, viewed dorsally with Panteliella Kieffer. Roman is synonymous broader than long; viewed frontally slightly of Endocaulonia to be reticulate- The type species appears broader than high; frons and vertex with Panteliella im- synonymous fedtschenkoi, although coriaceous; radiating striae of face weakly this case I do not venture to establish female antenna in formally pressed, incomplete; 13-seg- the synonymy without a previous examination of mented; male antenna with 14 antennomeres; A3 moder- the type. clearly longer than A4. Pronotum dorsally I the of the submedial present following diagnosis genus ately long, without distinct pits; based on Quinlan's redescription (1968). mesoscurum reticulate-coriaceous; notauli usu- —Female antenna scutellar foveae transverse, Diagnosis. 14-segmented; ally faint anteriorly; in male 15-segmented; A3 and A4 of same length confluent, not well defined or indistinctly closed the two sexes. Mesoscurum reticulate; notauli and reticulate. posteriorly; mesopleuron Wings fully indi- median mesoscutal impression very faintly developed in females, sometimes reduced in males; a broad cated; scutellar foveae distinct separated by radial cell of forewing usually closed, sometimes on distal and striate septum; mesopleuron longitudinally incompletely on the margin; hair fringe cell of on the mar- abdominal striated. Radial forewing open wing margins moderately long. Third on distal distinct. Third Tarsal gin; wing fringe margin tergum without lateral pubescent patch. without lateral abdominal tergum pubescent patch; claws simple. on lateral mar- following tergites weakly punctate Comments. —Phanacis is easily distinguished Tarsal claws gins. simple. from other genera of Aylacini excepting Timaspis is close to on the The genus Neaylax Nieves-Aldrey by the absence of pronotal submedian pits It can be from and to Rhodus Quinlan. separated pronotum and the reticulate sculpture of the obsolete scutellar foveae the former by the notauli, mesopleuron. The genus was considered by Eady female antenna broadly separated and 14-seg- and Quinlan (1963) to comprise also the species the mented. From Rhodus it is distinguished by included in Timaspis Mayr. However, when some distinct on distal of forewing, the not examined fringe margin species of the latter genus by Eady shorter than A4. obsolete notauli and A3 not being and Quinlan, are included in Phanacis, the hetero- Distribution and — one known evident. This is Biology. Only geneity of the grouping becomes Panteliella 1896) cichorii Kieffer and T. species: fedtschenkoi (Rubsaamen, the case with, e.g., Timaspis in central and eastern and between the distributed Europe phoenixopodos Mayr. The boundary It in the de- eastern Mediterranean. produces galls two genera does not, however, seem clearly of Phlomis tuberosa L. P. caulicola leaves and stems (Lamiaceae). fined and some species, particularly and is a Panteliella bicolor (Ionescu Roman) posibly (Hedicke), P. hypochoeridis (Kieffer) and Timaspis of the known from to one or second species genus, only lampsanae Perris, are not easily assigned Rumania. Volume 3, 1 994 191

hair on distal the other genus. All the known species of Phanacis (Fig. 110) open; fringe wing margins on of the absent. Third abdominal without lateral (s. str.) produce galls plants family tergum Asteraceae. pubescent patch. Tarsal claws simple. — — based on one Distribution and Biology. The genus has a Comments. The genus was spe- intro- Rhodus oriundus found in Rhodes palaearctic distribution with two species cies, Quinlan, duced in North America. All the known species Island, galling apical buds of Phlomis cretica Presl are associated with Asteraceae. Galls are produced (Quinlan 1968). Its morphological features relate it mostly on stems. to the genera Isocohis Forster and Neaylax Nieves- Aldrey; it differs from the former in the conforma- Phanacis taraxaci (Ashmead) tion of the notauli, the sculpture of the mesoscutum and the distal margin of the wing completely Gillettea taraxaa Ashmead 1897:69. lacking hairs. The wing venation with Rl and Rs Phanacis taraxaci (Ashmead): and Quinlan, 1963:18. Eady not quite reaching the anterior margin of wing and the absent hair fringe on distal wing margin sepa- Examined material. —The material con- type rate it from Neaylax. sists of two female specimens; one in the USNM — Distribution and Biologi/. The genus comprises and the other in the NHMW (both examined); the one described species, only recorded from Rhodes from Wien, in better condition, is specimen desig- (Greece) associated with Phlomis (Lamiaceae). nated lectotype. The is studied in detail species being by TIMASPIS Mayr Shorthouse (personal communication). Appar- the has been introduced in North ently species Timaspis Mayr 1881:18. Type species:— Diastrophus lampsanae America, together with the host plant Taraxacum, Karsh. Monotypic. (Type. Timaspis phoenixoptodos Mayr 1882. Ashmead 1903 from Europe. But the existence of the species in Designated by p.214). 1963:18. Phanacis Forster (part): Eady and Quinlan Europe is documented only by a few gall records, Aylacopsis Hedicke 1923:81. Syn. n. Type species: Aylacopsis so the european origin of this species does not heraclei Hedicke 1923. Monotypic and original designa- is distributed and seem clear. The species widely tion. relatively abundant in North America (Shorthouse, pers. comm.). The type material of Timaspis lampsanae, in the Museum of Vienna, consists of a 2 mounted on RHODUS Quinlan micro-pin on polyporous strip labeled "Tim. lampsanae det Karsch Type"; "lampsanae Karsch Rhodus 1968:282. Rhodus oriundus Quinlan Type species: Type"; "collect G. Mayr" and red label "Holo- Quinlan 1968. Monotypic and original designation (ex- type". amined) — Diagnosis. Radiating striae present laterally — 104-110. Head viewed dor- Diagnosis. Figs. on the face; incomplete. Female antenna of 13-14 two times broader than viewed sally (Fig. 104) long; segments; 14-15 in males; A3 relatively long; at broader than f rons frontally (Fig. 105) clearly high; least 1.5 times longer than A4. Pronotum dorsally and vertex with reticulate sculpture; radiating moderately long, without distinct submedial pits; striae of face area present laterally; supraclypeal notauli usually complete but often not well im- female antenna raised; (Fig. 106) 13-segmented; pressed anteriorly; median mesoscutal impres- male with 14 A3 shorter antennomeres; clearly sion usually present; scutellar foveae small, than A4. Pronotum with (Fig. 107) dorsally long, confluent, usually indistinctly closed posteriorly; distinct submedial mesoscutum pits; (Fig. 108) mesopleuron reticulate-rugulose, rugulose- with dull reticulate sculpture and some scattered alutaceous or striate-alutaceous. Radial cell of notauli distinct in piliferous punctures; posterior forewing closed on the margin, sometimes incom- of faint or obsolete 3/4 mesoscutum; anteriorly; pletely so anteriorly; hair fringe on distal margins scutellar foveae large, rounded, sculptured; sep- of wing long. Third abdominal tergum without tum lon- relatively broad; mesopleuron (Fig. 109) lateral pubescent patch. Tarsal claws simple. striated. Rl and Rs not gitudinally quite reaching Comments.—Timaspis can be distinguished anterior of radial cell of margin wing; forewing from Phanacis on the basis of its rugulose-reticu- 192 Journal of Hymenoptera Research late sculpture of the mesopleuron and complete orly; scutellum ovate, weakly rugulose; Forew- notauli. Timaspis was synonymyzed with Phanacis mesopleuron (Fig. 116) rericulate-rugulose. Forster by Eady and Quintan (1963) but these ing with radial cell (Fig. 117) open anteriorly on authors' concept of Timaspis was incomplete be- margin but R, clearly reaching wing margin; length hair cause it was limited to the type species of the of radial cell 2.6 times width; fringe on wing abdominal genus, T. lampsanae. But if other species described margins long. Third tergum (Fig. 118) within Timaspis are considered, especially T. without lateral pubescent patch; following seg- it not ventral of phoenixopodos Mayr, I consider advisable to keep ments punctate; spine hypopygium It must be how- visible. Tarsal claws the genera separate. emphasized, quite short, slightly— simple. ever, that the boundary between the two genera is Comments. Within Timaspis Mayr this spe- is close to cichorii not clearly defined and some species are difficult cies morphologically Timaspis it is the to assign to one genus or the other. from which differentiated by coloration, of the and conformation of Aylacopsis Hedicke, was described from a spe- sculpture mesoscutum foveae. to the cies that produces galls in the stems of Heracleum the scutellar According original this in the stems sphondylum L. (Apiaceae) (Hedicke 1923). After description, species produces galls

I of Heracleum at that time having examined the type species, consider that sphondylum (Apiaceae) to be it fits well within Timaspis Mayr. Consequently, I being the first cynipid species demostrated with this of Afterwards synonymize Aylacopsis with Timaspis— Mayr. associated family plants. Distribution and Biology. As understood this species has not been recorded again and, herein, the genus comprises 9 species in West- consequently, the host has not been confirmed. It be out that Phanacis Europe. All the species, excepting one, are associ- may, however, pointed eryngii ated with Asteraceae Diakonchuk, another Aylacini species, was re- cently found in Eastern Europe associated with Timaspis heraclei (Hedicke), comb. n. Apiaceae (En/ngium) (Diakonchuk 1984).

Aylacopsis heraclei Hedicke 1923:81. Syn. n. Timaspis lusitanica Tavares

in the Berlin The type of Aylacopsis heraclei Hedicke Museum Timaspis lusitanicus Tavares 1904:301 of a 9 on a card The consists specimen glued triangle. Timaspis lusitanica Tavares 1904: D.T. and Kieffer 1910 Emend. carries the labels: white handwrit- specimen following Phanacis crepidos Weidner 1965:1. Syn. n. ten "Doicul, Leipzig 26-2-15"; "Aylacopsis heraclei 9 Hedicke" and red printed "Type". Timaspis lusitanica Tavares was described from Portugal (Tavares 1904) reared from galls on We can contribute the following complemen- Barkausia taraxacifolia DC. (Asteraceae). More re- tary data to the original description: Figs. 111-118. cently, it has been recorded from Spain, on the Head, viewed dorsally (Fig. Ill), about two times same host plant (Nieves-Aldrey 1992). Phanacis broader than long; viewed frontally (Fig. 112) crepidos was described from Germany (Weidner slightly broader than high; frons and vertex finely 1965) from material reared from galls on Crepis coriaceous; POL slightly less than two times OOL; biennis L. (Asteraceae). Having examined the type transfacial line about 1.6 times height of eye; face material from the Museum of Hamburg, I have laterally with radiating striae; supraclypeal area concluded that it is the same species described by without striae; with weak median protuberance- Tavares, so I establish the corresponding new malar space about 0.7 times height of eye; antenna 14 synonymy. (Fig. 113) slender, filiform, with 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- Volume 3, 1 994 193

XESTOPHANES Forster LITERATURE CITED

Xestophanes Forster 1869:332, 337. Type species: Cynips Ashmead, W. H. 1885. A Bibliographical and Synonymical potentillae De Villers. Monotypic and original designa- Catalogue of the North American Cynipidae, with de- tion. scription of new species. Transactions of the American Entomological Society 12: 291-304. — Ashmead, W. H. 1887. On the Cynipidous galls of Florida, Diagnosis. Face laterally with radiating striae; with descriptions of new species and synopses of the supraclypeal area without striae; with weak me- described species of North America. Transactions of the dian frons and vertex almost smooth protuberance; American Entomological Society 14: 125-158. antenna of female and shining; 13-segmented, A3 Ashmead, W. H. 1896. Description of new cynipidous galls and in the United as long as or slightly longer than A4; antenna of gall wasps States National Museum. the United States National Museum 19: 113- male 14 segmented. Pronorum dorsally long, lat- Proceedings of 136. erally pubescent; submedial pits conspicuous; Ashmead, W. H. 1897a. Description of some new genera in mesoscutum smooth and notauli com- shining; the family Cynipidae. Psyche 8: 67-69. faint plete or anteriorly; median mesoscutal im- Ashmead, W. H. 1897b. Description of five new genera in the pression slightly indicated; scutellar foveae family Cynipidae. Canadian Entomologist 29: 260-263. Ashmead, W. H. 1903. Classification of the and smooth; mesopleuron smooth and shining. Forew- gall-wasps the parasitic cynipoids, or the super-family Cynipoidea. ing with radial cell open on the margin but R, I-IV. Psyche 10: 140-155 to hair reaching wing margin; fringe moderately Askew, R. R. 1984. The biology of gall wasps, pp. 223-271 . In T. Third abdominal long. tergum laterally smooth N. Ananthakrishnan (ed.), Biology of Gall Insects. Ed- and glabrous. Tarsal claws with weakly devel- ward Arnold, London. Barbotin, F. 1964. Sur une nouvelle et deux en oped basal lobe or tooth . galle cynipides provenance d'Algerie. Marcellia 31: 151-157. Comments. —The genus, together with the al- Belizin, V. I. 1959. Gall wasps of the tribe Aylaxim (Hy- lied form a distinct unit within genus Diastrophus, menoptera, Cynipoidea) new for the fauna of the USSR. the Aylacini: morphologically, by the tarsal claws Entomologkheskoe Obozreme 38: 662-674. having an acute basal lobe or tooth and frons, Burks, B. D. 1979. Cynipoidea, pp. 1045-1107. In K. V. Krombein, P. D. Hurd Jr., D. R. Smith, and B. D. Burks vertex and mesopleuron without sculpture; bio- (eds), Catalog of Hymenoptera in North America North ofMexico. Volume logically, by their association exclusively with 1, Smithsonian Institution Press, Washington, DC. Rosaceae (the genera Potentilla and Rubus). The Dalla Torre, C. W. and J. Kieffer. 1910. Cynipidae. Das Tierreich genus is very likely monophyletic, based on the 24: 1-891. fusion of third and fourth abdominal terga in the Diakonchuk, L. A. 1984. New species of Cynipidae (Hy- females menoptera) from Georgian SSR. Vestnik Zoology 3: 74-77 Distribution and — Two of (In Russian). Biology. species R. D. and 1963 Eady, J. Quinlan. Hymenoptera: Cynipoidea. Xestophanes are known in Western both Europe, Key to families and subfamilies and (includ- makers on stems and runners of of gall species ing galls). Handbooks for the Identification ofBritish,lnsects

Potentilla. . 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: ACKNOWLEDGMENTS 407-564. Forster, A. 1860. Die zweite Centune neuer Hymenoptera I am very grateful to the curators of the collections listed (Eucoela et Phanacis). Verhandlungen derNaturhistorischen Vereins der Preussischen und under materials for the loan of type material and other Rhemlande Westphalens 17: 143-146. specimens. Special thanks to A. Menke (USNM) and R. A. 1869. Ueber die der Danielsson (MZLU) also for sending some interesting bibli- Forster, Gallwespen. Verhandlungen ography. I am also most indebted to one anonymous re- kaiserlich-koemglichen zoologisch-botamschen Gesellschafi in Wien 19: 327-370. viewer and specially to F. Ronquist for a critical revision of 1859. de nouvelles the manuscript including linguistic suggestions. M. A. Alonso- Giraud, J. Signalements quelques especes de et de leurs der Zarazaga helped with nomenclatonal questions. I would like cynipides galles. Verhandlungen to thank Inaqui Diez for the rotulation of the drawings made kaiserlich-koemglichen zoologisch-botanischen Gesellschaft in Wien 9: 337-374. by myself. Financial support was provided by the Research T. 1840. project "Fauna Ibenca II", DGICY PB89 0081. Hartig, 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. (Hymenoptera, Cynipidae). Boletim da Sociedade Hedicke, H. 1923. Beitrage zur kenntnis dercvnipiden (Hym.) portuguesa de entomologia, suppl n°3, Adas do II Congresso XII. Ein neues cecidozoon an Heracleum spondylium. Iberico de Entomologia v. 1: 493-500. L. "In of Zeitschrift fur Pflanzenkrankheiten 33(1-2): 81-83. Nieves-Aldrey, J. press". The types Aylax Hartig Hedicke, H. 1928. Beitrage zur kenntnis der cynipiden (Hym.) described by C.G.Thomson (Hymenoptera, Cynipidae). XIV. Der Erzeuger der levantimschen salviagalle Entotnologica Scandmavica. Deutsche Entomologische Zeitschrift 14: 81-85. Nordlander, G. 1984. Vad vet viom parasitiska Cynipoidea?. Ionescu, M. A. and N. Roman. 1960. Un genre nouveau de Entomologisk Tidsknft 105: 36-40 bicolor n n. 1968. on Cynipides gallicoles Endocauloma g., sp. Quinlan, J. Cynipinae (Hymenoptera) occurring (Hymenoptera, Cynipoidea). Revue de Bwlogie.Acadenuci Phlomis L. Transactions of the Royal Entomological Society Republicu Populare Ronune Bucaresti 5(3): 221-225. of London 120:275-286.

Ionescu, M. A. and N. Roman. 1962. New genera of gall maker Ritchie, A. J. 1984. A revision of the higher classification of the Cynipidae in the fauna of the rumanian people's repub- mquiline gall-wasps (Hymenoptera: Cynipidae) and a revi- lic: Weldiella aequalis n. gen. and Cecconia valerianeliae sion of the Neartic species of Periclistus Forster. Ph. D. Thomson. Revue de Biologie Academia Republicii Populare Thesis, Carleton University, Ottawa, ON. 368 pp. Romine Bucaresti 7: 551-559. Rohwer, S. A. and M M. Fagan. 1917. The type-species of the

KiefferJ.J. 1901 . Monographic desCynipides d' Europe et d' Algene genera of the Cynipoidea, or the gall wasps and para- Hermann edit. Paris, t.I, 687 pp. sitic cynipoids. Proceedings of the United States National et 53: 357-378. Kieffer, J. J. 1902. Revision du genre Aulax des genres Museum, limitrophes d'Aulax, avec quelques notes sur divers Ronquist, F. In press. Evolution of parasitism among closely autres Cynipides. Bulletin de la Societe d Histoire N aturelle related species: phylogenetic relationships and the ori- deMetz. 10:91-97. gin of inquilinism in gall wasps (Hymenoptera, Kinsey, A.C. 1920. Phylogeny of cynipid genera and biologi- Cynipidae). Evolution. cal characteristics. Bulletin of the American Museum of Ronquist, F. and G Nordlander 1989. Skeletal morphologv of Natural History 62: 357-402. an archaic cympoid, Ibalia rufipes (Hymenoptera: Kovalev, O. B. 1982. Cynipid gall-makers of the subfamily Ibaliidae). Entomological scandmavica Suppl. 33: 1-60. their S. 1904. novo. Aylacinae (stat. n.) (Hym., Cynipidae) and species Tavares, J. da Descnpcao de un Cynipide described within the family Figitidae. Trudy Broteria 3: 301-302. Zoologicheskogo Instituta Akademiya Nauk SSR 110:85-93. Thomson, C. G. 1877. Oefversigt af sveriges Cynips Arter. (in russian) Opuscula Entomologica_8: 732-841. Mayr, G. 1881. Die Genera der gallenbewohnenden Wachtl, F. 1880. Beitrage zur Kenntnis der gallenerz. Insecten Cynipiden. 20, fahresbenchte der Communal-Oberrealschule Europas. Verhandlungen der kaiserlich-koeniglichen im l. Bezirke pp. 1-38. zoologisch-botamschen Cesellscliaft in Wien 30: 531-545. Mayr, G. 1882. Die europaischen arten der gallen- Wachtl, F. 1891. Eine neue Gallwespe. Wiener Entomologische bewohnenden cynipiden. 21, fahresberichte der Commu- Zeitung pp. 277-280. nal-Oberrealschule im I. Bezirke pp. 1-44. Walsh, B. D. 1869. Galls and their architects. The American L. los 2: Nieves-Aldrey, J. 1984. Notas sobre Aylaxini (Hvm, Entomologist 70-74. Cynipidae, Cynipinae) de la Peninsula Iberica con Weidner, H. 1965. Phanacis crepuios sp. nov., eine neue descripcion de una nueva especie de Isocolus. Eos 60: gallwespe an Crepis biennis L. Entomologische Mttteilungen 235-250. 52: 1-7.

Nieves-Aldrey, J. L. 1985. Nuevos Aylaxini (Hym., Cynipidae) Weidner, H. 1968. Zur Kenntnis der Gallwespentnbus para la Peninsula Iberica con descripcion de una nueva Aulacini (Hymenoptera, Cynipidae). Entomologische especie de Aylax Htg. Boletim da Sociedade Porttuguesa de Zeitschrift 10: 105-120. entomologia, suppl. 1, Adas do II Congresso Iberico de Weld.L. H. 1930. Notes on Types (Hymenoptera: Cynipidae). Entomologia 117-128. Proceedings of the Entomological Society of Washington 32: L. 8. Nieves-Aldrey, J. 1987. Estado actual de conocimiento de la subfamiha Cynipinae (Hym.. Parasitica, Cynipidae) Weld, L. H. 1952. Cynipoidea (Hym.) 1905-50. Privately printed, en la Peninsula Iberica. Eos 63: 179-195. Ann Arbor, MI, 351 pp.

Nieves-Aldrey, J. L. 1988. Descripcion de una nueva especie Wiebes-Rijks, A. A. 1979. A character analysis of the species de Isocolus Forster con notas de otras especies de Aylaxini of Synergus Hartig, Section II (Mayr, 1872) (Hy- nuevas para la Peninsula Iberica (Hym., Cynipidae). Eos menoptera, Cynipidae). Zoologische mededelingen 53(28): 64: 221-227. 297-321.

la lista Nieves-Aldrey, J. L. 1992. Adiciones a de cinipidos Zerova, M. D., Diakonchuck, L. A. and V. M. Ermolenko.

ibericos, especialmente Aylaxini, incluyendo la 1988. Gall-formers of the european part of the USSR Part 2 . descripcion de una nueva especie de/lH/aniifa Ashmead 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 tavaresi Nieves-Aldrey, 1984 syn. n. scabiosae (Giraud, 1859) Centaurea AULACIDEA Ashmead, 1897 rogenhoferi Wachtl, 1880 syn. n. abdominalis (Thomson, 1877) ? serratulae (Mayr, 1882) comb, n Serratula arnicae Hoffmeyer, 1930 Arnica follioti Barbotin, 1972 Sonchus LIPOSTHENES Forster

hieracii (Bouche 1834) Hieracium glechomae (Linnaeus, 1758)... . Nepeta artenusiae (Thomson, 1877) latreilki (Kieffer, 1898) crassinervis (Thomson, 1877) foveiger (Thomson, 1877) NEAYMXgen.n. kerneri (Wachtl, 1891) comb, n Nepeta salviae (Giraud, 1859) comb, n Salvia

' kieffen Cotte, 1915 verbenacus (Nieves-Aldrey, 1988) comb, n Salvia laurae Nieves-Aldrey, 1992 ? versicolor (Nieves-Aldrey, 1985) comb, n ? nibletti Quinlan, 1969 Hieracium

pilosellae (Kieffer, 1901) Hieracium PANTELIELLA Kieffer, 1902 scorzonerae (Giraud, 1859) Scorzonera ENDOCAULONIA Ionescu and Roman, 1960 syn. n. subterminalis Niblett, 1946 Hieracium fedtschenkoi (Rubsaamen, 1896) Phlomis tragopogonis (Thomson, 1877) Tragopogon bicolor (Ionescu and Roman, 1960) Phlomis

AYLAX Hartig, 1840 PHANACIS Forster, 1860 hypecoi Trotter, 1912 Hypecoum GILLETTEA Ashmead, 1897 minor Hartig, 1840 Papavcr caulicola (Hedicke, 1939) Picris papaveris (Perris, 1839) Papaver centaureae Forster, 1860 Centaurea artenusiae Weidner, 1968 misident. punctipleuris (Thomson, 1877) hypochoendts (Kieffer, 1887) Hypochoens BARBOT1N1A gen. n. taraxaci (Ashmead, 1897) Taraxacum oramensis (Barbotin, 1964) comb, n Papaver RHODUS Quinlan, 1968

CECCON1A Kieffer, 1902 onundus Quinlan, 1968 ... . Phlomis WELD/ELM Ionescu and Roman, 1962 syn. n. valerianellae (Thomson, 1877) Valenanella T1MASP1S Mayr, 1881 aequalis (Ionescu and Roman, 1962) Valenanella AYLACOPSIS Hedicke, 1923 syn. n. cichorn Kieffer, 1909 Cichonum DIASTROPHUS Hartig, 1840 heraclei (Hedicke, 1923) comb, n Heracleum rubi (Bouche, 1834) Rubus lampsanae Perris, 1873 Lampsana mayri Remhart, 1876 Potentilla lusitanica Tavares, 1904 Crepis crepidos (Weidner,1968) syn. n. HED1CK1ANA gen. n. phoenixopodos Mayr, 1882 Lactuca levantina (Hedicke, 1928) comb, n Salvia pilicornis (Thomson, 1877) ? rufipes Ionescu and Roman, 1959 Crepis IRAELLA gen. n. sonchi (Steffarti, 1900) Sonchus luteipes (Thomson, 1877) Papaver urospermi Kieffer, 1901 Urospermum papaveris Kieffer in Goury and Gignon, 1905 XESTOPHANES Forster, 1869 1SOCOLUS Forster, 1869 brevitarsis (Thomson, 1877) Potentilla fitchi (Kieffer, 1898) Centaurea potentillae (Retzius in Degeer, 1873) Potentilla jaceae (Schenck, 1863) Centaurea foveicollis (Thomson, 1877) lichtensteini (Mayr, 1882) comb, n Centaurea abreviatus (Thomson, 1877) 196 Journal of Hymenoptera Research

divisa 9. Figs. 1-15. Cynipidae. 1-3, Pronotum, dorsal view. 1, Aylax papaveris, 2, Aulacidea hieracii. 3, Cynips 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. 35, Head, dorsal view. frontal view. Antistrophuspisum. 36, Head, 37, Antenna ( 9 ). 38, Antenna ( 6 ). 39, Pronotum. 40, Mesoscutum and scutellum. 41, Body, lateral. 42, Forewing (hairs omitted). Volume 3, 1994 199

43-55. Aulacidea. Aulacidea dorsal Figs. 43-50, liarringtom. 43, Head, 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

56-62. Barbotinia oraniensis. dorsal view. frontal Figs. 56, Head, 57, Head, view. 58, Antenna ( 9 ). 59, Antenna ( 6 ). 60, Pronotum. 61, Mesoscutum and scuteilum. 62, Body and wings. Volume 3, 1 994 201

%vj&

63-70. levantina. Figs. Hedickiana 63, Head, dorsal view. 64, Head, frontal view. 65, Antenna ( 9 ) 66, Pronotum. 67, Mesoscutum and Scutellum. 68, Mesopleuron. 69, Forewing (hairs omitted). 70, Gaster, lateral. 202 Journal of Hymenoptera Research

Pronotum. Figs 71-79. hocolus scabwsae. 71, Head, dorsal view. 72, Head, frontal view. 73, Antenna (9). 74, Antenna (6). 75, 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

96-103. salviae. dorsal view. frontal Figs. Neaylax 96, Head, 97, Head, view. 98, Antenna ( 9 ). 99, Antenna ( 9 ). 100, Pronotum. 101, Mesoscutum and scutellum. 102, Body, lateral. 103, Forewing (hairs omitted) Volume 3, 1994 205

104

105

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 Sierra of (central 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 California at Riverside museum and were col- crabronine genera in the tribe Oxybelini. Bohart lected by P. H. Timberlake. Diligent work by Bohart and Menke (1976:359-370) gave an overview of Museum collectors at Davis has made some 3,000 present day knowledge of the tribe. They pointed more specimens available for study. out that the presence of metanotal projections Principal cooperating museums and individu- (squamae) and a propodeal projection (mucro) in als are listed below. Museum designations are addition to a lateral carina on terga I-III, and identified by their city locations in capital letters. sometimes IV-V, characterize Belomicrus. As far as The manuscript was read by Dr. Arnold Menke known, distribution of the genus is holarctic and and Dr. Eric Grissell. Their comments and correc- Ethiopian. Bohart and Menke listed 63 species of tions were quite helpful, and their assistance is 25 I which were North American. now recognize gratefully acknowledged. 40 species in the latter region, of which 14 are new. earlier BERKELEY— P. D. Hurd Among writers the work of Pate J. Powell, (deceased), Essig Mu- seum, of California at (1940a,b) was outstanding. His overall assessment University Berkeley. DAVIS— L. S. Kimsey, S. Heydon, Bohart Museum, Univer- of Belomicrus and its relatives was brilliant, and his sity of California at Davis. were meticulous. The criticisms descriptions only LOGAN—G. E. Bohart, Entomology Museum, Utah State that can be made are that some of his phylogenetic University.— discussions suffered from lack of material, and his NEW YORK ]. Rozen, Entomology Museum, American Museum of Natural descriptions contain much detail irrelevant at the — History. OTTAWA L. Masner, Biological Resources Division, Agri- species level. culture Canada. Belomicrus are ground-nesting wasps, mostly PHILADELPHIA—S. Roback, Daniel Otte, D. Azuma, Acad- in areas. sandy Prey of the forbesii group are emy of Natural Sciences. nymphal Miridae (Bohart 1956, Evans 1969). The RIVERSIDE—P. H. Timberlake (deceased), S. Frommer, En- Museum, of California at River- franciscus group provisions with adult dasytid tomology, University side. beetles (Williams 1936, Bohart and Menke 1976). SAN W. P. FRANCISCO— J. Pulawski, Arnaud, Entomology Prey of the other are unknown. groups Museum, California Academy of Sciences. of the Members genus are few in most collec- VIENNA—M. Fischer, Naturhistorische Museum, Vienna, tions. Through museum visits and borrowings I Austria. WASHINGTON—A. S. K. V. U.S. Na- have been able to examine about 600 specimens. Menke, Krombein, tional Museum. The majority of these are in the University of 208 Journal of Hymenoptera Research

Terms used in the species key and in descrip- projection is divided by a median carina. tions which may need explanation are: epipleural The ten groups of North American Belomicrus "button", small raised area just behind pronotal which Pate (1940b) proposed are now reduced to lobe; T-I, T-II, etc. terga after propodeum; S-I, S-II, five. Bohart and Menke (1976) considered only 2 etc., sterna after propodeum; LID, least interocular groups, forbesii and cladothricis. With the discovery distance; PD, puncture diameter; post-tegula, basal of additional species it now seems desirable to wing sclerite. divide the forbesii group into four according to the The generotype, Belomicrus italicus A. Costa following key. (1871), has a palearctic distribution and is similar The purpose of this and other keys in the to species in the forbesii group. However, it should paper is to give a quick means of making identifi- be placed in a different group because the clypeus cations. It is understood that the key user should has a considerable flattened area on the median have access to at least a representative named lobe, and the more posteriorly blunt metanotal 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 Instead of the six now valid species recog- nized by Pate, this figure has been raised to 10, The only previous key to the forbesii group is which includes four new species. that as of a given part more extensive one by Pate Members of theforbesii group have relatively (1940b:210). The forbesii part was based on about few "structural" differences. However, the colora- 65 specimens, of which some 43 were forbesii and tion and color patterns seem to be quite constant in columbianus, 10 were cookii, seven were penuti (as a long series. Therefore, I am treating columbianus as subspecies of forbesii), 3 were coloratus, and two distinct from forbesii, sierrae as distinct frompenuti, were querecho. I have studied about 1,500 speci- and desertus as distinct from cookii. Details on mens of the group. 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 weak or faint on T-III and — T-II, 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 all collar 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 bands indistinct at least Tergal markings weak, pale 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 II and III in T-I, sometimes 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 coloration in front of Tergal yellow bands dark brown to nearly black, postocular area nearly always with but definite low tubercles forbesii Robertson

Belomicrus coloratus Baker Belomicrus columbianus (Kohl)

Belomicrus coloraia Baker 1909:29. Holotype female, Ormsby columbianus Kohl 1892:208. female Oxybelus- Lectotype (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. larimerensis Rohwer 1908:417. Nevada and California. They were all taken in the Oxybelus Holotype female, Lanmer Co " Colorado (WASHINGTON), examined, months of June and July at elevations of 4,000 to • Treated as a of Bohart and oonnr lm j j i i ^ , synonym forbesii bv Menke 8,000feet. Nevada county records include Ormsby, (1976) Revised syn;nymy . Douglas, and Eureka. California counties are Al- El Dorado, pine, Inyo, Lassen, Mono, Nevada, In all/ 54 males ^j 69 fem ales have been Placer, Sierra, and Tuolumne. A single Oregon srudied. These were taken during the months of record is e. 5 mi Bly, Klamath Co. ra May ( rely), June, July, and August, mostly at The principle diagnostic features, especially elevations above 4,000 feet. Western state records the are in the tergal markings, given 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.; paratypes) were from California counties: San Ber- Morrison Meadow and Snowslide Park, Trinity nardino (Victorville, Morongo Valley, Red Moun- Co. tain), Imperial (Palo Verde), Los Angeles (Llano, The new status of columbianus and its syn- near Lovejoy Buttes), Ventura (Chuchupate Ranger onym, larimerensis is indicated by the difference in Sta., Frazier Mt.), Kern (Dove Well), inyo (Panamint 8 mi w. in markings from forbesii as outlined in the key. Mts., Brown). Paratypes cooperating museums. Belomicrus cookii Baker This species is similar to powelli which also has well developed postocular tubercles, the basal Belomicrus cookii Baker 1909:29. Lectotype male, Claremont, slope of T-I red, and the epipleural "button" Los Co., California (WASHINGTON). Lecto- Angeles scarcely indented. However, desertus usually has type designated by Pate (1940a). extensive whitish yellow tergal markings. Some- times, these may be on T-I only. See also poioelli. This species is known from a few southern — Etymology. specific name derived from Latin I 8 and 11 California localities. have seen males = adjective, desertus abandoned. females from Los Angeles Co. (Claremont, Newhall) and Riverside Co. (Riverside, Gavilan), Belomicrus forbesii (Robertson) all taken during the months of April and May at Figs. 12, 17 low elevations in foothill locations.

Oxybelus forbesii Robertson 1889:85. Holotype male, Colo- Belomicrus desertus R. Bohart, new species rado (PHILADELPHIA), examined.

— red are: Female holotype. Length 5 mm. Black, I have studied 118 males and 122 females, clypeal apex, foretarsus partly, tegula dully, terga collected from May to August at elevations of on mostly, sterna except for dark median blotches 4,000 to 10,000 feet in the following states and are: mandible T-I to T-III; whitish yellow basally, counties: California (Alpine, Lassen, Modoc, Mono, on scape partly, flagellum beneath, apical spots Nevada, Sierra, Siskiyou, Trinity), Colorado femora, tibiae outwardly, squamal triangle, apical (Larimer, Denver), Montana (Missoula), Nevada bands on T-I to III, weak on IV-V; wings nearly (Elko, Pine, Washoe), Utah (Box Elder, Kane, Sum- and clear. Pubescence moderate, silvery appressed mit), and Wyoming (Sublette, Teton, Uinta). Fe- lateral frontal band on clypeus, enlarged above, male face (Fig. 17). short supraclypeal spot, postocular, mesopleural, Diagnostic characters, particularly the abun- lateral areas silvered. Punc- and tergal moderately dant tergal markings, are given in the key. tation fine and close, a little shiny across frons below ocelli; pygidial plate with coarse, separated Belomicrus montanus R. Bohart, new species punctures. Postocular tubercles well developed, "button" indented. — are: epipleural weakly Female holotype. Length 5 mm. Black, red Male. — 4-5 mm. Whitish are: Length yellow clypeal apex, terga and sterna mostly except for tarsal mandible mostly, clypeus across apex, basal dark blotches on S-I-II; whitish are: mandible ba- bands on T-I to VII segment, apical (usually). sally, scape apically, flagellum beneath, pronotal female 3 mi s. Kramer Holotype (DAVIS), collar and lobe, apical spot on forefemur, tibiae San Bernardino IV-6-66 Junction, Co., California, outwardly, squamal triangle, T-I apex faintly; 123 31 females (R.M. Bohart). Paratypes, males, wings lightly stained. Pubescence light, silvery in of various topotypical, collected April years by and appressed on clypeus, lateral frontal area, E.I. M.E. D.S. R.M. Bohart, Schlinger, Irwin, short supraclypeal spot, postocular and Parker. Other Horning, J.C. Hall, F.D. paratypes mesopleural areas moderately. Punctation fine from San Bernardino Co., California: 3 males, 5 and close, not shiny across frons below ocelli; Tree National females, Joshua Monument, (H.K. pygidial plate with coarse, separated punctures. 13 4 Adelanto Court); males, females, (E.I. Postocular tubercles quite weak, epipleural "but- et 6 2 Schlinger, al.): males, females, Apple Valley ton" slightly indented. R. P. D. Other (W. Mason, Hurd). specimens (not Male. —Length 4-4.5 mm. Foretarsus partly Volume 3, 1 994 211

T-I off-white, whitish apically. outwardly, foretarsus dully, squamal triangle; Holotype female (DAVIS), White Mts., Mono wings moderately stained. Pubescence silvery Co., California, 10,000 ft., VII-10-68 (R.M. Bohart). appressed on clypeus, lateral frontal area, stout Paratypes, 14 males, 13 females, topotypical but supraclypeal triangle, postocular and mesopleural collected from 22 to 23 areas fine June July by J. W.MacSwain, (weakly). Punctation and close, not shiny G.I. across frons J. Powell, and Stage. Other California below ocelli, somewhat shiny on terga; paratypes: 1 pair, Benton, Mono Co., V-23-86 (R.M. pygidial plate with coarse, separated punctures. Bohart); 3 males, Westgard Pass, Inyo Co., V-VI-37 Postocular tubercles well developed, epipleural (G.A. Hamsher, CD. Michener); 2 females, Deep "button" slightly indented. Springs, Inyo Co., V-13, V-14 (L.D. French, N.J. Male. —Length 4-4.5 mm. Characters about as Smith). Paratypes in cooperating museums as far in female. Scape whitish at apex. as possible. Holotype female (SAN FRANCISCO), La There is some similarity to querecho, powelli, Panza, 12 mi ne. Pozo, San Luis Obispo Co., Cali- and since the basal of T-I is IV-29-62 desertus, slope mainly fornia, (J. Powell). Paratypes, 5 males, 16 red, and the epipleural "button" is scarcely in- females (all from California collected in April and dented. From all three of these the weakly indi- May): Monterey Co.: 1 male, 5 females, Hastings cated tubercles of are differ- postocular montanus Reserve (D.L. Linsdale); 1 pair, Arroyo Seco (D. entiating. Also, it may be separated from querecho Burdick, P. Torchio); female, 12 mi n. Cholame the distinct sternal of the by punctation latter, as (D.H. Janzen); San Luis Obispo Co.: 2 pair, well as its broad on S-II. R.W. 5 mi —depression topotypical (J. Powell, Thorp); female, Etymology. specific name derived from Latin ne. Santa Margarita (R.W. Thorp); Fresno Co.: 12 = adjective, montanus dwelling on mountains. mi w. Coalinga (J.W. MacSwain); Ventura Co.: 3 males, 8 females, Chuchupate Ranger Sta., Frazier Belomicrus Pate Mt. P.D. penuti (J. Powell, Hurd). Paratypes in SAN FRANCISCO, DAVIS. Belomicrns Pate 1940:27. forbesii penuti Holotype male, The well developed postocular tubercles are Yosemite Valley, Mariposa Co., California (PHILADEL- also found in desertus, which see. The all red terga PHIA). Raised to species status by Bohart and Menke and broken band of are (1976). yellow pronotal powelli distinguishing. Although primarily a coastal form, occurs with desertus on Frazier Mt., Ventura The 1,115 males and 51 females I have studied poioelli Co., at Station. In were collected during April to September in Cali- Chuchupate Ranger fairly long series the differences in the seem to fornia and neighboring Nevada at elevations given key hold. The is named for mostly above 5,000 feet. California counties are: species my friend, Jerry Powell, a who has also worked with Alpine, El Dorado, Fresno, Lassen, Mariposa, lepidopterist Mono, Nevada, Placer, Calaveras, Sierra, Tulare, wasps. Amador, and Tuolumne. Nevada records are In- Belomicrus Pate cline Village and Little Valley, Washoe Co. Also, querecho Pate (1940a:28) recorded penuti from southern Belomicrus Pate 1940:36. male, Oregon: Crater Lake and Lake of the Woods, Kla- querecho Holotype Alamogordo, Otero Co., New Mexico (PHILADEL- math Co. I have not studied these specimens. PHIA), examined. The relationships to columbianus and forbesii are outlined in the key. This southwestern species is largely restricted to the Chihuahuan Life Zone. I have studied 3 Belomicrus powelli R. Bohart, new species males and 20 females, collected in April, May, and June in the following states: Arizona (near Apache Female — 5 mm. red are: holotype. Length Black, and Willcox, Cochise Co.; Tubac, Santa Cruz Co.; and sterna for clypeal apex, terga mostly except Oak Creek Valley Road, Yavapai Co.; 10 mi w. dark blotches on S-II-III; whitish are: mandible Jacob Lake, Coconino Co.). Utah (25 mi s. Moab, beneath, collar medi- basally, flagellum pronotal Grand Co.). New Mexico (Skeleton Canyon, on ally, pronotal lobe, apical spot femora, tibiae Peloncillo Mts., and Rodeo, Hidalgo Co.; 212 Journal of Hymenoptera Research — Alamogordo, Otero Co.; 5 mi e. Las Cruces and Etymology. the specific name refers to the Leasburg Dam, Dona Ana Co.). Nevada (4 mi s. California Sierra. Co. and Texas Warm Springs, Nye ) (Rankin, Upton Co.). Belomicrus apache group

Belomicrus sierrae R. new Bohart, species The — apache group contains only the single Female holotype. Length 5 mm. Black or dark known species, apache Pate. It was described in brown, red are: flagellum beneath dully, foretarsus detail from a single female by Pate (1940a) and dully, apical half of pygidial plate; whitish yellow placed in a key by Pate (1940b). This is one of the are: mandible basally, femorotibial joints, tibiae smallest known Belomicrus, approximately 2 mm outwardly, apicomedial spot on squamal triangle in length. The female has a peculiar anchoriform (all yellow in some paratypes); wings moderately terminal process figured by Pate (1940a) and herein stained. Pubescence silvery appressed on clypeus (Fig. 16). The male clypeus is more ordinary. Fea- laterally, narrowly on lateral frontal area, short tures of the group are characterized in the key. A supraclypeal spot, weakly on postocular and brief description of the species, including the pre- mesopleural areas. Punctation fine, close, dull; viously unknown male, and additional locality clypeus wrinkled, with a pair of coarse punctures records follow: above polished apex, pygidial plate with coarse, separated punctures. Postocular tubercles present Belomicrus apache Pate but weak, epipleural "button" deeply indented Figs. 13, 16 but not covered with a thin membrane; lateral propodeal carina flattened, somewhat bent in- Belomicrus apache Pate 1940a: 15. Holotype female, Las Cruces, ward. Dona Ana Co., New Mexico (PHILADELPHIA), exam- ined. Male. —Length 4.5-5 mm. About as in female. Pronotal lobe sometimes dully pale. Female. — Length 2 mm. Black with red Holotype female (DAVIS), Sagehen Creek, unhanded abdomen; legs partly whitish yellow, Nevada Co., California, 6,500 feet, VI-25-68 (R.M. frons at narrowest point about equal to eye breadth, Bohart). 118 males, 97 females, Paratypes, with an anchoriform to VII-25 clypeus apical projection topotypical, V-29 (R.M. Bohart, J. A. Skin- (Fig. 16), no postocular tubercles or omaulus, ner, L.S. Kimsey, M.E. Irwin, D.S. Horning, J.E. translucent in- B. squamal triangle laterally, deeply Slansky, J. Powell, Villegas). Also paratypes, 10 cised posteriorly (Fig. 13). males, 19 females, Grass Lake, El Dorado Co., Male. —Length 2 mm. Much as in female but California, 8,000 ft., VII-5-62 and VII-16-62 (R.M. LID slightly greater and clypeal bevel triangular Bohart et ai). Other (not specimens paratypes) rather than anchoriform. have been studied from other California counties I have seen three males and seven females at 6,000 ft. or above: Alpine, Calaveras, Plumas, from the following western states: New Mexico Lassen, Placer, and Trinity. Paratypes in all coop- (Las Cruces, Doha Ana Co.; Rodeo, Hidalgo Co.), erating museums. Arizona (Aztec, and 7 mi s. Quartzite, Yuma Co.; The dark basal slope of T-I and the deeply Quijotoa and Sells, Pima Co.; Willcox, Cochise Co. indented epipleural "button" ally sierrae with on Euphorbia mat), California (Cathedral City, Riv- penuti. However, the "button" is not as thinly erside Co. on Euphorbia polycarpa). Collections were lidded above as in penuti. Futhermore, the latter made from August to October. seems always to have at least several of the terga with whitish bands. The dark abdomen ot sierrae is Belomicrus vanyume group similar to that of colwnbianus, and conceivably the two occur species might together in California The vanyume group consists of three species of near the Oregon border. In addition to the which one is described below as new. Pate (1940a, epipleural "button" difference, columbianns has b) placed vanyume in a monotypic group of its the lateral carina whereas in I propodeal simple, own. The vanyume group, as consider it, is unified sierrae it is not but flattened bent in- sharp and by the distinct omaulus running vertically along ward. 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 in tooth (Fig. 15), T-I to V usually red, especially females maricopa Pate

Belomicrus maricopa Pate supraclypeal patch. Punctation of mesonotum and Figs. 14, 15 interocellar area coarse and separated by 1-2 PD of microsculpture, mesopleuron similar but rugose Belomicrus maricopa Pate 1947:54. Holotype female, Higley, below, tergal punctation moderate and a little Co., Arizona (WASHINGTON), examined. Maricopa shiny, coarse on pygidial plate. LID about 1.7x Belomicrus mariposa Pate 1947:55. In error. scape length, clypeus with median raised area I have seen 44 males and 58 females from bearing longitudinal depression, clypeal bevel Arizona (Higley, Willcox, 28 mi s. Quartzite, transverse and nearly linear, frontal groove well Picacho Canyon), New Mexico (5 mi e. Las Cruces, impressed, postocular tubercles well developed, 9 mi n. Cotton City), California (18 mi w. Blythe, gena in side view bent outward at middle below, Palo Verde), and Sonora, Mexico (Guaymas). mandible strongly expanded at distal one-third, This species is similar to texensis, described front dorsal margin of pronotum obtusely cor- below and differences enumerated. The unde- nered, squamal unit as in Fig. 14, forefemur scribed male is much like the female but is a little rightangled near base, pygidial plate narrowly smaller and the mandible is somewhat less ex- triangular.— panded. Also, the last few terga are darker. The Male. Length 4 mm. Legs mostly brownish principal differences between the two species are red, forewing veins dark, clypeal bevel broadly the shinier mesopleuron and apically toothed triangular, T-I-II red, III a little red laterally and clypeus of maricopa (Fig. 15). posteriorly, VII dully red. Female holotype (DAVIS), Kingsville, South texensis R. Belomicrus Bohart, new species Pasture, Kleberg Co., Texas, VI-16-68 (J. E. 4 females NEW — Gillaspy). Paratypes, (DAVIS, Female holotype. Length 5 mm. Black, light YORK), same data as holotype but V-6-67 and V- yellow are basal two-thirds of mandible, foretibia 6-68; on Ratibida columnaris; paratype male (SAN outwardly; brownish are: legs partly; reddish are: FRANCISCO), Mexico: Baja California Sur, e. edge forefemur at T-I to 111-24-84 base, III; wings clear, veins of Sierra Placeres, (W. J. Pulawski). mostly orange. Silvery appressed pubescence on The large size (for Belomicms\), large postocular clypeus mostly, broad patch along eye margin tubercles, expanded mandible subapically and all reaching up as far as an imaginary line drawn dark antenna are also found in maricopa. However, across beneath midocellus, short but large texensis has many differences: pronotum all dark, 214 Journal of Hymenoptera Research punctation of mesopleuron and mesonotum much (Darwin Falls). These were all taken in April, May, closer and not shiny, female clypeus with a longi- and June. The previously undescribed male is tudinal flattened and shallowly grooved median much like the female but the last two terga may be ridge, clypeus also without a forward-pointing dark. apical tooth, wing veins of female mostly orange Belomicrus franciscus group instead of black, marginal cell of forewing some- The absence of an omaulus, triangular median what shorter, and T-III to V mostly or all black. 11), mandible not —the name indicates "of or squamal complex (Fig. angled Etymology- specific beneath toward and with an from" Texas. apex, clypeus apical deflected bevel or polished triangle, when taken the I have identified Belomicrus vanyume Pate together distinguish group. eight species, of which three are herein described as new. Belomicrus vanyume Pate 1940a:17. Holotype female, characters are the extent of Victorville, San Bernardino Co., California Species tergal yel- (PHILADEPHIA), examined. low markings, flagellarcoloration, pronotal macu- lation, punctation of the postmandibular area of In addition to the type I have seen five males males, development of the mandibular midtooth, and three females, all from California counties: and form of the clypeal apex. Since females may be Riverside (Box Canyon, Thousand Palms), San difficult to distinguish, the following key is based Bernardino (Apple Valley, Kramer Hills, Adelanto, on males. 1 1 mi w. Ludlow), Imperial (Fish Creek Mts.), Inyo

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 beneath I-IX or tooth Flagellum pale yellow on X; mandible 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 beneath before tooth Flagellum orange apex; mandible 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 T-I to Ground color on 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. bevel often with a 1); clypeal 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 rather Flagellum 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 counties: Placer (Dutch Flat, Baxter, Colfax), Tuolumne (Strawberry), El Dorado (Pyramid Belonucrus cahuilla Pate 1940:39. Holotype male, Andreas Ranger Station and near Icehouse Road). Collec- Canyon, Riverside Co., California (PHILADELPHIA), tors were J. G. Rozen, W. R. M. examined. J. MacSwain, Bohart, W. J. Pulawski, and H. M. Kimball Court. Paratypes are deposited in museums listed in The male differs from that of mono by the acknowledgements. simple and convex clypeal bevel. In the female the This species is close to Serrano. Both have the bevel is more indented than that in mono or quemaya. male area Both postmandibular practically impunctate, sexes have dark median areas on the terga, but in californicus the mandible tooth is well devel- which may extend all across on T-III and follow- oped, and punctation of the interocellar area is a ing. Only a single female has been identified, so little less dense. variation cannot be assessed. The type series of MacSwain and I observed a four males came from Andreas Canyon (Palm J. large nesting area in 1956 near Pyramid Ranger Station north of Springs). I have also seen five males from close by Placerville, California. Females were provisioning (Taquitz Canyon, F. D. Parker and L. A. Stange). ground nests with dasytine beetles, Amecocerus Other southern California records are: male, fe- cervicalis Blaisdell male, near San Bernardino Co. (Melyridae). Cajon Junction, (J. — Etymology. the specific name indicates "of or C. Hall, E. I. males, Walker Pass and Schlinger); from" West Wofford California. Heights, Kern Co. (J. Powell); male, 2 mi e. Banner, Riverside Co. (H. C. Dickson). Belomicrus costalis R. Bohart, new species Dates of capture ranged from April 16 to June 7 Fig. 5 (males), and July 4 (female).

Male — 4.5 mm. Black marked Belomicrus californicus R. Bohart, new holotype. Length species with whitish yellow: scape and mandible mostly, Figs. 7, 9 flagellum beneath, pronotum all across, post- unit, forefemoral small Belomicrus franciscus Pate of Bohart, in Bohart and Menke tegula, squamal large spot, 1976:363. Misidentification. distal midfemoral spot, tibiae and metatarsi out- wardly, subapical band on T-I, attenuate laterad; — red Male holotype. length 4.5 mm. Black, whitish are: clypeal bevel, T-I to III mostly, T-IV to VI are: mandible mostly, scape in front, flagellum except for median dark blotches, T-VII, wings beneath except at apex, pronotal collar medially, weakly stained. Silvery appressed pubescence in of inner a little lobe, squamal unit, tip mucro, femora distally, strip along eye margin, expanded tibiae outwardly, foremetatarsus dully; red are: above, supraclypeal spot, postocular area and bevel clypeal mostly, abdomen except for median mesopleuron moderately, terga lightly. Puncta- dark blotches on T-III to VI; wings weakly stained. tion sparse on mostly polished postmandibular area of Silvery appressed pubescence along inner eye head venter, fine and close on dorsum of margin, stout supraclypeal spot, weak pubescence body, somewhat reflective on interocellar area. on and bevel somewhat arched postocular area, mesopleuron, terga. Punc- Clypeal — (Fig. 5). tation sparse on mostly polished postmandibular Female. Length 5 mm. Terga all red, scape area of head venter, fine and close on dorsum of pale in front only, flagellum reddish yellow be- neath. body, a little shiny on interocellar area. Clypeal Clypeal bevel flat, not margined below; bevel quite thin (Fig. 7), distinctly darkened later- mandibular inner tooth well developed. ally; mandibular inner tooth well developed. Holotype male (DAVIS), Cache Creek Can- Female. — 5 Yolo Length mm. Clypeal bevel slightly yon, Co., California, IV-30-54 (R. M. Bohart). margined below, mandible (Fig. 9). Paratypes (all from California at low to moderate Holotype male (DAVIS), Baxter, Placer Co., elevations): male (DAVIS), 4 mi nw. Lake Berryessa, elev. California, 5,000 ft., VI-16-56 (R. M. Bohart). Napa Co., V-12-61 (F. D. Parker); 2 males (DAVIS, Paratypes: 18 males, 21 associated females, all WASHINGTON), Sacramento, Sacramento Co., from the California V-19-61 M. Sierra at 4,000 to 5,000 ft. (R. Bohart); male (BERKELEY), Alpine elevation to in Marin VI-6-57 during May July, the following Lake, Co., (J. Powell); 2 males 216 Journal of Hymenoptera Research

(DAVIS), Midlake, Lake Co., V-10-56 (E. I. Silvery appressed pubescence in strip along inner Schlinger); male (BERKELEY), near Quincy, eye margin nearly to level of midocellus, stout 2 Plumas Co., Vl-22-49 (J. W. MacSwain); males supraclypeal spot, postocular area, mesopleuron (NEW YORK), Tuolumne City, Tuolumne Co., V- moderately, terga lightly. Punctation moderate on area of head fine and close 30-53 (J. G. Rozen); male (DAVIS), Boca, Nevada postmandibular venter, Co., VI-19-62 (R. M. Bohart); male (DAVIS), Kyburz on dorsum of body, slightly reflective on Flat, Sierra Co., VII-15-76 (R. M. Bohart); male interocellar area. Clypeal bevel thin all across (Fig. Daffodill Amador VI-5-63 (DAVIS), Hill, Co., (R. 6). — M. Bohart). Also, 6 presumed and associated fe- Female. Length 5 mm. Whitish bands on T-I- males from localities listed above: Cache Creek V sometimes better developed in para types, prono- Canyon, Boca, Daffodill Hill, Sacramento. tum sometimes white all across, pygidial plate The band on the pronotal collar is sometimes partly red. Postmandibular area and rest of head broken in both sexes. The arched clypeal apex (Fig. venter polished. Mandibular inner tooth well de- 5) differentiates males from serrano and californicus veloped, clypeal bevel thin but impressed over its which agree with costalis in having the entire breadth. postmandibular area polished. Females are best Holotype male (DAVIS), Sonora Pass, distinguished by association with males. Tuolumne Co., California, elev. 9,624 ft., VII-6-61

. — Etymology specific name derived from Latin (R. M. Bohart). Paratypes (all from California Si- = noun, costa side. The species occurs especially on erra at elevations above 5,000 ft.): female, same the western side of California. data as holotype; 3 females, Hope Valley, Alpine Co., VII, 1948, 1978 (R. M. Bohart); male, Fred's Belomicrus franciscus Pate Place, El Dorado Co., VII-10-67 (R. M. Bohart);

Figs. 2, 3, 11 male, Gold Lake, Sierra Co., VII-8-54 (R. M. Bohart). Paratypes all in DAVIS museum.

Belomicrus Pate 1 331 Lone Moun- franciscus :77.Holotype male, This species belongs to the subgroup in which tain, San Francisco, San Francisco Co., California (PHILA- males have the postmandibular area somewhat DELPHIA), examined. punctate. It differs from other members of the subgroup by the extensively black (rather than I have studied several of the type series. The red) ground color of the terga. In cahmlla the terga rather stout male flagellum (Fig. 2), extensively may be partly black but the male clypeal bevel is yellow-banded terga, and partly punctate male not rimmed above, and the terga are not spotted postmandibular area distinguish this species. Fe- with whitish yellow. male face The series was collectecd (Fig. 3). type by — Etymology. specific name derived from the F. X. Williams (1936) in sandhills of San Francisco. Greek melanos = black. Williams found a colony of the species in late April and of 1930. Females were early May provisioning Belomicrus mono Pate with Trichochrous antennatus Mots. (Melyridae). In all probability the type locality has been taken Belomicrus serrano mono Pate 1940:46. Holotype female, 5 mi over a by housing development. n. Round Valley, Mono Co. California (PHILADEL- PHIA), examined. Belomicrus mono of Bohart and Menke 1976:363. Belomicrus melanus R. Bohart, new species Pate, Fig. 6 The male belongs to the subgroup with punc- — tate area on the head venter. The Male holotype. Length 4.5 mm. Black or postmandibular characters in the it. Females are brown, whitish are: scape in front, mandible mostly, given key separate best identified association with males. flagellum beneath, pronotum medially, pronotal by I have seen 83 males and 65 associated females lobe, post-tegula, squamal unit, tip of mucro, outer from to in the California Sierra of distal spots on fore and midfemora, tibiae and 5,000 7,000 feet, the counties: Mono holo- foremetatarsus outwardly, subapical band on T-I, following (Round Valley Placer Sierra weak subapical bands on T-II to VI, apex of VII; type), (Lake Tahoe), (Yuba Pass, Si- erra Nevada dark red: clypeal bevel; wings weakly stained. Valley, Independence Lake, Sattley), Volume 3, 1 994 217

Glenn (Sagehen Creek), (Plaskett Meadows). The southern California. The range includes low el- Creek is a Sagehen locality represented by long evation localities in the following counties: Los series. Collection dates are in and mostly June Angeles, Riverside, San Bernardino, San Diego, July. Kern, Ventura, and San Luis Obispo. Collection dates are from March 27 to May 7. Belomicrus quemaya Pate 10 Figs. 1, 4, Belomicrus cladothricis group

Belomicrus Pate 1940:47. quemaya Holotype male, PalmSprings, The deeply divided squamal lobes of the met- Riverside Co., California (PHILADELPHIA), examined. anotum (Figs. 28-43), and short body length, char- Belomicrus jurumpa Pate 1940:53. Holotype male, Gavilan, acterize the With the of Riverside Co., California (PHILADELPHIA), examined. group. exception cladothricis Cockerell New synonymy. (1895), all previously known species Belomicrus of Bohart and were franciscus quemaya Pate, Menke described by Pate (1940a, 1947). The group 1976:363. appears to be largely confined to western United a few The punctate postmandibular area of the head States, only species occurring east of the 100th meridian or in Sonora and the venter in the male places quemaya in its subgroup. Baja Califor- The nian of Mexico. a small unusually long and lightly colored flagellum peninsula Considering of amount of and the 6 new de- both sexes and the distinctive clypeal bevel of synonymy species the male scribed herein, the total number of known (Figs. 1, 4) are distinguishing. Wings are species in the is now 18. In nearly clear in both sexes. Females are best identi- group most of these the squamal fied lobes are at the base dark by association with males. B. jurumpa appears separated by integu- ment. In some to be a simple synonym based on an examination species, however, the pale color and setal texture be of the holotypes. may continuous basally, at least in females. These are I have studied 60 males and 30 associated cladothricis, eriogoni, and sechi. characters of the females collected from March to May. California Species tiny forms in this are county records are: Riverside (Whitewater Can- group squamal structure, scutal and interocellar punctation, frons breadth, bevel yon, Gavilan, Thousand Palms, 18 mi w. Blythe, clypeal details, Palm and the somewhat variable coloration. In one Springs), San Bernardino (Kramer Jet., spe- cies, sechi, the unusual amount of Adelanto), Kern (Short Canyon, Iron Canyon), silvery pubes- cence, and of the male are Imperial (Kane Springs, Fish Creek Mts., Palo shape scape distinguish- Verde), Inyo (Mazourka Canyon, Lone Pine), San ing. little Diego (Borrego). Other records are Clark Co., Although is known about the life history Nevada of these are all (Searchlight, Jean), and Sonora, Mexico (9 species, they presumably ground mi e. Kino). nesting (personal observation of eriogoni at Antioch, California), and they are most frequently collected Belomicrus serrano Pate on the flowers of Eriogonum. Other flower hosts are Fig. 8 reported Chilopsidis, Lepidospartum, Euphorbia, Cladothrix, Cleomella, Rhamnus, Chrysothamnus, Belomicrus serrano serrano Pate 1940:42. Holotype male, Solidago, Cercidium, Prosopis, Sphaeralcea, Asclepias, San Tujunga Wash, Gabriel Mts., Los Angeles Co., Cali- Baeria, and Salsola. Because of the small size of fornia (PHILADELPHIA), examined. in the species cladothricis group (2-4 mm long), it can be surmised that like other The they, many di- polished postmandibular area of the male minutive sphecids, provision with thrips. the in the same places species subgroup as costalis A deal of great descriptive matter on the group and califomicus. The dull interocellar area and was given by Pate (1940a, b, 1947). However, he weak mandible tooth serrano. (Fig. 8) separate had a limited amount of material to work with and the Also, relatively simple male clypeal bevel, and many of his new species were based on one or two the indented one of the female are additional specimens, often of a single sex. His remarks on characters. must not geographical range be taken too literally. I have seen 46 males and 27 females in addi- For instance, on timberlakei he said (1940a:93), tion to the series. The type species is widespread in "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 — T-I or T-I-I1I red Area between squame truncate (Fig. 41), scutal punctation quite coarse, with some 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 Bohart and longer than their interapical distance (Fig. 38) longiceps R. interocellar area vertex 12. Body with exceptional amounts of silvery pubescence (Fig. 30), and silvery setose, Pate propodeum usually extensively red, male scape strongly swollen distally (Fig. 27) sechi — 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 This rather wide-ranging and relatively abun- Fig. 43 dant species is represented in the DAVIS collec- tion by 450 males and 198 females taken at 51 Belomicrus bridwelli Pate 1940a:93. Male holotype, Clifton, localities during March to October. All of the Fairfax Co., (PHILADELPHIA), examined. Virginia captures were made south of 41.5° longitude and west of 100° latitude. Outlying states in the range In addition to the and four holotype topotype are Nebraska, Texas, Utah, Idaho, and California. I have also studied the male, paratypes, following: Elevations were moderate to low and many were Branford, Suwannee Co., Florida, VII-31-30; 2 in strictly desert areas. As might be expected of males, female, Fort Bragg, Cumberland Co., North such a common and widespread species, there is a Rhea Carolina; female, Dayton, Co., Tennessee, considerable amount of variation. The short and VI-20-54; male, 15 mi nw. Howard Co., Big Spring, rather broad squamae are a consistent feature. Texas, VI-1 3-63. According to these meager records, Also, the narrow LID of females (Fig. 18) helps bridwelli occurs east of 102° latitude in May, June, identification. Many western examples have yel- and July. Female squamae and mucro (Fig. 43). low spots on basal terga. Punctation of the scutum and interocellar area is close but allows some Belomicrus cladothricis (Cockerell) shininess not found in several related species such Fig. 18 as eriogoni. The legs are usually extensively red- dish but this occurs in some other species. T-I-II Oxybelus cladothricis Cockerell 1895:309. Female lectotype, are often reddish in males but II be all dark Las Cruces, New Mexico (PHILADELPHIA), examined. may brown as in the of or I-II brown as Lectotype designated by Cresson (1928). type prosopidis, Belomicrus cladothricis prosopidis Pate 1940a:72. Male holo- in the type of minidoka. type, Palm Springs, Riverside Co., California (PHILA- DELPHIA), examined. Belomicrus minidoka Pate 1940a:79. Male holotype, Hagerman, Gooding Co. Idaho (WASHINGTON), examined. 220 Journal of Hymenoptera Research

Belomicrus cucamonga Pate frontal suture characterize the species. The flagel- Fig. 25, 31 lum is a little longer and more slender than in other species. Belomicrus Pate 1940a:76. Female — cucamonga holotype, Camp Etymology. named for Dr. Darwin French, San Gabriel San Bernardino California Baldy, Mts., Co., who discovered Darwin Falls in 1860. (PHILADELPHIA), examined.

Belomicrus eriogoni Pate In addition to the holotype I have seen three Fig. 22 females from Big Pine, Inyo Co., California, VII- 10-61 (R. M. Bohart); and a female from Batchelder Belomicrus cladothrias eriogoni Pate 1940a:70. Male holotype, VI-10 K. Springs, Inyo Co., (H. Court). The more San California Lucas, Monterey Co , (PHILADELPHIA), abundant whitish of the scape and tarsi (Fig. 25), as examined. well as the larger postocular tubercles and broader Belomicrus eriogoni Pate of Bohart and Menke (1976). frons, distinguish cucamonga from inyo. Female I have studied 240 males and 165 females in squamae and mucro (Fig. 31). the DAVIS collection. These were taken during to in Belomicrus darwini R. Bohart, new species May September California, Arizona, New and Utah. California are Fig. 32 Mexico, Nevada, records from Mono Co., Lassen Co., Sacramento Co., and — Contra Costa south to the Mexican border. Female holotype. Length 3 mm. Black or dark Co., Mexican records are from California. brown, white are: mandible mostly, scape in front, Baja At times the is on the flowers pronotal lobe, squama, mucro distally; reddish species abundant, especially of Female facial are: clypeal bevel, flagellum (pale orange, spotted Eriogonum fasciculatum. propor- tions laterally), tibia and tarsi mostly (brownish red), (Fig. 22). hindtibia lighter in basal ring, dull apical bands on Belomicrus R. T-I to III; wings clear, veins black. Silvery ap- inyo Bohart, new species pressed pubescence extending upward along two- Fig. 20, 26, 28 thirds of inner eye margin, short supraclypeal Female — patch, weak on genal and mesopleural areas. Punc- holotype. Length 2.5 mm. Black, white tures of scutum and interocellar area medium fine, are: mandible mostly, scape in front, pronotal foretibia slightly separated by microsculpture. LID about lobe, squama, outwardly, mid and 1.4x hindtibiae on basal one- reddish are: scape length, frontal suture relatively deep third; clypeal abdomen brownish on T-II to halfway from midocellus to clypeus, clypeal bevel bevel, (more VI); veins black. margined and triangular, postocular tubercle small, wings clear, Silvery appressed pubes- short and curved cence is narrow lower half of inner squamae — slightly (Fig. 32). along eye Male. Length 2.5 mm. Scape mostly brown- margin, supraclypeal patch conical, moderate on and Punctures of ish, flagellum a little spotted laterally as viewed in gena mesopleuron. scutum and front. interocellar area medium fine, slightly separated about to Holotype female (DAVIS), Darwin Falls, Inyo by microsculpture. LID equal scape length frontal suture not Co., California, May 17, 1970 (R. M. Bohart). (Fig. 20); deep; faintly visible Paratypes, 2 males, 2 females, same data as holo- halfway from midocellus to clypeus; clypeal bevel type but collected by E. E. Grissell and R. M. triangular, postocular tubercles small; squamae well Bohart. Other paratypes, female (DAVIS), Wikiup, curved, separated distally (Fig. 28). Male. — Mohave Co., Arizona, IV-11-57 (T. R. Haig); fe- Length 2.5 mm. Hindtibia mostly or- male (RIVERSIDE), Kyle Canyon, Clark Co., Ne- ange with dirty white basal area, mid and vada, V-4-41 (P. H. Timberlake); female (DAVIS), hindmetatarsi orange or reddish brown. LID about 1.7x near Boron, Clark Co., Nevada, VI-7-41 (P. H. scape length. Timberlake). Holotype female (DAVIS), Antelope Springs, Co. As indicated in the key, the rather close punc- Inyo California, V-10-61 (R. M. Bohart). 2 1 female tation, moderate LID, brownish legs and terga, Paratypes, males, (DAVIS, BERKE- short squamae, light orange flagellum, and long LEY), topotypes, VI-29-61, VII-1-61, VIII-24-60, on Volume 3, 1 994 221

Solidago (P. D. Hurd, G. I. Stage). Spliaeralcea (R. M. Bohart). Paratypes collected April The orange to brown metatarsi (Fig. 26), dark to August, 4 males, female (DAVIS, WASHING- flagellum, narrow female LID (Fig. 20), basally TON), 18 mi w. Blythe, Riverside Co., California banded mid and hindtibiae (Fig. 26), close and (M. E. Irwin, F. D. Parker, R. M. Bohart, R. W. dull punctation, and well separated squamae (Fig. Brooks); male (DAVIS), 15 mi w. Baker, San Ber- characterize the nardino California female 28) — species. Co., (N. J. Smith); (BER- Etymology. specific name is an acronymic KELEY), Surprise Canyon, Inyo Co., California (P. noun inspired by Inyo Co., where all of the type D. Hurd); male (DAVIS), Patrick, Washoe Co., series was collected. Nevada (E. E. Grissell); female (NEW YORK), 2 mi n. Rodeo, Hidalgo Co., Nevada (J. G. Ehrenfeld). Belomicrus istam Pate Characteristic are the long head (Fig. 24), dark Fig- 37 antenna, sharp postocular tubercles, extensively silvery female gena and mesopleuron, brown wing Belomicrus istam Pate 1940a:81. Male holotype, 4.5 mi nw. veins, and unhanded abdomen. The only other Edom, Riverside Co., California (PHILADELPHIA), species with a rather long head (compare Figs. 21, examined. 24) is the much more abundantly marked oraibi, which see. I have seen a total of 37 males and 62 females collected from March to June in areas of 11 sandy Belomicrus mescalero Pate localities. These range in California from Sacra- Fig. 34 mento to Borrego Valley, and in Nevada from Nixon to Stillwater. I collected a long series of both Belomicrus mescalero Pate 1940a:87. Female holotype, sexes on a in Sand lighting sandy slope Canyon, Alamogordo, Otero Co., New Mexico (PHILADEL- Kern Co., California, on April 14, 1962. Female PHIA), examined squamae and mucro (Fig. 37). This species is known to occur in April and Belomicrus longiceps R. Bohart, new species May. In addition to the three female types, one of Figs. 24, 38 which came from Needles, California, I have stud- ied the following: male, female, 3 mi sw. Rodeo, Female — 2 holotype. Length mm. Black, white Hidalgo Co., New Mexico, IV-30-65 (J- G. Rozen). are: mandible mostly, pedicel partly, pronotal lobe, The partly polished scutum, together with the post-tegula, squama, mucro, tibiae outwardly, dark wing veins, mostly red terga, and whitish- tarsi; reddish are: clypeal bevel, abdomen; wings yellow spotted T-I are distinguishing. The male is clear, veins brown. Silvery appressed pubescence in general agreement with the female description present along lower half of eye margin, short given by Pate, but T-VI-VII are reddish. Female supraclypeal patch, dense on genal area and squamae and mucro are shown in Fig. 34. mesopleuron. Punctation on scutum and interocellar area fine and close but a little irregu- Belomicrus oraibi R. Bohart, new species lar, somewhat shiny. Head unusually long and Figs. 21, 36 narrow (Fig. 24), inner eye margins nearly straight frontal — (Fig. 24), carina distinct one-third of dis- Female holotype. Length 3 mm. Black, white tance from midocellus to clypeus, LID 1.7x length are: mandible mostly, scape in front, pronotal of scape, clypeal bevel broader than high, lobe, tegula, post-tegula, metanotum, mucro dis- tubercle postocular prominent and sharp, genal tally, tibiae and tarsi outwardly (a little reddish), area flattened to a little inside concave row of erect T-I to VI mostly; reddish are: scape partly, flagel- short hairs, squamae but curved and well sepa- lum beneath, clypeal bevel, femora distally, tibiae rated most (Fig.—38). partly, terga basally; wings clear, veins light Male. length 1.8 mm. T-I-II partly black, T- orange. Silvery appressed pubescence along lower III-VII black. three-fourths of eye margin, short supraclypeal Holotype female (DAVIS), 6 mi nw. Adelanto, patch, dense on genal area and mesopleuron. Punc- San Bernardino Co. California, V-ll-79, on tation fine, close and dull on scutum, a little reflec- 222 Journal of Hymenoptera Research tive on interocellar area. Head longer than usual Belomicrus pallidus R. Bohart, new species frontal 33 (Fig. 21 ), inner eye margins nearly straight, Fig. 23, carina distinct, LID 1.7x length of scape, clypeal — bevel broader than high, postocular tubercles Male holotype. Length 2 mm. Black, white but dark prominent and sharp, genal area slightly concave are: mandible mostly, antenna mostly inside row of erect hairs; squamae short, curved, distally in front, pronotum all across, tegula and and well separated (Fig. 36). post-tegula, wing veins, squamae mucro, legs

Male . —Length 2.5 mm. About as in female but beyond basal two-thirds of femora; reddish are: T-IV-VI black. Genal area not concave. legs basad of white markings, abdomen; wings Holotype female (DAVIS), Oraibi, Navajo Co., clear. Silvery appressed pubescence on face below Arizona, VIII-30-76, on Chrysothamnus (R. M. midocellus except for a median bare spot (Fig. 23), Bohart). Paratype female (WASHINGTON), genal and mesopleural areas densely silver, scutum Dinnebito Wash, 5 mi nw. Hotevilla, Navajo with scattered silvery setae. Punctation fine and Co., Arizona, VIII-31-86 (A. S. Menke); paratype sparse on extensively polished mesonorum, a little male (DAVIS), 23 mi nw. Page, Kane Co., Utah, VI- less so on interocellar area. LID (just above anten- male nal 1.4x inner 22-70 (N. J. Smith); paratype (SAN FRAN- insertion) scape length, eye margin CISCO), Green River, 5 mi ne. Jensen, Uinta Co., weakly curved below midocellus (Fig. 23); frontal and bevel obscured Utah, VI-26-82 (W. J. Pulawski). groove clypeal (if any) by tubercle Although generally similar to lotigiceps, which silvery pubescence; postocular promi- also has an unusually long head, oraibi differs by nent, sharp; squamae short, curved, pubescent, its pale wing veins and extensively white terga. well separated (Fig. 33). The concavity of the lower gena of the female is Female.—Unknown. even more pronounced. The presumed males from Male holotype (SAN FRANCISCO), Borrego, Utah agree with the female in punctation and San Diego Co., California IV-25-54 (P. D. Hurd). wing vein color. However, they have T-IV to VI Paratype male (DAVIS), Blythe, Riverside Co., dark. The presently known range includes north- California, on Asclepias, VII-8-56 (M. Wasbauer). ern Arizona and southern Utah. The rather extensive silvery pubescence is Etymology. —the name is an acronymic noun reminiscent of sechi, but there are many points of inspired by the Oraibi indians of Navajo Co., difference, one of which is the simple scape in Arizona. pallidus male. General features characterizing pallidus are the considerably polished mesonorum, whitish and red and white Belomicrus pachappa Pate wing veins,— legs. Fig. 40 Etymology, specific name based on the Latin = adjective pallidus pale. Belomicrus pacliappa Pate 1940a:73. Female holotype, River- side, Riverside Co., California (PHILADELPHIA), ex- amined Belomicrus potawatomi Pate I have studied 11 males and 30 females col- Fig. 41 lected from April to August at 4 localities in Cali- Belomicrus Pate 1947:57. Female Sioux fornia: Weed, Siskiyou Co. (J. Powell); Sand Can- potawatomi holotype, City, Woodbury Co., Iowa (WASHINGTON), exam- yon, Kern Co. (R. M. Bohart); Chuchupate Ranger ined. Station, Ventura Co. (J. Powell, P. D. Hurd); River- side, Riverside Co. (P. H. Timberlake). Nevada The holotype was collected July 12, 1935 by C. records axe from Washoe Co.: Wadsworth (R. M. N. Ainslee. The only other record of which I am Bohart, G. I. Stage); Patrick (R. M. Bohart). The 3 aware is: female, Regnier, Cimarron Co., Okla- species, pachappa, tuktum, and istam, have the squa- homa, June 9 (NEW YORK). This species is similar mae elongate and nearly touching distally. The to bridwelli which also has the squamae far apart. close and dull punctation of pachappa separates it. However, the metanorum is posteriorly pointed The female has T-I-VI red, but the male has the (compare Figs. 41, 43) in bridwelli and its scutal abdomen dark following T-I or T-II. Female squa- punctation is less coarse. mae and mucro (Fig. 40). 223 Volume 3, 1994

Belomicrus sechi Pate veins, and orange-tinted leg markings character- ize the Figs. 27, 30 species. — derived from the Etymology. specific name = Belomicrus sechi Pate 1940a:60. Male holotype, 6 mi n. Palm Latin adjective siccus dry. Co. California Springs, Riverside (PHILADELPHIA), examined. Belomicrus timberlakei Pate

Fig. 35 I have seen 17 males and five females of this were extensively silvery pubescent species. They Belomicrus timberlakei Pate 1940a:91. Male holotype, River- in the desert ex- taken from May to August following side, Riverside Co., California (PHILADELPHIA), localities of southern California: Palm Springs, amined. Magnesia Canyon, and Rancho Mirage, all River- has been collected from to side Co.; 11 mi w. Ludlow, San Bernardino Co.; This species April I have studied 41 males and 1 1 females and Borrego Valley, San Diego Co.; a male (SAN September. and New FRANCISCO) comes from Arizona: Roper Lake from California, Nevada, Utah, Arizona, Mexico. In California 13 localities from State Park, Graham Co., V-26-83 (W. J. Pulawski). range Co. to Riverside Co., in the Tran- The male has a unique, distally swollen scape (Fig. Siskiyou mostly the sition Life Zone. One record is at 8,000 ft. on 27). Otherwise, the female agrees with descrip- and mu- Carson Pass, Co. Non-California records tion given by Pate. The female squamae Alpine are: Cochise Co., Arizona (Willcox, Santa cro are shown in Fig. 30. Apache, Rita Mts.); Eureka Co., Nevada (Eureka); Churchill Grant Utah Belomicrus siccatus R. Bohart, new species Co. Nevada (Frenchman); Co., (Moab); Co. Utah Co., New Figs. 19, 29 Emery (Greenriver); Hidalgo Mexico (Rodeo); Baja California, Mexico (10 mi e. Female — 2.5 mm. Black, white holotype. Length San Quintin). are: mandible basally, pronotal tubercle, squa- Of the 3 species with all dark pronotum are: mae, mucro distally; light dull orange wing (timberlakei, potazoatomi and bridzoelli) only tibiae and tarsi; dark red are: T-I-II, py- veins, tnnberlakei has the squamae only moderately sepa- gidium; wing membrane clear. Silvery appressed the male in rated (Fig. 35). The female resembles in a broad patch reaching up along 2x the pubescence nearly all respects but its LID equals scape two- thirds of inner eye margin, supraclypeal patch length instead of 1.7. undeveloped, postocular and mesopleural patches of scutum and interocellar moderate. Punctation Belomicrus tuktum Pate area fine and a little separated, somewhat shiny. Fig. 39 LID 2x scape length, inner eye margin gently curved tubercles but (Fig. 19), postocular present Belomicrus tuktum Pate 1940a:84. Male holotype 5 mi e. Edom, not sharp, squamae nearly straight (Fig. 29). RiversideCo., California (PHILADELPHIA), examined. Holotype female (DAVIS), Thousand Palms Canyon, Riverside Co., California, IV-9-64 (R. M. In addition to the holotype I have studied 24 Bohart). Paratypes, from Riverside Co.: 2 males males and 25 females collected from March to IV-17-87 7 include the (DAVIS), Whitewater, (N. J. Smith); May. The 13 localities represented males (DAVIS), SAN FRANCISCO, WASHING- following California counties: Inyo (Darwin Falls, TON), same data as holotype; 2 male topotypes Little Lake, Surprise Canyon), Kern (Sand Can- N. 2 Riverside (DAVIS), 111-29-77 (R. M. Bohart, J. Smith); yon), Ventura (Quatal Canyon), (Thou- male topotypes (DAVIS), 111-31-63 (F. D. Parker); sand Palms, Deep Canyon, 18 mi w. Blythe, female topotype (DAVIS), IV-11-70 (R. M. Bohart). Whitewater Canyon), San Bernardino (5 mi n. Also a paratype female (RIVERSIDE), Apple Val- Barstow), Imperial (Glamis), and San Diego V-20-41 H. is similar to ley, San Bernardino Co., California, (P. (Borrego Valley). This species quite Timberlake). istam. Females have the abdomen essentially all The combination of unusually broad frons dark red, but males usually have T-I dark red, the are (Fig. 19) black flagellum, nearly straight squamae rest brown. Female squamae and mucro shown to brownish (Fig. 29) orange (female) (male) wing in Fig. 39. 224 Journal of Hymenoptera Research

south of 38.6 Arizona Belomicrus vierecki Pate low elevations longitude. records are from Wickenburg, Tucson, and Santa Fig. 42 Rita Mts. New Mexico records are Alamogordo, Las Cruces, and near Valencia Co. Mexi- Belomicrus vierecki Pate 1940a:56. Male holotype, Alamogordo, Laguna, Otero Co., New Mexico (PHILADELPHIA), examined. can material has been seen from Sonora (Alamos), Puebla (Petlalcingo), Durango (Nombre de Dios), This is a relatively abundant and widespread and Baja California Sur (La Ribera). The most species. I have seen 272 males and 47 females eastern record in the United States is Presidio, collected from March to July. I swept a long series Texas at about 104.5° latitude. from flowers of Chilopsis linearis at 15 mi w. Baker, The dark brown abdomen, white tibiae and San Bernardino Co., California in May. Other pre- tarsi, extensively polished interocellar area, nar- ferred hosts are Cercidium and Prosopis. The 15 row LID, and well separated crescentic squamae at localities for California are all sandy situations (Fig. 42), make identification relatively simple.

REFERENCES

Baker, C. F. 1909. Studies in Oxybelidae I. Pomona journal of Pate, V. S. L. 1931 . A new Belomicrus from the West. Ent. News Entomology 1:27-30. 42: 77-78. Bohart, R. M. 1956. Prey captures of Belomicrus penuti and B Pate, V. S. L. 1940a. The taxonomy of the Oxybeline wasps I. coloratus. In Bohart and Menke 1976: 363. Ibid. A review of the genera Belomicroides, Bnmocelus and Bohart, R. M. and A. S. Menke. 1976. Sphecid Wasps of the Belomicrus with particular reference to the nearctic spe- World. A Generic Revision, ix + 695 pp. Univ. California cies. Transactions of the American Entomological Society 66: Press, Berkeley 1-99. Cockerell, T. D. A. 1895. The second Anacrabro and the small- Pate, V. S. L. 1940b. The taxonomy of the Oxybeline wasps II. est American Oxybelus. Canadian Entomologist 27: 308- The classification of the genera Belomicrus and

309. Enchenncrum . Transactions of the American Entomological

Costa, A. 1871 . Prospetto sistematico degh Immenotterologia Society 66: 209-264. Itahana. Ann Mus. Zool. Univ. Napoli 6: 28-83. Pate, V. S. L. 1947. New North American Belomicrus. Proceed- Cresson, E. T. 1928. The types of Hvmenoptera in the Acad- ings of the Entomological Society of Washington 49: 54-57. emy of Natural Sciences of Philadelphia other than Robertson, C. 1889. Synopsis of North American species of those of Ezra T. Cresson. Memoirs of American Entomo- Oxybelus. Transactions of the American Entomological Soci- logical Society 5: 1-90. ety 5&. 77-85. Evans, H. E. 1969. Notes on the nesting behaviour of Pisonopsis Rohwer, S. A. 1908. Four new Hvmenoptera. Entomological clypeata and Belomicrus forbesu. Journal of the Kansas News 19:417-420. Entomological Society 42: 117-125. Williams, F. X. 1936. Notes on two oxybehd wasps in San Kohl, F. F. 1892. Neue Hymenopterenformen. Ann. Nat Hist Francisco. Pan-Pacific Entomologist 12: 1-8. Hofmus Vienna 7: 197-234. Volume 3. 1 994 225

1. 3. quemaya 2. franciscus franciscus

4. quemaya 5. costalis 6. melanus 7. californicus 8. serrano

forbesii 16. 17. 15. maricopa apache

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

19. siccatus 20. inyo 21. oraibi 18. cladothricis

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 Digonogastra Viereck (Quicke 1988b, Wharton et al. 1989). As Myers implied; however, I. pennipes Diatraea, a large genus of New World pyralid does not belong to Digonogastra, and in fact, it is a moths, whose larvae bore in the stems of member of a new genus near Myosoma Brulle graminaceous plants, includes several species of which is described below. considerable economic importance in both the By coincidence, Westwood (1882) had earlier temperate and particularly the tropical regions described another species belonging to the same where they attack sugarcane, corn and sorghum new genus under the name of Myosoma pennipes, (Elias 1970, Chippendale 1979, Agnew et al. 1988, which therefore becomes a senior, subjective hom- Pashley et al. 1990). Several genera of parasitic onym of pennipes Myers. In addition to the type wasps of the braconid subfamily Braconinae have specimens of Myosomatoides pennipes (Westwood) been recorded as attacking species of Diatraea, viz. comb. nov. (Hope Entomological Collections, Ox- Bracon Fabricius (Muesebeck 1925, Shenefelt 1978), ford) and M. myersi comb, et nom. nov. (Natural C Digonogastra Viereck (Wharton et al. 1989), History Museum, London: type no. 3 418) the Myosoma Brulle (Quicke 1989) and Palabracon author has seen seven other specimens belonging Quicke (Quicke 1988a). All of these are believed to to the new genus from Argentina, Brazil, Colom- be ectoparasitoids attacking relatively late instar bia, British Guiana, Ecuador, Paraguay, and Peru, host larvae. In addition to these, Myers (1931) though none of these has any associated host data. described and illustrated three species of Both M. myersi and M. pennipes appear to be wide- Braconinae reared from Diatraea spp. in British spread in South America. Guiana, all of which he placed in Ipobracon Available material suggests that there are only Thomson, though he recognized that one of his two species of Myosomatoides in South America, new species, I. pennipes Myers (non Westwood), viz. M. pennipes and M. myersi nom. nov. A key is was an "isolated species". Ipobracon was subse- provided to enable their separation. Some varia- to be a quently shown junior synonym of tion is apparent for both species. In particular, Cyanopterus Haliday (Quicke 1985, 1987) and the specimens of M. pennipes vary in the extent of the limits of that genus were restricted such that the black coloration on the body, the wing coloration, of great majority New World species described the extent and length of setosity, the degree of under both and Ipobracon Iphiaulax Foerster, and compression of the hind leg, the shape of the head, including the other two species described by Myers, the relative lengths of hindwing veins lr-m and to a related actually belong distinct, though genus, SC+RI, and the relative length of the ovipositor. 228 Journal of Hymenoptera Research

face Malar area Because no two specimens examined were found separated from by weak groove. Face setose to be entirely similar, I consider it most likely that weakly impressed, punctate. densely they represent a single rather variable species except for smooth, shiny, glabrous supraclypeal area. Antennal sockets rather than a diverse polyspecific complex though triangular approximately ex- level with middle of Frons im- the latter possibility cannot be completely eye. very weakly cluded. pressed with indistinct midlongitudinal sulcus; de- with dense short Myers (1931) provided a fairly complete moderately setosity. Mesosoma. Mesosoma Notauli scription of M. myersi (= pennipes Myers non smooth, shiny. Westwood), in which he illustrated the rather indicated only by weak depressions at anterior of smooth. distinctive metasoma. Following Myers' original mesoscutum. Scutellar sulcus narrow, without carinae. record, the species was recorded as a parasitoid of Propodeum simple, Diatraea, on several further occasions (Myers 1932, Forewing. Veins 1-M and 1-SR+M straight, cu-a interstitial the 2nd sub- Thomson 1953) and, in particular, of D. angustella vein (or virtually so), vein 3-SR 0.67-0.95 Dyar in Argentina (Parker et al. 1953, De Santis & marginal cell rather long (i.e. vein less than or to 0.5 times Esquival 1966). Members of the related genus times SRI), m-cu equal 2-SR. Myosoma are similarly idiobiontectoparasitoids of length of Vein lr-m or less stem-boring pyralid moth larvae, including Hindwing. more straight, vein Diatraea and Chilo species (Quicke 1989, Quicke & slightly shorter than, to slightly longer than, of with one Wharton 1989). At least one species of Myosoma is SC+R1. Apex vein C+SC+R especially thickened bristle. Base of and gregarious. wing evenly densely Myosomatoides is being described here to make setose. Vein 2-1A absent. the name available to biocontrol workers in South Legs. Claws with large, acutely pointed, basal and Central America and for a forthcoming manual lobes. Telotarsi swollen in dorsal aspect compared to the New World genera of Braconidae (Marsh & to basal tarsal articles. Hind leg and often mid-leg Wharton, in preparation). (though less so) extremely laterally compressed, Terminology follows that of Achterberg (1979, tibia markedly expanded in lateral aspect (Figs. 1988). 1,2,4). Metasoma. Smooth, shiny, sparsely setose. MYOSOMATOIDES Quicke gen. nov. First metasomal tergite rather long, approximately 1.5-2.0 times longer than wide, and without cari- 1882 Type species: Myosoma pennipes Westwood, nae. Second tergite transverse, with — midlongitudinal carina. Third to seventh tergites Diagnosis. Superficially similar to Myosoma without transverse grooves subbasally or sub- that the 2nd metasomal is except tergite strongly posteriorly, without anterolateral areas, and mem- in the middle to form a pinched-up mid-longitudi- branous posteriorly. Hypopygium not extending In the hind femur and nal ridge (Fig. 4). addition, beyond apex of metasoma, pointed in lateral as- tibia are broad and of extremely laterally compressed pect. Ovipositor (part exserted beyond apex and have very long, dense setosity (Fig. 3). In metasoma) approximately two-thirds length of to the North Ameri- Quicke & Sharkey's (1989) key metasoma, with a pre-apical dorsal nodus and of will can genera Braconinae, Myosomatoides key apicoventral serrations. to Bracon from which it can be distin- Fabricius, Male genitalia. Digitus with two rather small, its hind femur guished by extremely compressed well separated apical teeth. Parameres not extend- and tibia. ing beyond base of digitus; setation restricted to a —Head. Antenna with Description. approxi- fairly narrow band at apex. Volsellae virtually Terminal mately 50 flagellomeres. flagellomere glabrous. Basal ring moderately produced wider than acuminate. Medial flagellomeres long. medioanteriorly, approximately as long as wide. Scapus small, shorter ventrally than dorsally in Internal anatomy. Rectum small with four not emar- lateral aspect, emarginate apico-laterally, circular rectal pads. Male with a pair of weak,

. with- ginate apicomedially Clypeus very shallow; pouch-shaped, intertergal gland reservoirs be- out a transverse carina reflexed ventral separating tween abdominal tergites 7 and 8, and with similar from dorsal dorsal part part; part punctate. Clypeus 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 2. Detail of hind 3. aspect. leg. 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 habitus and 7; abdominal tergum 8 without obvious glands. general simply cannot be excluded. Etymology.—Name based on the existing ge- represent convergence totally neric name Myoscma and the Greek suffix -oides As with Myosomatoides, members of the genus are of indicating close affinity; gender masculine. Myosoma parasitoids lepidopterous grass- and Sesiidae Distribution. —M. myersi: Argentina, Brazil, stem borers including Pyralidae British Guiana, Colombia, Ecuador, and Paraguay; (Maeto 1992). and Peru. M. pennipes:—Brazil (Amazonia) Biology. M. myersi nom. nov. is a larval para- ACKNOWLEDGMENTS sitoid of members of the pyralid borer genus

I would like to thank Tom Huddleston Natural Diatraea. Nothing is known of the biology of M. (The History Museum, London), Paul Marsh (USNM), Chris pennipes' (Westwood). 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- tific visit to funded the 1 Wings uniformly brown exchange Budapest by Royal Society, dr, a to and Mike Fitton M. muersi nom. nov. d by research grant jointly DLJQ (BMNH) from the NERC. — of the Wings largely hyaline with the apical third forewing brown (Fig. 1) LITERATURE CITED M. pennipes (Westwood) Achterberg, C. van, 1979. A revision of the subfamily Zelinae auct. (Hymenoptera, Braconidae). Tijdschrift voor DISCUSSION Entomologie 122: 241-479. Achterberg, C. van, 1988. Revision of the subfamily Blacinae Foerster (Hymenoptera, Braconidae). Zoologische nov. most re- Myosomatoides gen. closely Verliandelmgen, Leiden 249: 1-324. related to the L. W. sembles and is probably closely Agnew, C.W., Rodriguez-del-Bosque, A., Smith, J. Jr. Misidentification of Mexican stalkborers in the small and widespread, but principally New World 1988. Crambinae Folia genus Myosoma Brulle. Individuals of both genera subfamily (Lepidoptera: Pyralidae). Entomologica Mexicana 75: 63-75. have the hind femur rather strongly laterally com- Chippendale, G.M. 1979. The southwestern corn borer, and clothed with setae pressed densely (Mason Diatraea grandiosella: case history of an invading insect. in 1978): Myosomatoides differs from Myosoma hav- Bulletin of the Missouri Agricultural Experiment Station, no. 1-52. ing an even more strongly laterally compressed 1031, De Santis, L., Esquivel, L. 1966. Tercera lista de himenopteros hind femur and tibia (apomorphous) (Figs. 1, 2, 4), parasiticos y predatores de los insectos de la Republica and in having the 2nd metasomal tergite 'pinched- Argentina. RevistadelMuseo de La P/

1932. D. L. Myers, J. G. Biological observations on some Neotropi- Quicke, ]. 1990 Tergal and inter-tergal metasomal cal parasitic Hymenoptera. Transactions of the Royal En- glands of ma le braconine wasps (Insecta, Hymenoptera, tomological Society of London 80: 121-136. Braconidae). Zoologica Scripta 19: 413-423. H. P. A. Silveira A. D. L. Parker, L., Berry Guido 1953. Host- Quicke, J. Sharkey M. J. 1989. A key to and notes on the parasite and parasite-host lists of insects reared in the genera of Braconinae Hymenoptera: Braconidae) from South American Parasite Laboratory during the period America north of Mexico with descriptions of two new 1940-1946. Revistade la AsociaciondelngenieraAgronomica, genera and three new species. Canadian Entomologist Montevideo, no. 92: 1-101. 121:337-361. D. A. A. 1990. D. L. R. 1989. Pashley, P., Hardy T„ Hammond M, Mihm J Quicke, J., Wharton A. Myosoma nyanzaensis Genetic evidence for sibling species within the Sugar- sp.n. (Hymenoptera: Braconidae) parasitic on the stem- cane Borer (Lepidoptera: Pyralidae). Annals of the Ento- borer pest Chiio partellus (Swinhoe) (Lepidoptera: mological Society of America 83: 1048-1053. Pyralidae) in Kenya. Bulletin of Entomological Research D. L. 1985. Quicke, J. Two new genera of Braconinae (Insecta, 79: 1-5. Hymenoptera) from the Afrotropical Region. Zoologica Shenefelt, R. D. 1978. Hytnenopterorum Catalogus (nov. ed.). ScriptaU: 117-122. Part 15. Braconidae, vol. 10. The Hague, Junk, pp. 1425- Quicke, D.L.J. 1987. The Old World genera of braconine 1865. wasps (Hymenoptera: Braconidae). Journal of Natural Thomson, W. R. 1953. A catalogue of the parasites and predators History 21: 43-157. of insect pests. Section 2, host parasite catalogue. Part 2, D. L. 1988a. of hosts of Quicke, J. A new genus and species Braconinae the Hymenoptera (Agaonidae to Braconidae). (Hymenoptera: Braconidae) parasitic on Diatraca (Lepi- Commonwealth Institute of Biological Control, Ottawa, doptera: Pyralidae) in Ecuador. Bulletin of Entomological 190 pp. Research 78: 15-18. 1882. of Westwood, J. O. Descriptions new or imperfectly D. L. 1988b. the correct for - Quicke, J. Digonogastra: name known Ichneumones adsciti. Tijdschrift voor Entomologie Nearctic Iphiaulax of Authors. Proceedings of the Entomo- 25: 17-48. 90: 196-200. R. Smith L. logical Society of Washington Wharton, A., J. W. Jr, Quicke D. J., Browning H. W. D. L. 1988c. in the 1989. Quicke, J. Inter-generic variation male Two new species of Digonogastra Viereck (Hy- of the Braconinae genitalia (Insecta, Hymenoptera, menoptera: Braconidae) parasitic on Neotropical pyralid Braconidae). Zoologica Scrinta 17: 399-409. borers (Lepidoptera) in maize, sorghum and sugarcane. D. L. 1989. for Quicke, J. Further new host records genera and Bulletin of Entomological Research 79: 401 -410. species of Braconinae (Hym., Braconidae). En tomologist's Monthly Magazine 125: 199-205.

J. HYM. RES. Vol. 3, 1994, pp. 233-239

Immature Stages of Aganaspis pelleranoi (Brethes) (Hymenoptera: Cynipoidea: Eucoilidae), a Parasitoid of Ceratitis capitata (Wied.) and Anastrepha Spp. (Diptera: Tephritidae).

Sergio M. Ovruski

Centre de Investigaciones para la Regulation de Poblaciones de Organismos Nocivos (CIRPON), Pasa|e Caseros 1050, C.C.C. 90, S. M. de Tucuman (4.000), Argentina.

— Abstract- The embryological-larval development of Aganaspis pelleranoi (Brethes), a larval parasitoid of Ceratitis capitata (Wied.) and Anastrepha spp. is described and illustrated. A. pelleranoi possesses a stalked egg, eucoiliform first- and second- instar larvae, and hymenoptenform third- and fourth-instar larvae, followed by a prepupal and pupal stage. The duration of each stage at 25-26°C was as follows: egg, 3-4 days; first-, second-, and third-instar larvae, 2-3 days each; fourth-instar larva, 3-4 days, prepupa, 1-2 days; , 9-14 days.

Aganaspis pelleranoi (Brethes) is a parasitoid of MATERIALS AND METHODS Ceratitis capitata (Wied.) and several Anastrepha species (Diptera: Tephritidae) in Argentina (De Aganaspis pelleranoi was reared on last instar Santis that 1965), attacks the final larval instar of larvae of C. capitata in the laboratory as described these fruit (Turica and Mallo 1961). Aganaspis by Ovruski (in press). The host larvae were reared pelleranoi has also been reported from Lonchaea on carrot/corn meal/yeast medium as described spp. (Diptera: Lonchaeidae) in Brazil (De Santis by Nasca (1977). 1980). The life cycle and immature stages of A. 1941 From to 1945, A. pelleranoi was released pelleranoi were studied by exposing 20 to 30 C. as a control agent of fruit flies in Tucuman prov- capitata larvae to individual parasitoid females. ince, Argentina (Nasca 1973). In 1942 this species The rearings and experimentation were conducted was introduced and reared in Peru for the control at 25-26°C, 70-80% RH, and 12:12 LD photoperiod. of A. fratercuhts (Wied.) (Clausen 1978). Host larvae and host puparia were removed at Ovruski (in press) carried out studies on host intervals of 24 h after parasitism and dissected in detection behaviour in this fruit fly parasitoid. Insect Ringer's solution on depression slides. The Otherwise little is known of the biology of A. parasitoid larvae were removed from each host pelleranoi, and descriptions of the immature stages puparium and preserved in 70% ethanol for later of this have not been In species published. fact examination by light microscopy. Eight fourth- very little at all is published on the development of instar larvae were refixed in 4% glutaraldehyde eucoilid parasitoids of fruit flies, except that and 0.2 M phosphate buffer, and dehydrated in 35- Clausen et al. (1965) briefly described the imma- 50-70-90-100% ethanol. Thereafter they were placed ture stages of A. daci (Weld) (=Trybliograplia daci in 100% acetone. These specimens were critical Weld), a parasitoid of Dacits dorsalis Hendel, the point dried before examination by scanning elec- oriental fruit fly. tron microscopy (SEM). Descriptions are based on In this paper, studies on the life cycle and pre- several specimens. Thirty-four eggs in different imaginal development of A. pelleranoi are pre- stages of development were mounted on slides sented. and examined by light microscopy. Six first-, five 234 Journal of Hymenoptera Research

five formed. stalk 4 second-, and third-instar larvae were mounted The disappears (Fig. ) and the size and examined by light microscopy. Eight fourth- of the egg is similar to the 48 h old egg. Eclosion of instar larvae were observed with SEM and fifteen the first-instar larva occurs after 78-80 h. The fourth-instar larvae were examined by light mi- chorion is broken through anteriorly by the man- croscopy. Six females of A. pelleranoi of different dibles and posteriorly by movements of the tail dissected in Insect solution to 5 ages were Ringer's (Fig. ). observe the ovarian mature eggs. The terms used for morphological structure follow Kopelman and LARVAL DEVELOPMENT Chabora (1984) and Evans (1987). The drawings were made with a camera lucida. Data are pre- After hatching, the first-instar larva actively sented as range and /or mean. Specimens, in etha- feeds on the internal tissues of the host. For this nol and mounted, are deposited in the Miguel purpose, the mandibles are protruded from the Lillo Institute Fundation, S. M. de Tucuman Ar- oral cavity, although it was not possible to verify gentina. their use in feeding by direct observation. During this stage, the tail becomes less prominent and the LIFE CYCLE head wider. The second-instar larva appears on the fifth to Females of A. pelleranoi deposited a single egg sixth day after oviposition, when the host pu- in the body cavity of the host larva and one para- parium is formed. The larva continues to feeds sitoid adult emerged from each host puparium. inside the host pupa. Under laboratory conditions, the period from The third-instar larva appears on the eigth to oviposition in the host larva to the emergence of a ninth day after oviposition. In this stage the para- parasitoid adult varied from 25 to 27 days in males sitoid larva partially emerges from the host pupa and from 26 to 30 days in females. Thus, males near the middle region, and it then feeds exter- tended to emerge somewhat earlier than females. nally within the host puparium. The parasitoid The egg stage inside the host larva lasted about 3- larva occupies both internal and external posi- 4 days. Observations of the postembryological tions in the host pupa and develops two respira- development showed the presence of four larval tory mechanisms: cuticular respiration through instars. The first, second, and third instar each submerged posterior parts of the body and tra- lasted 2-3 about days and the fourth instar 3-4 days cheal respiration through prothoracic spiracles. (9-13 days in total). The prepupal stage lasted 1-2 In the fourth-instar larva, which appears on days, and the pupal stage from 9 to 14 days. approximately the tenth day after oviposition, feeding continues externally on what remains of EMBRYOLOGICAL DEVELOPMENT the host pupa. The parasitoid larva is very slug- gish and occupies 3/4 of the host puparium. After The newly laid egg is 0.32 mm-0.56 mm long feeding is completed, the meconium is released. It - and 0.06 mm 0.10 mm wide; the stalk is 0.42 mm appears as a dark crust in the posterior apex of the - 0.73 and the is not distin- 14 mm long embryo yet puparium (Fig. ). guishable (Fig. 1). 24 h after oviposition inside the host larva, the DESCRIPTIONS OF IMMATURE STAGES parasitoid egg is 0.26 mm-0.48 mm long and 0.08 - 0.15 is in — mm mm wide. The stalk reduced length, Mature ovarian egg. (Fig. 1.) Stalked; total and the embryo, vitelline membrane and the length on average 0.60 mm; stalk 1 .2 X longer than chorion are well discernible 2 the latter 3.5 (Fig. ). egg body, portion X longer than wide; By 48 h after oviposition, the egg has length- chorion translucent and smooth. ened and swelled (0.52 mm-0.66 mm long and 0.24 - 0.26 and the stalk has First-instar — mm mm wide), decreased larva. (Fig. 6, 7). Length 0.93 mm still more 3 The of - 1.39 2.2 (Fig. ). embryo shows signs mm; X longer than wide; eucoiliform; body segmentation, a caudal portion and thoracic translucent; subcylindrical in shape with elongate appendages. distinct head; mouth surrounded by several oral After 72 h the parasitoid larva is completely papillae, inside with a pair of unidentate Volume 3. 1 994 235

1

I = Figs. 1-7. Aganaspis pelleranoi. 1, ovarian egg. St., stalk. Scale bar 0,12 mm. 2, egg at 24 h. after oviposition. ch., chorion; V.M., = = vitelline membrane; Em., embryo. Scale bar 0,12 mm. 3, egg at 48 h. after oviposition. Scale bar 0,12 mm. 4, egg at 72 h. after = = oviposition. Scale bar 0,12 mm. 5, egg at 80 h. after oviposition. Scale bar 0,12 mm. 6, first-instar larva. OP., oral papillae; = Th. App., thoracic appendages; V.P., ventral process; 77., tail; An., anus. Scale bar 0,12 mm. 7, mandibles of first- instar larva. Scale bar = 0,02 mm. 236 Journal of Hymenoptera Research

n.

Ml. G. F.

8-9. second-instar larva. oral = Figs. Aganaspis pelleranoi. 8, OP., papillae; 77., tail. Scale bar 0,25 mm. 9, third-instar larva. Sp. spiracle; Mt., mouth; C.F., globular fat; An., anus. Scale bar = 0,32 mm. 237

= fourth-instar larva. Lat. lateral anus. Scale 1 1 , anterior view Figs. 10-12. Aganaspis pelleranoi. 10, Sw., swelling; An. , bar 1,0 mm. = of head, fourth-instar larva. Ant. Or., antennal orbit; Ma., mandible; Mi., maxilla; La., labium; Lb , labrum. Scale bar 0,10 mm = lateral of fourth-instar larva. Lat , lateral Scale bar mm. 12, view body segment 6-8, Sp , spiracle; Sw swelling. 0,10

7 ventral surface with short tail and without ventral subtriangular mandibles (Fig. ); ment very of head with a prominent tubular projection; gut process; anus opening dorsally. easily discernible; each thoracic segment with a — pair of long slender ventral appendages; seven Third-instar larva. Fig. 9. Length 2.78 mm- well-defined abdominal segments; caudal segment 3.45 mm; 3 X longer than wide; more typically with a long tail and short ventral process; tail hymenopteriform; yellowish with several white bearing several small setae apically; ventral mar- globular fat particles in thorax and abdomen; cy- gin of 7th abdominal segment and basal end of lindrical in shape with slender posterior portion; caudal segment with scalelike ornamentation; tail broad head; small unidentate mandibles, difficult 6 X longer than ventral process; anus opening to discern; prothoracic segment with a single pair dorsally. of spiracles; ten distinguishable body segments; — brain and gut easily discernible; without tail and Second-instar lari'a. (Fig. 8). Length 1.53 mm ventral process; anus opening ventrally. - 2.31 mm; 2 X longer than wide; modified — eucoiliform; whitish yellow with some white Fourth-instar larva. Figs. 10, 11,12, 13). Length globular fat particles; cylindrical in shape with 3.48 mm-4.13 mm; 2 X longer than wide; short and fleshy head; mouth with external oral hymenopteriform; whitish yellow with many small papillae; unidentate mandibles; gut well discern- white globular fat particles dispersed throughout ible; without thoracic appendages; caudal seg- dorsolateral parts of thorax and abdomen; head 23S Journal of Hymenoptera Research 13

Me.

= Figs. 13-15. Aganaspis pelleranoi. 13, mandibles of fourth-instar larva. Scale bar 0,04 mm. 14, prepupa inside the host Scale bar = Scale bar = pupanum. Ph., pupanum; Pr., prepupa; Me , meconium. 1,0 mm. 15, pupa. 1,6 mm.

relatively large, subcircular in front view, without instar larva; body curved and reduced in size; setae; antennal orbits large; prominent bidentate occupying 2/3 of the host puparium. mandibles, 1.2 X broader than long basally, very - dark apically (figs. 1 1 and 13); labrum 3.2 X broader Pupa.— (Fig. 15). Length 2.68 mm 2.99 mm, than long, with rounded lateral sides and median initially white, later becoming dark; antennae 1,7 - depression on apical margin; maxillae circular in mm 2.1 mm in length, extending to first abdomi- shape and joined to labium; body very swollen; nal segment. only eleven body segments discernible; except prothoracic and terminal abdominal segment, all DISCUSSION segments with a pair of spiracles; each body seg- ment provided with lateral voluminous swellings Aganaspis pelleranoi produces a stalked egg, behind spiracles (Fig. 12); integument smooth and typical of the Eucoilidae. The transition from en- without setae; anus opening ventrally. doparasitic to ectoparasitic life has also been re- for other of Eucoilidae 1951 ported species (Jenni ; - Prepupa.— (Fig. 14). Length 2.55 mm 2.98 Wishart and Monteith 1954; Sychevskaya 1974; - mm; width 1.18 mm 1.56 mm; similar to fourth- Kopelman and Chabora 1984). The embryological Volume 3, 1994 239 and larval development is similar to other eucoilid Book, N° 480. 545 pp. Clausen, C. P., D. W. Clancy and Q. C. Chock. 1965. parasitoids: Tn/bliograplia rapae (Westwood) (James Biological control of the oriental fruit fly {Dacus dorsalis Hendel) 1928; Wishart and Monteith 1954); Hexacola sp. and the other fruit flies in Hawaii. Technical Bulletin Hexacola near Websteri (Simmonds 1952); sp. United States Department of Agriculture, N" 1322. 102 pp. (Crawford) (Eskafi and Legner 1974); Leptopilina De Santis, L. 1965. Nota sobre un parasito de la mosca sudamencana de la fruta boulardi (Barbotin et al.) (Kopelman and Chabora (Hymenoptera: Cynipidae). Rei'ista de la Soaedad Entomologica Argentina 27 (1-4): 73- 1984); Eucoila frichopsila (Hartig) (Sychevskaya 74. 1974) and Kleidotoma Huz. japonica (Huzimatu De Santis, L. 1980. Catalogo de los himenopteros brasilenos 1940). de la Sene Parasitica, incluyendo Bethyloidea. Clausen et al. (1965) reported three larval in- Universidad Federal do Parana, Curitiba. 395 pp. Eskafi, F. M. and E. F. Legner. 1974. of immature stars in A. daci . This study proves the presence of Descriptions stages of the cynipid Hexacola sp. near Websteri four larval instars in A. pelleranoi. The first-, sec- (Eucoilinae: Hymenoptera), a larval-pupal parasite of ond-, and fourth-instar larvae described here are Hippelates eye gnats (Diptera: Chloropidae). The Cana- similar to the first-, second-, and supposed third- dian Entomologist 106: 1043-1048. Evans, H. E. 1987. and instar larvae of A. daci respectively. The third- Cynipoidea. Cynipids, others, pp. 665-66. In Stehr, F. W., ed. Immature insects. instar larvae of A. pelleranoi is a transition between Kendall Hunt, Dubuque, Iowa. 754 pp. the eucoiliform and the type hymenopteriform K. 1940. The life of a Huzimatu, history new cynipid fly, This is a type. stage distinguished by single pair of Kleidotoma japonica, n. sp. Science Report, Toboku Imperial Sendai 15: 457-480. open spiracles, in the prothoracic segment, as in University, James, H. C. 1928. On the life histories and economic status of the supposed third stage of T. rapae (Wishart and certain parasites of dipterous larvae, with descriptions Monteith 1954), the supposed fourth stage of £. of some new larval forms. Annals of Applied Biology 15: frichopsila (Sychevskaya 1974) and the supposed 287-316. fourth stage of L. boulardi (Kopelman and Chabora Jenni, W. 1951. Beitrag zur Morphologie und Biologie der Pseudeucoila bocliei eines 1984). Further comparative work would be needed Cynipide Weld, larvenparasiten von Acta 32: 177- to confirm these differences between Drosophila melanogaster Meig. Zoologica apparent 254. as intermediate instars of larvae species, apocritan Kopelman, A. H. and P. C. Chabora. 1984. Immature stages of are notoriously easy to overlook. Leptopilina boulardi (Hymenoptera: Eucoilidae), a Furthermore Clausen et al. (1965) stated that protelean parasite of Drosophila spp. (Diptera: Drosophilidae). Annals of the the first-instar larva of A. daci does not possesses Entomological Society of America 77(3): 264-269. discernible mandibles, but the first-instar larvae of Nasca, J. A. 1973. Par^sitos de "moscas de los frutos"

A . has a of well unidentate pelleranoi pair developed establecidos en algunas zonas de Tucum^n. Revista mandibles with their ends projecting from the oral Agricola del Noroeste Argentino 10 (1-2): 31-43. Nasca, A. 1977. Estudios relacionados con las cavity, similar to the first-instar larvae of £. J. posibilidades del Control Integrado de las moscas de las frutas. trichopsila (Sychevskaya 1974). Universidad Nacwnal de Tucumdn, Informe CAFPTA, N" Prominent bidentate mandibles in last instar 529. 61 pp. larvae have also been reported for A. daci (Clausen Ovruski, S. M. In press. Comportamiento en la deteccion del en et al. 1965), Trybliograplia Forster and Hexacola huesped Aganaspis pelleranoi (Brethes) (Hymenoptera: Eucoilidae) Cera fitis Fbrster (Evans 1987). Cynipoidea, parasitoidedelarvasde capitata (Wied) (Diptera: Tephritidae). Revista de la Sociedad Entomologica Argentina 53. ACKNOWLEDGMENTS F. 1952. Simmonds, J. Parasites of the frit-fly, Oscinellafrit (L.), in eastern North America. Bulletin Entomological Re- search 43: 503-542. I am especially grateful to Dr. Patricio Fidalgo (CIRPON, Sychevskaya, V. L. 1974. The biology of Eucoila Argentina), Dr. Abraham Willink (CONICET-UNT, Argen- trichopsila Hartig (Hymenoptera, Cynipoidea), a parasite of the tina), and Dr. Goran Nordlander (Swedish University of larvae of synanthropic flies of the family Sarcophagidae Agricultural Sciences, Uppsala, Sweden) for suggestions and (Diptera). Entomological Review 53: 36-44. critical reviews of the manuscript Turica, A. and R. G. Mallo. 1961. Observaciones sobre la poblacion de las "Tephritidae" y sus endoparasitoides LITERATURE CITED en algunas regiones citricolas . IDIA 6: 145- 161. Wishart, G. and E. Monteih. 1954. (Westw.) Clausen, C. P. 1978. Tephritidae, pp. 320-334. In Clausen, C. Trybliograplia rapae (Hymenoptera: Cynipidae), a of P., ed. Introduced Parasites and Predators of arthropod parasite Hylemya spp. (Diptera: Anthomiidae). The Canadian 86: pests and weeds: a World Review. U.S. Agriculture Hand Entomologist 145-154.

J. HYM. RES. Vol. 3, 1994, pp. 241-277

Phylogenetic Implications of the Mesofurca and Mesopostnotum in Hymenoptera

John M. Heraty, James B. Woolley and D. Christopher Darling

(JMH) Biological Resources division, CLBRR, Agriculture Canada, C.E.F., Ottawa, Ontario, Canada Kl A 0C6 (present address: Department of Entomology, University of California, Riverside, California, USA 92521); (JBW) Department of Entomology, Texas A&M University, College Station, Texas, USA 77843; (DCD) Department of Entomology, Royal Ontario Museum, Toronto, Ontario, Canada M5S 2C6

— Abstract. The skeleto-musculature of the mesofurca and the mesopostnotum is examined in Hymenoptera. Character systems based on internal structure of the mesothorax support recent hypotheses that suggest sawflies are paraphyletic with respect to . Unique character states for Hymenoptera include the presence of two mesofurcal-laterophragmal muscles, a mesofurcal-third basalare muscle, and a scutellar-metanotal muscle. Other possible apomorphies include the medial emargination of the mesopostnotum and the formation of anterior furcal arms. The arrangement of mesofurcal muscles that attach to the and the profurca laterophragma are described and interpreted in light of recent phylogenetic hypotheses. in attachment Changes sites, fusion or loss of the anterior arms of the mesofurca and features of the laterophragma provide characters that are consistent with + the monophyly of Tenthredinoidea (Cephoidea + (Siricoidea (including Anaxyelidae) + + (Xiphydriidae Orussoidea + Apocrita))). Groundplan states for the Apocrita are proposed that include retention of a mesofurcal bridge, retention of an anterior process on the bridge that supports the interfurcal muscles, reduction of the mesofurcal-laterophragmal muscles from two to one, retention of the mesotergal-laterophragmal muscle, loss of the mesofurcal-third basalare muscle, and loss of the metafurcal-spina muscle. Within Apocrita the distribution of character state is less informative than in changes Symphyta, but provide evidence for relationships of some taxa. The mesofurcal bridge is lost convergently in Ceraphronoidea, Pelecinidae, Platygastroidea, Mymarommatoidea, Mymandae and some Chalcidoidea. The tergal-laterophragmal muscle and associated posterior lobe of the laterophragma are postulated to have been lost in nine of independently lineages Apocrita. The development of the laterophragma into an axillary lever is a synapomorphy for Vespoidea and Apoidea, and in Apiformes the lever is an independent sclerite. The distribution of states for 12 characters is discussed for 62 families of Hymenoptera. Parsimony analysis of these data result in trees that generally agree with the current hypotheses for Symphyta but not for Apocrita.

INTRODUCTION 1969, 1980,Matsuda 1970,Shcherbakov 1980, 1981, Gibson 1985, Johnson 1988, Whitfield et al. 1989). "Students of these [hymenopteran] parasites discover that The determination of homologous structures and the thorax valuable characters for the determination presents polarity of characters in Hymenoptera are crucial and classification of but are r > species, they handicapped by the _• ,, .• i .• i for a lack of reliable studies on the structure of the thorax" understanding phylogenetic relationships, to lc recentl addressed various authors Snodgrass 1910 p 37 P y by (Rasnitsyn 1969, 1980, 1988, Brothers 1975, Konigsmann 1977, 1978a 1978b, 1986, Since Snodgrass (1910) first attempted to ex- Carpenter Brothers ^d Carpenter 1993). pand our knowledge of the structure of the hy- The mesofurca is an of the ster- menopteran thorax, additional studies have de- invagination scribed num mto the thorax that forms a central oint of the skeleto-musculature of single species P attachment f°r the ventral muscles, or single families (Weber 1925, 1927, Tulloch 1935, longitudinal the muscles, the coxal and trochant- Maki 1938, Duncan 1939, Michener 1944, Bucher sterno-pleural

eral muscles ' and the muscles 1948, Alam 1951, Saini et al. 1982, Daly 1964, mesopostnotal

( Kelse 1957 < Matsuda The mesofurca is Gibson 1986, 1993). Fewer studies have compared Y 1970). com nsed of a basal thoracic structures among families of Hy- P plate (discrimenal lamella) that rises vertically f™ the discrimen, slopes menoptera (Snodgrass 1942, Daly 1963, Rasnitsyn posteriorly to the furcal base, and divides dorsally 242 Journal of Hymenoptera Research into two lateral arms, termed the sternal apophy- MATERIA! S AND METHODS ses (Snodgrass 1927, Chapman 1992, Lawrence et al. 1992). The mesopostnotum is one of the pri- Terms for structures and muscles generally mary dorsal sclerites involved in flight through follow Snodgrass (1910, 1942), Daly (1963), Gibson the posterior inflection of the antecosta (second (1985, 1986, 1993) and Ronquist and Nordlander phragma) which forms the posterior attachment (1989). Muscles were identified using the systems of the longitudinal flight muscles. Recent studies proposed by Kelsey (1957) and Daly (1963) (Table involving skeleto-musculature of the hy- 1). Figures 1 and 2 are used to place the skeleto- menopteran thorax have focused on the pleural musculature within the context of the mesosoma. attachments (Shcherbakov 1980, 1981, Gibson 1985, Muscles and stuctures are extensively labeled in 1993), the extrinsic musculature of the mesocoxa Figs. 3 and 4. The Kelsey system uses a fixed set of (Johnson 1988), and the development of the numbers and is useful for comparisons across the metapostnotum (Whitfield et al. 1992). Rasnitsyn Endopterygota. Daly's system is preferred for clar- (1969, figs. 187-194) was the first person to com- ity because the insertion-origin of attachment sites pare the different skeletal structures for the are readily identified and new muscles can be mesofurca of 8 families of Symphyta. His illustra- added to the system; for example, the new muscle tions show the transformation series for was the abbreviation fbl for the Symphyta fu,-ba 3 given that are discussed in this paper. Rasnitsyn (1988) Kelsey system as it could not be assigned a nu- refers to the furca for features supporting meric value that would signify its relative position Tenthredinoidea and for Cephoidea + Siricoidea + to other muscles in the mesothorax. Terms pro- Apocrita. Similarly, Snodgrass (1942) presented a posed by Matsuda (1970) are comprehensive and pictorial evolutionary history for development of may be referenced across orders of insects; how- the axillary lever of Apoidea. This work expands ever his abbreviated system is difficult to use and upon these initial studies and extends the com- is not followed here. parative aspects of these works to include most Several new classifications of families within families of Apocrita. Hymenoptera have been proposed recently that This study of the mesofurca and differ largely in placement of certain families as mesopostnotum began as an attempt to under- separate superfamilies, families, or subfamilies. stand the polarity and homology of mesofurcal We follow the classification of Huber and Goulet structures and muscles of Aphelinidae (1993), as it represents the most current synthesis (Chalcidoidea) and the phylogenetic implications of information across the order. of these attributes within the Chalcidoidea. Even- Dissections were based on specimens pre- the entire to tually Hymenoptera needed be sur- served in 70%ethanol or initially fixed in Dietrich's veyed to resolve what we initially thought were or Kahle's solution and then transferred to etha- In relatively simple questions. this study, all nol. All specimens were critical point-dried prior muscles attaching to the mesofurca and to dissection. The mesosoma of Monomachns mesopostnotum are identified and compared to (Monomachidae) was rehydrated using Barber's homologous muscle groups in Neuropterida and solution, transferred through increasing concen- Mecopterida (sensu Kristensen 1992), as they are trations of ethanol to 98% and then critical point- considered to be phylogenetically close to Hy- dried. For each dissection, the mesosoma was menoptera (Kristensen 1992), and have a mesotho- anchored onto a standard SEM stub using chloro- rax is which structured similar to the Symphyta. form-based silver paint. Dissections were made Within Hymenoptera, we have concentrated our using hooked minuten pins or fragments of razor analysis on the skeletal structure of the mesofurca blades. Dried haemolymph and extraneous tis- and mesopostnotum, and on the muscles attach- sues were removed from dissections using small ing between the thoracic furcae and the amounts of glue obtained by dragging a hooked of the laterophragma mesopostnotum. The evi- minuten pin across clear sticky tape (Gibson 1985). dence the provided by mesofurca and Exemplar taxa were chosen to represent the mesopostnotum for relationships within the maximum variation within taxa. In some groups Chalcidoidea will be discussed in a subsequent (e.g. Apoidea), there was virtually no variation; within paper. whereas some taxa (e.g. Diapriidae) both Volume 3. 1994 243 structure and presence of muscles varied and more + remaining Hymenoptera), the anterior furcal genera were dissected to characterize this varia- arms are either absent, short, or elongate and well tion. Our primary concern was for establishing separated along their entire length. The anterior groundplan states for higher taxa, although arms, or the equivalent region on the lateral arms autapomorphies are discussed. The taxa exam- of the mesofurca, form the posterior attachment ined for internal characters are listed in Table 2. sites for the ventral intersegmental muscles (muscle Numerous Chalcidoidea were also dissected as and muscle With fusion 124, fiij-fUj, 127, fu-rsps,). of a part comprehensive study of the mesofurca in of the anterior arms, the furca separates the gastric that superfamily. Representative dissections are and nervous systems and the ventral nerve cord housed at the Canadian National Collection passes through the foramen bounded by the (CNCI), Royal Ontario Museum (ROM) and Texas mesofurcal bridge and the lateral arms of the A&M University (TAMU). When possible, con- mesofurca. specific adults of the dissected specimens are de- In most Neuropterida and Mecopterida, the posited as voucher specimens in the above collec- interfurcal muscles attach directly to the anterior tions. The majority of specimens were obtained face of the lateral arms of the mesofurca (state 0, no from the CNCI alcohol collection. arms) (Snodgrass 1927, Kelsey 1957, Matsuda 1970). The In mesofurcal-mesopostnotal complex was Xyelidae (Fig. 3) and Pamphiliidae (Fig. 6) (and broken into 12 characters with a total of 36 up probably Megalodontidae, cf. fig. 187, Rasnitsyn character states. Seventy hymenopteran taxa were 1969), the interfurcal muscles attach to anterior scored based on the of (Appendix 2) examination projections (af, anterior furcal arms) that are long, internal characters for 119 species (Table 2). A robust, and separated along their entire length taxon was scored single outgroup based on dissec- (state 1). Of the outgroup taxa examined, only tions of 5 families of Neuropterida and 3 families Brachynemurus (Myrmeleontidae) have anterior of Mecopterida. Characters 1, 3 and 9 are postu- arms (state 1) similar to Xyelidae, and thus their lated as unique characters for Hymenoptera; in presence could be plesiomorphic or apomorphic each of these cases, the outgroup is coded as a for Hymenoptera. Because anterior arms are unique character state (state 0). The state value "?" present only in a derived member of the outgroup, was used to denote uncertain homology, not miss- we consider the presence of long anterior arms in ing data. Xyeloidea and Megalodontoidea to be apomorphic Illustrations were made with a camera lucida. for Hymenoptera. of Outlines muscles present but not illustrated are In Tenthredinoidea, the anterior arms (af) can dashed lines. represented by Some muscles not be reduced in size (Cimbicidae Fig. 8, central to this study were not consistently figured Blasticotomidae Fig. 20, and Nematus muscle 180 for and caution should (e.g. Symphyta) (Tenthredinidae) Fig. 21), modified into support- be exercised in additional deriving interpretation ing cup-like structures (Diprionidae, Fig. 7), or lost the from illustrations. Abbreviations referring to entirely so that the interfurcal muscles attach to muscles are circled in all skeletal charac- the anterior face of the lateral furcal arms figures, ( Argidae, ters are not. The mesofurcal-mesopostnotal com- Pergidae Fig. 5, and most Tenthredinidae). plex is abbreviated as MF-MPN complex. Rasnitsyn (1988) treated "fore arm short" as a character state (his 2-3d) supporting Tenthredinoidea and "fore arm reduced" (his 6- RESULTS AND DISCUSSION 8a) for Argidae + Pergidae. Short arms, as in Blasticotomidae (Fig. 20), are probably Character Analysis plesiomorphic for Tenthredinoidea, and further modifications of the arms or complete loss are Character 1. Mesofurcal bridge derived within Tenthredinidae. It is possible to The most significant modification of the code for several different character states within mesofurca in is the fusion of the Hymenoptera Tenthredinoidea (small, absent, cup-shaped, etc.), anterior arms into a elongate (af) mesofurcal bridge but this would only introduce unnecessary ho- (fb) that characterizes all In virtually Apocrita. moplasy into the analysis (i.e. "absence" derived 2 basal groups of Symphyta (excluding Cephoidea or more times) or a series of autapomorphic char- 244 Journal of Hymenoptera Research

and mesofurca examined in and Table 1. Homology and terms for muscles of the mesopostnotum Neuroptera Hymenoptera Letter m added to distinguish metathoracic muscles. Muscle fbl described in text.

Kelsey 1957— Con/dalus Hymenoptera

No Name Matsuda Daly Johnson Present Notes 1970 1964 1988 designation

Mesothoracic dorsal indirect muscles muscle both 112 internal (ventral] longitudinal tl4 lph-2ph lph-2ph longitudinal flight Kelsey (1957) recognized ventral (internal, 112) and dorsal (external, 113) muscles Recognition of the two muscles in Hymenoptera is unnecessary

and attached in most 112m metathoracic longitudinal tl4 2ph-3ph 2ph-3ph reduced laterally Hymenoptera

in medial muscles from meta- 114 scutellar-metanotal tl3 t,-t, t,-t, Hymenoptera, paired passing external in notum (t,) to scutellum over mesopostnotum (PN,), some sawflies From PN in ; Neuroptera

on to anterior face of 116 second dorsal diagonal t,-2ph t,-pn, dorsomedial attachment mesoscutum to axillar laterophragma (pn,), attachment dorsal surface where transscutal arhculahon present Mesothoracic ventral muscles lateral face of to anterior face or arms of 124 mesothoracic interfurcal sl3 fu,-fu, fu,-fu, from fu, 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

of to of lateral arm of 127 mesofurcal-spina sl4 fu,-lsps fu,-sps, spina profurca apex fu,.

interfurcal sl3 • face of lateral arms of fu to (1957) treats as 181 metathoracic fuj-fu, fu,-fu, posterior ; fu, Kelsey three muscles (181-183) Hymenoptera have maximum of two

mesofurca to of lateral arm of 180 metafurcal-spina sl4 fUj-spSj spina of apex fu,

Tergopleural muscles 7 t face of = to of basalare 137 laterophragmal-basalare t-p8 -ba, posterior laterophragma ( t,) apodeme (ba,) Synonomy with t-p8 is questionable Furcal muscles attachment on lateral arm of to anterior of 150a posterior furcal-laterophragmal t-sl fu,-2ph fu,-pn, posterior fu, process laterophragma (ap)

150b anterior furcal-laterophragmal t-sl fu,-pn, attachment anterior to 150a on anterior or lateral arm of fu, to posterior lobe of laterophragma (pn,)

lateral surface of to 151 furcal-pleural arm p-sl pl,-fu, pl,-fu,_ fu, pleural ridge

arhculation-furcal s-cx2 lateral surface of fu to coxal on 170 coxal pl,-fu, t 2 process pleuron

fbl furcal-basalare fu 7 of lateral arm of fu to basalare be p-s3? 2-pl 3 fu,-ba, apex 2 (ba,), may homologous with pupal muscle of Apis (Daly 1964), well developed with dorsal cap and apodeme in Dipnon (one preparation where ba, dissected with fu and both attached). ; -ba, pl,-ba, Possibly homologous with Matsuda's p-s3 from furca to anterior margin of succeeding episternum

Walking and indirect flight muscles anterior furcal-coxal s-cx6 fu-cx anterior base of fu to median nm of coxa. 169 fu,-cx fu,-cx ;

173 furcal-coxal s-cx3 fu -cx fu-cx fu -cx. attachment on to muscle 169 to rim of posterior 2 2 fu, posterior posteromesal

174 furcal of trochanter s-trl fu,-tr, anterior base of fu (or arms) to trochanteral depressor hy h", ; apodeme Volume 3, 1 994 245

Table 2. Taxa dissected for study of the mesofurcal-mesopostnotal complex.

Section: NEUROPTERIDA Trigonalyoidea Mantispidae Mantispa sp. Trigonalyidae Orthogonalys pulchella (Cresson) Chrysopidae Chrysopa sp. Corydalidae Corydalus sp. Ceraphronoidea Myrmeleontidae Brachynemurus sp. Megaspilidae Megaspilus fuscipennis (Ashmead), Rhaphidiidae Rhaphidia sp. Trichosteresis sp. Ceraphronidae Ceraphron sp. Section: MECOPTERIDA

Bittacidae Bittacus sp. Evanoidea Meropeidae Merope tuber Newman Aulacidae Prislaulacus strangahae (Rohwer) Panorpidae Panorpa sp. Evaniidae Hyptw thoracica (Blanchard), Evania sp. Gasteruptiidae Gasteruptwn sp. Order: HYMENOPTERA SYMPHYTA Ichneumonoidea

Xyeloidea Braconidae Alysiinae: Coelwius sp.; Hybrizontinae: Xyelidae Macroxyelinae: Macroxyela ferruginea Hybrizon sp.; Macrocentnnae: Pleuroneura (Say); Xyelinae: sp., Xyela Macrocentrus sp.; Miscogastennae: minor Norton ?. Apanteles sp.; Rogadinae: genus Ichneumonidae Ephialtinae: genus ?, Scambus sp., Megalodontoidea Megarhyssa sp.; Ophioninae: Enicospilus Pamphilndae Cephalciinae: Acantholyda sp., sp. Pamphiliinae: Pamphilius sp. Chrysidoidea Plumarndae Plumana sp. Tenthredinoidea Sclerogibbidae Probethylus sp. Blasticotomidae Blasticotoma sp. Embolemidae Etnbolemus nearcticus (Brues) Tenthredinidae Heteranthinae: Profenusa canadensis Dryinidae Anteoninae (female), Gonatopodinae (Marlatt); Nematinae: Nematus sp.; (male) Selandriinae: Aneugmenus flavipes

(Norton), Strongylogaster fflcirn(Norton); Bethylidae Anisepyris sp., Epyris sp. Filacus Tenthredininae: sp., Macrophya Chrysididae Amiseginae: Adelphe anisomorphae sp. Krombein; Chrysidinae: Chrysis sp., similis Dipnonidae Dipnoninae: Diprion (Hartig) Parnope sp.; Cleptinae: Cleptes sp. Cimbicidae Cimbicinae: Cimbex americana Leach; Ambnnae: Zaraea americana Cresson Vespoidea Acordulecerinae: Acordulecera Pergidae sp., Tiphiidae Myzinum sp. Imexicana Smith Syzygoninae: Lagideus Sapygidae Sapyga sp. Argidae Arginae: Arge sp., Durgoa matogrossensis Mutillidae Sphaeropthalminae (males, 3 genera) Mai. Sierolomorphidae Sierolomorpha canadensis Provancher Pompilidae Pepsinae: Calicurgus hyalinatus Fabr.; Cephoidea Pompilinae: Aporinella galapagensis cinctus Cephidae Cqjhus Norton, Rohwer, Aporus sp. trimaculata Hartigia (Say) Rhopalosomatidae Rhopalosoma sp. Bradynobaenidae Bradynobaenus sp. Siricoidea Formicidae Myrmicinae: Solenopsis invicta Buren Anaxyehdae Syntexis libocedrii Rohwer (queen & worker); : Siricidae Siricinae: Urocerus albicornis (Fabricius); Camponotus planus F. Smith (queen), Tremicinae: Tremex columba (Linnaeus) Paratrechina sp. (queen) Scoliidae Scolia sp. Xiphydrioidea Vespidae Eumeninae: Odynerus sp., abdommalis Xiphydriidae Xipiudria Say Parancistrocerus sp.; : Dolichovespula sp., Vespula sp. Orussoidea Orussidae Orussus terminalis Newman Apoidea

Crabronidae Ectetnnius sp., Larra sp. APOCRITA Heterogynaeidae Heterogyna sp. Stephanoidea Stephanidae Megischus bicolor (Westwood)

continued on next page 246 Journal of Hymenoptera Research

Table 2 continued arm are joined (state 3). The anterior arms are considered to be homologous with those of Andrenidae Andrena sp. Xyeloidea and Megalodontoidea; therefore, state Anthophondae Nomada sp., Ceratma sp. 3 is probably derived from state 1. Rasnitsyn (1988; Apidae Apis mellifera L., Bombus sp., Trigona sp. his character includes with Halictidae Agnpostemon sp., Laswglossum sp. 19b) Cephoidea + fore- Megachilidae Megachile sp. Siricoidea Vespina based on having the arms of the mesofurca long and fused for some Proctotrupoidca distance. Although true for Siricoidea + Vespina, Diaprndae Ambositrinae: Dissoxylabis sp.; the arms are only apically fused in Cephidae. Belytinae: Aclista sp., Acropiesta sp., In and Siricidae Aneurrynchus sp., Oxylabis sp., Anaxyelidae (Fig. 10) (Fig. Diaprunae: Coptera sp., Paramenia sp. 11), the anterior arms are elongate and fused along Spilomicrus sp., Trichopna sp. most of their length (state 4). The anterior arms of Monomachidae Monomachus sp. Siricidae are laterally flattened (Fig. lib) and in Vanhorniidae Vanhornia eucnemidarum Crawford dorsal view each arm can be distinguished along Helondae Helorus sp. its entire 11a). have a Ropronhdae Roprorua sp. length (Fig. Anaxyelidae vera Proctotrupidae Exallonyx sp., Mwta (Fouts) similar structure, including anterior placement of Pelecimdae Pelecinus polyturator (Drury) muscle 150b, but fusion of the arms is more com- furcal plete (Fig. 10a). Posteriorly, the anterior Cynipoidea arms of Anaxyelidae are connected by a thin hori- Ihalndae Jbfl/ifl sp. 7 zontal of cuticle. Eucoihdae genus plate AUoxvstidae Alloxysta sp. InXiphydrndae(Fig. 12), theanteriorarmsare completely fused and form a transverse mesofurcal Platygastroidea bridge (state 5), and the entire dorsal and lateral Scelionidae Scehonmae: y4n/cris sp., Calotelea sp., surfaces form the posterior attachment for muscle Oemasfobaeus sp., Gryon sp. (wingless), of 150a and 150b to the Macrcteleia absona Muesebeck, Sparasion 124. Displacement muscles Teleasinae: Tnmorus fusion of the in sp., sp. lateral arms suggests that arms Inostemmatinae: Inoslemma Platygastndae hocybus sp., Xiphydriidae may be independent of the fusion in s . P Siricidae and Anaxyelidae. Orussidae exhibit fusion of Mymarommatoidea (Fig. 13) complete the anterior arms into a smooth and bowed Mymarommatidae Palaeomymar sp mesofurcal bridge (state 6) with a strong median Chalcidoidea dissections for work in (additional progress anterior process that is the attachment site for not listed) muscles 150 and 124, as occurs in Anaxyelidae and Mymandae Gonafoccrus sp. Siricidae. Pteromalidae Cleonvminae: Cleonymus sp., Oodera sp. The groundplan states for the mesofurca of Apocrita consists of 1) a complete mesofurcal acter states. of character states must in Coding part bridge (as in Xiphydriidae and Orussidae), 2) an reflect the level of At a different level, for analysis. anterior medial projection supporting muscle 124 in an of the relationships of in and and example analysis (fu^-fUj) (as Xiphydriidae Orussidae), Tenthredinoidea that is associated with a more of 150a 3) lateral displacement muscle (fu,-pn, ) extensive of taxa, to survey itmightbe appropriate (as in Xiphydriidae). No Apocrita have muscle 150 further the various These modi- partition shapes. originating on the anterior projection of the fications are difficult to characterize and here we mesofurca as in Orussidae. Because of the lateral have combined them into one apomorphic state, displacement of muscle 150a, which is similar to the reduction of anterior arms well-separated (state Xiphydriidae (see character 2), Apocrita with a 2). mesofurcal bridge are coded as character state 4. A clear transformation series to the leading The mesofurcal bridge is absent (state 7) in of the mesofurcal is found development bridge Ceraphronoidea (Fig. 25), Pelecinidae (Fig. 31), within the ta with the Symphy beginning Cephidae. some Chalcidoidea (including all Mymaridae), In the anterior arms are narrow Cephidae (Fig. 9), Mymarommatidae, and Platygastridae (Fig. 28). and and the extreme of each elongate only apices We consider that absence of the bridge is an Volume 3, 1 994 247 apomorphic loss of the type of bridge found in muscles 150 and 116, and the loss of either of these Xiphydriidae (state5), which is similar to the type muscles is associated with a corresponding change found in most Apocrita. Assuming that presence in structure of the laterophragma. To avoid dupli- of a bridge is a groundplan state for Apocrita cation of character coding, the presence or absence (whether states 4, 5 or 6), losses within each of of certain structures of the laterophragma are these taxa are considered irreversible (unlikely treated under other characters. For example, the that a bridge can be regained) and probably inde- anterior apodeme is lacking in Neuropterida and pendent. In Ceraphronidae, Pelecinidae, Mecopterida but this was not coded as a different Mymarommatidae, Platygastridae, and some state for this character because it is reflected in the Mymaridae, the mesofurca is lyre-shaped with the differentiation of muscle 150a in Hymenoptera, lateral arm terminating in a cup-shaped process which is dealt with as Character 3. A broad that supports muscle 150a. When the furcal bridge laterophragma with a small anterior apodeme is lost in Chalcidoidea (Aphelinidae, Encyrtidae, (state 0) occurs in Xyelidae, Pamphiliidae (and Rotoitidae, Signiphoridae and Tricho- probably Megalodontidae), Blasticotomidae, grammatidae), the shape of the mesofurca and Tenthredinidae, Argidae, Diprionidae, and attachment of muscle 150a are considerably differ- Cephidae. ent. The laterophragma of Cimbicidae (Fig. 8) pos- sesses an exaggerated apodeme (ap) and an en- Synonymy for mesofurcal bridge: larged posterior lobe (pn,), which is fused with the Siricidae & Vespidae: mesofurcal ring second phragma (2ph). The laterophragma is (Tenthredinidae, Vespa, Weber 1925). Ichneu- unique in form and apparently autapomorphic. monoidea: mesofurcal bridge (Stenobracon, In Xiphydriidae (Fig. 12), Monomachidae (Fig. Vanhorniidae Alaml951). Aculeata:mesofurcalbridge(ArF 2 ) 14), (Fig. 15), Cynipoidea (Fig. 22), (Vespula, Duncan 1939); supraneural bridge Ceraphronoidea (Fig. 25) and most Diapriidae (Apis, Snodgrass 1942). Chalcidoidea: arch of (Fig. 29), the laterophragma is excised between the the furca (Monodontomerus, Bucher 1948); ten- elongate apodeme (ap) and the posterior lobe dinous arch of the mesofurca (Tetramesa, James (pn,) (statel). In most of these taxa the apodeme of 1926). the laterophragma extends medially and horizon-

tally into the thoracic cavity, but in Cynipoidea (cf . Character 2. Laterophragma of mesopostnotum Fig. 22), the apodeme is vertical. The posterior lobe

) is lost in most some (pn 2 Apocrita (incuding Diapriidae), In Neuropterida, Mecopterida, Xyelidae and but this was not coded as an additional state Pamphiliidae, the laterophragma of the change for this characterbecause it reflects the loss mesopostnotum forms a lobe (pn,, Figs. 4, 6) (state of the muscle 116 (Character 4). The anterior 0) that extends obliquely into the mesothorax from apodeme (=axillary lever) or associated attach- the lateral corners of the mesopostnotum, mediad ment for muscle 150a is present in all Apocrita. of the lateral attachment to the upper mesepimeron Additional character state changes for the and lateral to the second phragma. The laterophragma in Apocrita are based only on laterophragma in all of these taxa forms the attach- changes in the shape of the anterior apodeme. ment site for muscle 150 see Character In and Orussidae (fu 2 -pn,, 3), Pergidae (Fig. 5) (Fig. 13), muscle 1 16 see Character and muscle 137 the is reduced to a narrow (t 2 -pn„ 4) laterophragma triangu- (pnj-ba,). The posterior face of the laterophragma lar process that forms an attachment for the ten- forms the site attachment for muscle 137 (pn,-bav don of muscle 150 (state 2). This reduction is also Fig. 3). Muscle 137 is usually small and difficult to associated with the loss of muscle 116. State 2 trace, but it is apparently lost in Pergidae and all could be derived from either state or state 1. Apocrita. In Apocrita, the axillary lever (ap) occurs in a In Xyelidae (Figs. 3, 4) and most Symphyta variety of shapes that probably have different (Figs. 6, 8, 9), an apodeme (ap) is present on the effects on leverage of the laterophragma with re- anterolateral margin of the lobe that serves as the spect to the fourth axillary sclerite and the second attachment site for muscle 150a The In all the of the (fu,-pn 2 ). phragma. Apocrita apex axillary laterophragma is functionally coupled with lever maintains a connection with muscle 150a 24S Journal of Hymenoptera Research

lever for is muscle In the (fu^-piv )• The plesiomorphic Apocrita 150a) (state 4?). Formicidae (Fig. 43), narrow and inflected medially into the thorax as lever forms a narrow, twisted apodeme (state 1?). found in Xiphidriidae (statel) and occurs either in Neither Bradynobaenidae and Formicidae were association with a posterior lobe (most Diapriidae, broadly surveyed and other forms may exist. Monomachidae (Fig. 14), Vanhorniidae (Fig. 15), In Ichneumonidae the axillary lever is either Ceraphronoidea (Fig. 25) and Cynipoidea (Fig. 22) closely appressed to the second phragma and or without the posterior lobe (Tngonalyidae (Fig. similar to the lever found in Vespoidea and 23), Stephanidae (Fig. 24), Diapriidae (Fig. 29), Apoidea (state 5), or rod-like and extending medi- Proctotrupidae (Fig. 32), some Ichneumonidae, ally into the mesolhorax (state 1), or inflected Mymarommatidae, and Chrysidoidea (Figs. 34- medially and dorsally (Fig. 17a). A robust lever is 39) (except some Chrysididae). The axillary lever not found in Braconidae, including Hybrizontinae exists in a variety of forms in Apocrita but in (Fig. 33), in which it was knob-like or short and shapes that are difficult to separate into discrete slightly deflected ventrally (state 1), similar to states, especially without a more thorough survey Trigonalyidae and some Proctotrupoidea. The le- of the apocritan taxa. In all cases where the lever ver of Gasteruptiidae (Fig. 26) is similar to some was narrow and inflected medially the character Ichneumonidae but is oriented vertically with re- was coded as state 1. spect to the second phragma. In Oodera In Platygastridae (Fig. 28), some Scelionidae (Pteromalidae), the lever is robust, appressed to and most Chalcidoidea, the axillary lever is de- the second phragma and horizontal (state 5). This flected ventrally (state 3). In Mymarommatidae, form of the lever appears to be convergent with the lever is robust and inflected medially (statel). Ichneumonidae and Aculeata, since in Oodera In Sparasion (Scelionidae; Fig. 16), the lever is muscle 150a attaches to the entire ventral aspect of reduced to a small cup-shaped lobe on the the apodeme as in other Chalcidoidea. laterophragma (autapomorphic and not coded). All Apiformes have the axillary lever sepa- In other Scelionidae and Mymaridae the lever is rated (state 6) as an independent sclerite (Snodgrass short, narrow and medially inflected (statel). 1942). This feature was verified in all of the The axillary lever in Stephanidae (Fig. 24) and Apiformes examined here. The apical connection most Chrysidoidea (Figs. 35-38, also in Fig. 34 but of the mesopostnotum is weak and encased by the this view is slightly different) is elongate and cup-shaped basal process of the axillary lever strongly inflected medially (state 1). This confor- (Figs. 45,54). mation of the lever is likely the precursor to the lever found in Apoidea and Vespoidea, discussed Synonomy for laterophragma: below. However, in Cleptinae and Chrysidinae Symphyta: lateral lobe of second phragma (Chrysididae) the lever is reduced (state 4) form- (2ph) (Daly 1963). Icnneumonoidea: muscle ing a short, broad process attaching to a broad bearing process of mesopostphragma of the tendon shortened muscle 150a (Fig. 40). (Stenobracon, Alam 1951); axillary lever In Apoidea (Figs. 45, 46, 54) and Vespoidea (Snodgrass 1942). Cynipoidea: mesopostnotal (Figs. 41, 42, 44) the axillary lever is robust and apodeme (Ibalia, Ronquist and Nordlander strongly appressed to the inner surface of the 1989); 2ph (Daly 1963). Aculeata: axillary le- second phragma (state 5). This type of lever was ver (Apis, Snodgrass 1957; Bombus, Pringle, found in all examined and is associated 1961 sclerite Apoidea 1957, 1960, ); accessory of the fourth a with robust muscle 150a. Except in axillary sclerite (Apis, Snodgrass 1910); inner Bradynobaenidae and Formicidae, muscle 150a is process of mesopostphragma (Vespa, Weber conical and attached to the lever by a narrow 1925 [states that lever turns apex of tendon or robust and nearly tubular (Mutillidae mesophragma and the connected axillary 4]); and Sapygidae). In Sierolomorphidae, the lever is anterior process of mesopostphragma (MPPhJ not closely appressed to the second phragma (in- (Vespula, Duncan 1939). Hymenoptera: axil- flected medially about 30°), but this was not treated lary lever (Matsuda 1970). as a different state (intermediate between states 1 and 5). In Bradynobaenidae, the lever is reduced to Character 3. Furcal-laterophragmal muscle a rounded knob (attaching to a narrow tendon of In Neuropterida and Mecopterida, only a Volume 3, 1994 249

muscle fu acts between the lost in all without a single (150, 2 -pn 2 ) Apocrita posterior mesofurca and laterophragma of the laterophragmal lobe. Autapomorphic losses of mesopostnotum (state 0) (Kelsey 1957, Matsuda muscle 150a occurs in workers of Formica and 1970). In the majority of Symphyta, including Camponotus (Markl 1966, Saini et al. 1982). Xyelidae and Pamphiliidae, two distinct muscles (150a and 150b) operate antagonistically on the Synonymy for muscle 150: enlarged laterophragma (Figs. 2, 3, 4, 6). The pre- Tenthredinidae: Ilisml (Dolerus, Schizocerus, sumed division of muscle 150 into two muscles is Tenthredo, Weber 1927), 25 & 26 considered as an apomorphy for Hymenoptera (Euthomostethus, Maki 1938). Ichneumonidae: (state 1). Muscle 150a (fu,-pn^) originates on the 25 (Psilopsyche, Maki 1938); 63 (Stenobracon, lateral arms of the mesofurca posterior or lateral to Alam 1951). Aculeata: 25 (Vespa, Maki 1938); the of origin 150b and inserts onto an anterior 78 (Formica, Markl 1 966; Apis, Snodgrass 1942); or of the apodeme process laterophragmal lobe IIdv2 (Vespula, Duncan 1939); Ilism (Vespa,

4, 6, 7, 9). Muscle 150b ) Weber fu (Figs. (fu,-pn 2 originates 1925); 2 -2ph (Apis, Daly 1964). Hy- anteriorly or medially on the lateral or anterior menoptera: t-sl (Matsuda 1970). arms of the mesofurca and broadly attaches to the of the margin laterophragmal lobe (Figs. 4, 6, 7, 9). Character 4. Second dorsal diagonal (=phragmal The fibres of muscle 150b are often arranged so flexor) muscle that the anteriormost fibres attach to the Muscle 116 is found between (t 2 -pn2 ) attaching posteriormost or innermost margin of the the mesonotum and the anterior face of the laterophragmal lobe, although this is dependent laterophragmal lobe in Neuropterida, Mecopterida on the size of the muscle and lobe. and most Symphyta (Figs. 2, 3, 12, 14, 15, 20), and If both muscles are muscle 116 is present, (t,-pn,) therefore plesiomorphic for Hymenoptera (state is also of present. Loss muscle 150b (state 2) occurs 0). Muscle 116, and the associated posterior lobe of in Cimbicidae, Monomachidae, Vanhorniidae, the laterophragma, is lost (statel) in Pergidae, and all Diapriidae, Cynipoidea, Ceraphronoidea, Orussidae and most Apocrita. It seems unlikely of which have retained both the laterophragmal that a functional complex composed of both muscle lobe the former (presumably point of insertion of 116 and the corresponding posterior lobe of the muscle 150b) and muscle 116. All Apocrita have laterophragma could be regained, and we con- lost muscle 150b. In Anaxyelidae (Fig.10), muscle sider the apomorphic state (loss of muscle and 150b is attached to the posterior face of the poste- lobe) to be irreversible. An assumption of irrevers- rior lobe the and muscle 150a is (versus margin) ibility has obvious consequences for models of retained. Because of a similar placement and at- character state change which are discussed in the tachment of muscles in Siricidae 1 it (Fig. 1), can be later section on parsimony analyses. Muscle 116 that muscle 150a is lost postulated and 150b re- and the corresponding posterior lobe are present mains (state 3). When both the laterophragmal in Monomachidae, Vanhorniidae, Cynipoidea, lobe and muscle 116 are as in missing, Pergidae Ceraphronoidea and most Diapriidae, and it is and Orussidae it is (Fig. 5) (Fig. 13), difficult to likely that this complex is the groundplan for assess which or muscle, 150a 150b, has been lost. Apocrita (Daly 1963; Gibson 1985). The loss of this could We assume that muscle 150b is lost in both complex in Pergidae and Orussidae is therefore families. the forward However, attachment of convergent with the loss in most Apocrita. muscle 150 to the mesofurca in Orussidae suggests that it is homologous with muscle 150b of Synonomy for muscle 116: and Siricidae. Anaxyelidae Therefore, the con- Symphyta: 21 (Euthomostethus, Maki 1938). in and figuration Pergidae Orussidae could also Symphyta, Monomachidae, Diapriidae, be treated as an parallel loss of muscle 150a. Be- Vanhorniidae, Ceraphronoidea, Cynipoidea: cause of this we code the uncertainty, reduction to t,-2ph (Daly 1963; Gibson 1985). Hymenoptera: a in single muscle Pergidae and Orussidae as of tl2 (Matsuda 1970). questionable homology (state ?). Based on the presence of muscle 150a in the apocritan families mentioned above, we assume that muscle 150b is 250 Journal of Hymenoptera Research

Character 5. Mesopostnotum and scutellar- the mesosoma) and muscle 114 passes externally metanotal muscle over the mesopostnotum as a tendon enclosed by In Neuropterida and Mecopterida, the a sheath of connective tissue (state 2, broadly with an mesopostnotum is broadly exposed medially; al- exposed and 114 external). In groups an though in Corydalus (Megaloptera) the external muscle and and exposed anterior medial mar- mesopostnotum is weakly sclerotized medially mesopostnotum (state 2), the be or and appears to be split. The scutellar-postnotal gin may shallowly deeply emarginate (ap- underneath muscle 114. In some muscle t -t ; absent in (muscle 114, 2 3 Bittacus) passes pearing split) of Tenthredinidae and the internally from its origin on the scutellar area of genera Dipnonidae, the mesonorum to the anterior medial margin of mesopostnotum is obscured by posterior devel- conforms the mesopostnotum (Kelsey 1957; Matsuda 1970). opment of the scutellum, but otherwise Within Hymenoptera, changes in the structure of to state2 (see Character 6). the mesopostnotum and attachment of muscle 114 In Anaxyelidae, the mesopostnotum and are correlated; however, the attachment of muscle muscle 114 are the same as in Xyelinae (Fig. 47). coded this as the same character state 114 to the mesopostnotum is unique with respect Initially, we this caused in devel- to the outgroup, and the outgroup is scored as a (state 1). However, problems an additive scheme that would force separate state (state 0). oping coding In muscle 114 arises the state to be and not Hymenoptera, (t,-t 3 ) anaxyelid aucapomorphic dorsomedially from a fan-shaped attachment on transitional between states 2 and 4. To resolve this char- the mesonorum and passes medially over the problem, Anaxyelidae were assigned a new state which is treated as an mesopostnotum (rarely under), to a narrow me- acter (state 3), dial attachment on the anterior margin of the autapomorphy. In all of the analyses, coding for metanotum. In all Symphyta, our dissections indi- Anaxyelidae and Xyelinae as state 1 or state 3 had cate a posterior attachment of muscle 114 to the no effect on tree topology. metanotum. Our observations are supported by In Siricidae, a median vertical process on the illustrations of attachments to the metanotum in anterior margin of the metanotum forms the pos- Weber (1925) for Vespa (his IldlmJ and Markl terior attachment of muscle 114, which is internal (1966) for Apis (his 70). In Eutomostethus and lacks any connective sheath. Although the (Tenthredinidae), muscle 114 was described as process is unique for Hymenoptera, this confor- is to be a modification of state inserting on the median membrane that divides mation considered is not the mesopostnotum (Maki 1938, his 19 & 20), but 2 even though the mesopostnotum generally in Tenthredinidae muscle 114 passes through the exposed. The scutellum often extends over the membrane to the metanotum. The attachment of metanotal process and, as in some Tenthre- muscle 114 to the metanotum is apomorphic for dinoidea, the internalization of muscle 114 and Hymenoptera. mesopostnotum may be secondary. Snodgrass Determination of the groundplan condition (1910) reported that the mesopostnotum of Tremex for Hymenoptera is complicated by the presence cohtmba was exposed medially; however, we ob- of different character states in the two extant sub- served an exposed mesopostnotum only in some families of Xyelidae. In Xyelinae (Xyelidae) (based specimens of Urocerus. Some Cephidae (Fig. 49) a on Pleuroneura, Fig. 47, and Xyela), the also have a similar attachment to peg-like pro- mesopostnotum (PNJ is broadly exposed dor- cess (mp) on the metanotum. and the scutellar-metanotal muscle 14, t In 50), Orussidae and sally (1 ; -t,) Xiphydriidae (Fig. under the a small muscle 114 -t and the passes mesopostnotum through Apocrita, (t2 3) mesopostnotum muscle 1 14 emargination of its posterior margin (state 1, (medially) are completely internal and broadly exposed and 114 internal). In passes medially over the second phragma (state 4 is Macroxyelinae (Xyelidae) (based on Macroxycla 4). An autapomorphic modification of state cuticle of the ferrunginea (Say), Megaxyela tricolor Norton, found in Xiphydriidae, in which the the lobes of the Megaxyela aviingrata (Dyar) and Xyelicia neurotica second phragma (between encircles the muscle tendon as it Ross), Megalodontoidea (Fig. 48),Tenthredinoidea pseudophragma) is to the Further devel- and Cephoidea (Fig. 49), the mesopostnotum passes through metanotum. exposed dorsally (depending on the contortion of opment of the mesopostnotum and its associated Volume 3. 1 994 251

phragma in Apocrita is toward an even greater fied their presence only in Corydalus. Muscle 127 is degree of internalization and a lateral shift of the lost (state 1, character 7) in all Hymenoptera ex- the points of articulation with the scutellum and met- cept Xyeloidea and Megalodontoidea. illustrates a muscle con- anotum (Snodgrass 1910) (see character 10). In Rasnitsyn (1969, fig. 204) the is almost nection between and fu in Aculeata (Fig. 53), second phragma (127?) sps, (his sst) 2 entirely disassociated mesally and the notal at- Tremex (Siricidae), but we did not observe this in tachments are reduced to lateral lamellae along either Tremex or Urocerus. The loss of muscle 180 the extreme lateral margins (Snodgrass 1942). (state 1, character8) is a synapomorphy of Apocrita.

Synonomy for muscle 114: Synonomy for muscle 127: Tenthredinidae: 19 & 20 (Euthomostethus, Maki Apidae: fu,-lsps [pupal muscle only] (Apis, 1938). Ichneumonidae: 20 & 21 (Psilopsyche, Daly 1964). Hymenoptera: sl2 (Matsuda 1970; Maki Aculeata: IIdlm Weber not listed Matsuda for 1938). 2 {Vespa, by Hymenoptera). 1925); 23 (Vespa, Maki 1938)'; Ilisl (Vespula, Duncan 70 Markl for muscle 180: 1939); (Apis, 1966); t,-t 3 {Apis, Synonomy Daly 1963). Hymenoptera: tl3 (Matsuda 1970, Symphyta: living (Dolerus, Schizocerus, Brodskiy 1992). Tenthredo, Weber 1927).

Character 6. Formation of pseudophragma Character 9. Furcal-basalare muscle The at the The furcal-basalare muscle postphragma (2ph) originates junc- (fbl, fu,-ba 3 ) ap- tion of the mesopostnotum and metanotum and pears to be found only in Hymenoptera (state 1). forms the posterior attachment site for the longitu- The muscle is apparently absent in the outgroup dinal flight muscles (112, lph-2ph) (Snodgrass (state 0). Muscle fbl could be homologous with 1910). In Mecopterida, Neuropterida, Xyelidae (Fig. Matsuda's p-s3 which connects the furca to the 47), Megalodontoidea (Fig. 48) and Tenthre- anterior margin of the succeeding episternum. dinoidea, the anterior margin of the Muscle p-s3 is found in Neuropterida (IIpml7, mesopostnotum is broadly attached to the poste- Korn 1916 for Myrmeleon), but apparently not in rior margin of the mesonotum (scutellum) by a other Neuropterida or Mecopterida (Matsuda narrow conjunctiva (60, term from Ronquist and 1970). In Symphyta (Figs. 2, 3, 5-13) muscle fbl Nordlander 1989) (state 0). In Cephidae (Fig. 9b), extends from the extreme lateral apex of the Anaxyelidae, Siricidae (Fig. lib), Xiphydriidae mesofurcal arm to the apodeme of the basalare of and the anterior of the the This muscle was first illus- Apocrita, margin hindwing (ba ? ). is a mesopostnotum developed into broad bilobed trated, but not discussed, by Rasnitsyn (1969, fig. internal plate (pseudophragma) that extends an- 204) for Tremex. In Diprionidae (Fig. 7), muscle fbl teriorly and dorsally beyond the dorsal attach- ends in a sclerotized cap attached to the apodeme of the to the of the with muscle 137 ment mesopostnotum mesonotum basalare, along (t,-ba ) (state 1). The pseudophragma is an extention of which originates from the posterior face of the the second and both form the phragma posterior laterophragma (t3 ). attachment for the longitudinal flight muscles (lph- Muscle fbl is absent (state 0) in most Apocrita. 2ph). A pseudophragma is not apparent in However, what appears to be the same muscle Orussidae (Fig. 13) and some Ichneumonidae and occurs in Mutillidae (Fig. 18), Rhopalosomatidae has probably been lost. (Fig. 19) and Bradynobaenidae (state 1). The muscle in these taxa originates laterally on the mesofurcal Characters 7-8. Furcal-spina muscles arms (dorsal to the furcal-trochanteral muscle, In Xyeloidea (Fig. 3) and Pamphiliidae (Fig. 2, musclel74) and narrows to a fine tendon that indicated dashed enters the of by line), muscles 127 (fu,-sps,) pleural area near the base the hind and 180 connect the with the furca The dorsal of insertion was not ob- (fu ? -sps,) spina wing. point of the following segment (stateO). Muscle 127 is served but the muscle enters the pleural area pos- present in both Neuropterida and Mecopterida terior to the subalar muscle of the mesothorax and and muscle 180 is found only in Neuropterida may insert onto the basalare or axillary sclerites of (Maki 1938, Kelsey 1957, Matsuda 1970); we veri- the hind wing. No such muscle was observed in 252 Journal of Hymenoptera Research

in Character 12. Furcal for trochanteral other Apocrita. The loss of muscle fbl Apocrita process could be the result of a shift of the muscle origin muscle 174 the from the furca to the pleural wall. Gibson (1986) In most Hymenoptera and outgroup taxa, of the trochanter reported a unique muscle in Eupelmidae and the furcal depressor (muscle 174, Pteromalidae -3ax that attached between fu^-tr^) attaches to the lateral arms of the mesofurca (pl 2 3 a) In and the lower mesepimeron and possibly the metatho- (Fig. 3) (state 0). Chrysididae (Cleptinae muscle 174 attaches to anterior ex- racic third axillary sclerite (of uncertain insertion; Chrysidinae), fbl tensions of the lateral furcal arms Gibson pers. comm. 1993). The homology of (Fig. 40, vp) In with the pleural muscles is beyond the scope of (state 1). Pompilidae (Fig. 42), Sapygidae, this study. Scoliidae, Sierolomorphidae and Vespidae (Fig. 41), muscle 174 attaches to similar extensions (vp) Character 10. Lateral articulation of that arise instead from the furcal bridge (state 2). - mesopostnotum In Vespoidea, muscle 174 has two origins the lateral furcal arms and the ventral In Xyeloidea, Megalodontoidea (Fig. 51) and process (= scoop Tenthredinoidea and the outgroup taxa, the shaped processes, schT, Weber 1925; anterior pro- mesopostnotum is broadly and evenly joined to cess of mesofurcal bridge, PF„ Duncan 1939). the upper mesepimeron (state 0). In Cephidae the HYPOTHESIS FOR (Fig. 52), Anaxyelidae and Siricidae, juncture PHYLOGENETIC is invaginated and connected by weak cuticle HYMENOPTERA 56 within the invagination, but maintains a strong Figs. 55, cuticular connection at the anteriormost point of classification of has attachment (state 1). In Xiphydnidae, Orussidae The higher Hymenoptera reviews and Apocrita, the mesopostnotum is completely come under increasing scrutiny following internal and the lateral connection of the published by Konigsmann (1977, 1978a) and mesopostnotum to the upper mesepimeron is re- Rasnitsyn (1969, 1980). Konigsmann (1978a) pre- duced to the anteriormost point of attachment sented evidence that Hymenoptera was comprised the (state 2). of two monophyletic sister taxa, Symphyta (excluding Cephoidea) and Cephoidea + Apocrita, Character 11. Fusion of lateral arms of mesofurca with most apocritan families presented as a series and metafurca of unresolved lineages. Although this more tradi- In most Aculeata the lateral arms of the tional concept of the Symphyta is still used (Zessin mesofurca and metafurca are closely appressed, 1985), it has not been accepted in more recent but as in most other Hymenoptera and the papers except to present an alternate hypothesis of outgroup taxa, they are broadly separated and relationships (Gauld and Bolton 1988). Rasnitsyn that is joined by the interfurcal muscle 181 (state 0) (Fig. (1980) presented evidence Symphyta lateral of relative to with 3). In Apoidea (Figs. 46, 53), the arms paraphyletic Apocrita Cephoidea both furcae are fused at the junction with the as the sister group of Apocrita. Studies subsequent mesofurcal bridge and the interfurcal muscle is to Rasnitsyn (1980) (Gibson 1985, 1986; Naumann led to lost (state 1). The lower furcal arms form a four- and Masner 1985; Carpenter 1986) Rasnitsyn cornered brace for the furcal complex, and the modify his original hypotheses in a later paper metafurcal arms are divided lateral to the furcal (Rasnitsyn 1988), in which Cephoidea were placed bridge. A similar fusion of the mesofurca and as a more basal lineage of Hymenoptera and far metafurca occurs in Amiseginae (Chrysididae; Fig. Orussidae as the sister group of Apocrita. A the 39) and Gasteruptiidae (Fig. 26), but in both cases more resolved set of relationships among para- muscle 181 is retained and these are considered to sitic Apocrita (Fig. 56) were presented by Rasnitsyn be non-homologous and autapomorphic changes. (1988). Some of Rasnitsyn's hypotheses have been Additional coding as autapomorphic states is un- corroborated in recent morphological or phyloge- necessary for this analysis but may be warranted netic treatments (Johnson 1988; Whitfield et al. in more extensive studies of related taxa. 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 As a framework for evaluating our results, we and Aculeata. For our data, these modifications have compiled the results of Rasnitsyn (1988) and affect only the distribution of character 12 in the Gibson (1985, 1993) to derive the cladogram pre- Aculeata, and have no effect on the parsimony sented in Figure 55. Contrary to Rasnitsyn (1988), arguments presented in a later section. Cladograms Hymenoptera excluding Xyeloidea are treated as comparing the hypotheses of Rasnitsyn and monophyletic based on the presence of a Konigsmann are presented in Whitfield (1992). postspiracular sclerite (=detached anepisternum) Rasnitsyn (1980, 1988), Shcherbakov (1980, (Gibson 1985, 1993) and an apically simple Rs vein 1981) and Gibson (1993) have proposed Xyelidae that is not bifurcate (Rasnitsyn 1988). Xyelidae are as the most basal extant lineage of Hymenoptera. treated as paraphyletic to allow for discussion of As well as possessing a branched Rs vein, Xyelidae Character 5. Tenthredinoidea are monophyletic retain the anepisternum as an integral part of the and placed here as sister group to the remaining pleuron and retain the plesiomorphic condition of Hymenoptera, excluding Xyeloidea and a relatively equally proportioned meso- and meta- Megalodontoidea, based on a reduction in the thorax (Gibson 1993). Other than the structure of number of preapical tibial spurs (from 2-4 spurs the antenna, which may be plesiomorphic, there on the hind legs in Xyeloidea and Megalodontoidea are no features that support the monophyly of the tolessthanl in Tenthredinoidea, l-2inCephoidea Xyelidae, and Rasnitsyn (1988) proposed that the and in remaining Symphyta and Apocrita; H. two subfamilies of Xyelidae belong to two lin- Goulet and G. Gibson, pers. comm., 1993), and the of loss of the ventral fu - eages Hymenoptera (Hymenoptera biphyletic mesofurcal-spina muscle, 2 with to the respect two subfamilies). Gibson (1993 spSj, which is present only in Xyeloidea and and personal communication, 1993) argues that Megalodontoidea (see Character Analysis). Hymenoptera excluding Xyelidae are monophyl- Remaining infraorders of extant Vespina pro- etic but that no characters support the monophyly posed by Rasnitsyn (1988) include the of the Xyelidae. The mesofurcal-mesopostnotal Orussomorpha, Evaniomorpha (Stephanoidea, complex of Pamphiliidae and both subfamilies of Evanioidea), Proctotrupomorpha (Proctotru- Xyelidae are basically the same and we consider poidea, Cynipoidea, Chalcidoidea), Ichneu- them to represent the plesiomorphic condition for monomorpha (Ichneumonoidea), and Hymenoptera. Vespomorpha (Chrysidoidea, Sphecoidea, Rasnitsyn (1988) classified the suborder Pompiloidea, Scolioidea, Formicoidea and Siricina into three infraorders that include the Vespoidea). Recent classifications (Naumann 1992; following extant families: Xyelomorpha (Xyelidae), Huber and Goulet 1993) treat the five families of Tenthredinomorpha (Argidae, Blasticotomidae, Rasnitsyn's Stephanoidea as higher taxa Cimbicidae, Pergidae and Tenthredinidae [includ- (Ceraphronoidea [Ceraphronidae and ing Diprionidae]), and Siricomorpha (including Megaspilidae], Megalyroidea, Stephanoidea and Megalodontoidea [=Pamphilioidea]: Pamphiliidae Trigonalyoidea). Following Brothers (1975) and and Megalodontidae; Cephoidea: Cephidae; Huber and Goulet (1993), the Aculeata are re- Siricoidea: Anaxyelidae, Siricidae and ferred to only three superfamilies (Chrysidoidea, Xiphydriidae). Rasnitsyn (1988) treated Xyelidae Apoidea and Vespoidea) and Mymarommatidae as paraphyletic with respect to Tenthredinoidea are placed as a superfamily. and remaining Hymenoptera, and the three fami- lies of Siricoidea are paraphyletic with respect to GROUP ANALYSIS the Orussidae + Apocrita (Suborder Vespina). Rasnitsyn (1988) placed Cephoidea as basal to the Character states are referred to in brackets as + Siricoidea Apocrita, rather than sister group to [charactenstate]. Relationships are discussed ac- the as in Apocrita his earlier treatments (Rasnitsyn cording to the phy logenetic hypotheses illustrated 1969, 1980). 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 Tenthredinoidea + remaining Hymenoptera Hymenoptera of Tenthredinoidea and the re- Figs. 1, 2, 3, 4, 6, 47, 48, 51 Monophyly maining Hymenoptera is supported by the loss of of the mesofurcal- muscle 127 Four characters (fu 2-spSj) [7:1]. mesopostnotal (MF-MPN) complex are not found in the outgroups and therefore support the mono- Tenthredinoidea phyly of Hymenoptera: long anterior arms of the Figs. 5, 7-8, 20, 21 mesofurca [1:1]; two furcal laterophragmal muscles anterior mesofurcal arms are reduced or [3:1]; muscle 114 attaching to the metanotum (part The of Character 5, see below); and furcal-basalare lost in all Tenthredinoidea [1:2].

muscle (fbl, fu,-ba.,) present [9:1]. Xyeloidea and Megalodontoidea are Blashcotomidae (Fig. 20) of muscles is plesiomorphic for all characters postulated for The plesiomorphic combination Hymenoptera in the mesofurcal-mesopostnotal present. Anterior furcal arms are present, but they complex. Megalondontidae were verified for ex- are short and thin. As discussed earlier (analysis of anterior characterl because of uncertainties in the homol- ternal characters only and the presence of ), in of reduced have scored both re- furcal arms [1:1] was based on fig. 189 Rasnitsyn ogy arms, we (1969). Rasnitsyn (1969) also refers to the duced and absent anterior furcal arms as the same mesofurco-metapre-episternal muscle as found character state (1:2). The anterior apodeme of the only in Xyela. This muscle was not identified in our laterophragma is absent and both 150a and 150b dissections, although it may be synonymous with attach to the apex of the posterior lobe (Fig. 20a). This is considered to be an devel- muscle 127 (fu^sps,). autapomorphic The mesopostnotum is external and broad opment, but there is also a general similarity with medially in both Xyeloidea and Megalodontoidea. the laterophragma of Nematus (Fig. 19), in which In extant Xyelinae (Xyelidae), the mesopostnotum short anterior furcal arms are also present. is entire medially and muscle 114 is internal [5:1, Fig. 47]. A similar conformation is found in the Tenthredinidae (Fig. 21) outgroup, but the attachment of muscle 114 is The plesiomorphic combination of muscles is different [5:0]. In Macroxyelinae (Xyelidae), present in all taxa examined, and the structure of Megalodontoidea and Tenthredinoidea, the the laterophragma is the same as in Xyelidae and mesopostnotum is almost completely separated Pamphiliidae. The anterior arms of the mesofurca medially and muscle 114 passes externally [5:2, vary from narrow, elongate apodemes extending Fig. 48]. The deep emargination of the from the anterior face of the lateral arms (Nematus), mesopostnotum associated with state 5:2 is visible to short cup-like structures similar to those found externally and can be interpreted from fossil Hy- in Diprionidae, to completely absent (most menoptera. Based on illustrations in Rasnitsyn Tenthredinidae). Muscle 180 (fu^sps,) was not (1969), external passage occurs in the fossil xyelids: found in Eutomostethus (Blennocampinae) (Maki

Archexyelinae: Asioxyela (fig. 34), Leioxyela (fig. 1938).

41), Triassoxyela (fig. 32), Xiphoxyela (figs. 45, 46); Macroxyelinae: Agaridyela Rasn. (fig. 41), Diprionidae (Fig. 7) combination of muscles Chaetoxyela Rasn. (fig. 76), Ceroxyela Rasn. (fig. 77); The plesiomorphic structures for Tenthredinoidea are and Xyelinae: Eoxyela Rasn. (fig. 59), Lydoxyela and present the features: ante- Rasn. (fig. 68), Spathoxyela Rasn. (fig. 42), Xyelisca but with following apomorphic rior reduced to for muscle Rasn. (fig. 61), and Xyela mesozoica Rasn. (fig. 62). arms cup-like receptacles If correctly interpreted in fossil Xyelinae, the inter- 124; anterior process of laterophragma expanded nal passage of muscle 114 and an entire and forming a broad attachment for 150a; and mesopostnotum are derived within extant muscle 150b originates medially to 150a on the Xyelinae, and the external passage of muscle 114 lateral furcal arms. and an emarginate mesopostnotum are probably apomorphic for Hymenoptera Volume 3. 1 994 255

Cimbicidae (Fig. 8) lower Hymenoptera and presence in Cephidae, The anterior arms are reduced to small spoon- Anaxyelidae, Siricidae, Xiphydriidae and most shaped processes, the posterior lobe of Apocrita suggest that it is a synapomorphy at this mesopostnotum is fused to second phragma form- level. There are no published accounts which make ing attachment for a large muscle 116, and muscle reference to a pseudophragma occuring elsewhere 150b is lost [3:2]. The anterior process of the within Endopterygota. laterophragma is enlarged and extends ventrally Cephidae are plesiomorphic for other charac- into the cup-shaped lateral arm of the mesofurca ters of the mesofurcal-mesopostnotal complex: (autapomorphy). muscles 116, 150a, 150b, and fbl are present, and the laterophragma is divided into a large posterior Argidae + Pergidae lobe and a small anterior process. The The lateral arms of mesofurca are tubular and mesopostnotum is usually exposed medially but hollow (apomorphic), the anterior arms of the is sometimes concealed by the scutellum. Muscle mesofurca are absent, and the attachment site for 114 passes externally over the mesopostnotum in muscle 124 is to the rounded surface of the all (f Uj-fiij) genera (plesiomorphic) and may arise lateral arms. Although they would suggest a sister dorsomedially from a vertical process on the group relationship for these two families, these metatergum {Calameuta Konow, Ceplms Latrielle characters were not surveyed throughout Hy- [mp, Fig. 49], Monoloplopus Konow, Syrista Konow menoptera and were not coded for the parsimony and Tracheitis Jurine) as in Siricidae, or from an analysis. emargination of the metatergum (Caenocephalus Konow and Janus Stephens) as in Anaxyelidae. Argidae The laterophragma and its associated muscle Anaxyelidae + (Siricidae + (Xiphydriidae + attachments are as in Tenthredinidae. (Orussidae + Apocrita))) Fig. 10 Pergidae (Fig. 5) This family displays a number of Monophyly of the Anaxyelidae and remain- autapomorphies: loss of the tergal depressor ing Hymenoptera is based on a more complete muscle (116) [4:1] and the associated posterior fusion of the anterior arms of the mesofurca [1:4]. lobe of the laterophragma [2:2], and loss of one of Fusion of the anterior mesofurcal arms in the furcal laterophragmal muscles (possibly 150b) Anaxyelidae (Fig. 10a) appears to be more com- The is to a small [3:?]. laterophragma reduced plete than in Siricidae (Fig. 11a) in which the arms triangular lobe that ends in a narrow attachment are fused in about the anterior half. The anterior to muscle 150. The lateral arms form a of cup-shaped arms Anaxyelidae (Fig. 10a) are still traceable in attachment for muscles 150 and fbl. sclerotized A dorsal view and they are connected posteriorly by apodeme, probably derived from the lateral arm a thin plate of cuticle. The greater degree of fusion and muscle 151 (pl,-fu, ), extends to the of the arms is in rt laterally probably autapomophic pleural ridge. Ventrolateral extensions of the lat- Anaxyelidae; however, in Anaxyelidae and eral arms are developed as insertions for the en- Siricidae the fusion is intermediate between the muscle 170 -fu connection found in larged (pl2 2b ). apical Cephidae and the com- plete fusion including a strong furcal bridge found + Cephidae remaining Hymenoptera in Xiphydriidae + Orussidae + Apocrita. In Figs. 9, 49, 52 Anaxyelidae, Siricidae and Xiphydriidae, muscle 150b inserts onto the posterior face of the Monophyly of the Cephidae and the remain- laterophragmal lobe rather than to the ventral ing Hymenoptera is supported by a suite of char- margin as in Xyeloidea, Megalodontoidea, acters: connection apical of the anterior mesofurcal Tenthredinoidea and Cephoidea. As the muscle is arms formation [1:3], of a pseudophragma [6:1], absent in Orussidae and Apocrita, the importance and of the invagination mesopostnotum laterally of this unique attachment as a character for sup- The [10:1]. pseudophragma is not present in all porting monophyly of Siricoidea (including but its Hymenoptera, complete absence in the Anaxyelidae and Xiphydriidae) or Siricoidea + 256 Journal of Hymenoptera Research

of Orussoidea + Apocrita cannot be assessed. in the remaining genera Anaxyelidae including libocedrii Rohwer. In Anaxyelidae, the presence of both furcal- the only extant species, Syntexis the antennal is laterophragmal muscles (150a & 150b) and ante- However, flagellum apomorphic fureal arms is and in rior origin of muscle 150b on the (reduced divided) Xyeloidea, Megalo- is Tenthredinoidea, and plesiomorphic. The mesopostnotum broadly dontoidea, Cephoidea and thus its reduction in and muscle 114 (t,-t ) within exposed medially 3 passes Anaxyelidae, under the mesopostnotum medially (autapo- Siricidae and remaining Apocrita is not a reliable of muscle 114 character. The meet the morphy; 5:3). The dorsal enclosure postgenae broadly along line to form a in Anaxyelidae may result from the extension and medial postgenal bridge (=genaponta medial fusion of the anterior margin of the sensu Ross 1937) in Anaxyelidae, Siricidae and illus- we with Ross (1937) that the mesopostnotum. Rasnitsyn (1969, fig. 99) Xiphydriidae; agree for the fossil of and are trates a split mesopostnotum [5:2] postgenae Anaxyelidae Xiphydriidae This the con- similar and thus cannot be excluded Anaxyela gracilis Martynov. supports Anaxyelidae tention that an internal muscle 114 is derived from Siricoidea based on this character. Gibson demonstrated that a transscutal articula- within Anaxyelidae. (1985) tion is absent in Cephidae, Anaxyelidae and Siricidae + (Xiphydriidae + (Orussidae + Siricidae. The presence of a concealed prepectus is of Siricidae Apocrita)) useful for supporting the monophyly + + Orussidae + Fig. 11 Xiphydriidae Apocrita (Rasnitsyn 1988; Gibson 1985). No characters of the mesofurcal- + + mesopostnotal complex demonstrate the mono- Xiphydriidae (Orussidae Apocrita) 12 phyly of Siricidae and the remaining Hymenoptera. Fig. 1 1 The anterior mesofurcal arms of Siricidae (Fig. ) is in the MF-MPN com- are vertically flattened and in dorsal view are There strong support of Orussidae distinct along their entire length. The degree of plex for the monophyly Xiphydriidae, fusion of the anterior arms is intermediate be- and Apocrita. The anterior mesofurcal arms are tween Cephidae and Xiphydriidae + Orussidae + completely fused into an anterior furcal process and to form a trans- Apocrita, but, as discussed above, is comparable (fp) posteriorly strengthened to that found in Anaxyelidae. Siricidae are verse mesofurcal bridge (fb) [1:4], with the excep- taxa in which the mesofurcal autapomorphic for the loss of muscle 150a (along tion of some Apocritan the is with the anterior process of the laterophragma) bridge is lost. In addition, mesopostnotum of 114 internal and [3:3]. The posterior attachment muscle (t,-t,) completely medially [5:4] laterally incised to a vertical process of the metanotum is shared [10:2], and the laterophragma is strongly with some Cephidae but not with Xiphydriidae, with the anterior process (ap) developed into a The combi- Orussidae or Apocrita, which lack a process. Inter- strong apodeme [2:1]. plesiomorphic nal passage of muscle 114 in Siricidae may result nation of muscles in the MF-MPN complex is the Siricoidea and from the posterior extension of the scuteOum rather retained. Among Cephoidea, than internalization of the mesopostnotum as in Orussoidea, Xiphydriidae are unique in having the lateral arms of the Cephidae and Tenthredinoidea. muscle 150b attached to Rasnitsyn (1988) proposed that Siricidae + mesofurca (versus the anterior arms or furcal pro- 124 is attached to remaining Hymenoptera were monophyletic based cess). Muscle (fu^-fUj) broadly In the on four synapomorphies: 1) compound third an- the furcal process and bridge. Xiphydriidae, reduced and to the fol- tendon of muscle 114 -t is enclosed tennal segment subequal (t2 3 ) medially, cuticle of the second lowing segments, 2) head capsule with postgenae and internally, by the phragma subcontiguous, divided with narrow hypostomal (autapomorphic). bridge, 3) transscutal articulation present, and 4) + prepectus concealed. In Anaxyelidae, the primi- Orussidae Apocrita tive state of a compound third antennal segment is Fig. 13 found only in the fossil anaxyelid genus in the Sphenosyntexis Rasnitsyn; the derived state is found No synapomorphies were found Volume 3, 1994 257 mesofurcal-mesopostnotal complex that support Symphy ta and attaches laterally to the mesofurcal the monophyly of Orussidae + Apocrita without arms (Figs. 3, 5, 9b, 12b, 13). In Apocrita, the Xiphydriidae. It is possible that the loss of muscle position of muscle 170 shifts from the apex of the 150a supports the monophyly of Orussidae + furcal arms to the base of the furca and is more

= . Apocrita [3:? 3:2] However, the homology of the nearly horizontal. This is most apparent in remaining muscle 150 in Orussidae is uncertain Vespoidea (cf. Duncan 1939, muscle IIfpl2 (71)). (see discussion of Character 3). Therefore, it is also The laterophragma of Cynipoidea, some possible that muscle 150b is lost in Orussidae and Diapriidae, Monomachidae (Fig. 14) and 150a is lost in Apocrita. Vanhorniidae (Fig. 15a) are similar in structure Orussidae have at least three apomorphic fea- with an anterior apodeme (ap) narrowly sepa- tures in the MF-MPN complex, two of which are rated by a deep incision from the posterior lobe also found in Pergidae and Cimbicidae. The furcal (pn,) [1:5], which supports muscle 116 [4:0]. In bridge (fb) is rounded and bowed anteriorly with Megaspilidae (Fig. 25), the apodeme and lobe may a distinct median process (fp) [1:6]. The attach- be broadly or narrowly separated. Muscle 116 ment of muscle 150 to the furcal is found 1 is in the families listed process [4: ] present only above and only in Orussidae. In Apocrita, muscle 150a is probably has been lost numerous times within usually attached to the furcal bridge or lateral arm. Apocrita (Gibson 1985). It is unlikely that this Otherwise the furcal bridge and process in muscle and associated posterior lobe could be Orussidae are similar to Xiphydriidae and other regained once lost. Under the present hypothesis Apocrita. Muscle 116 is lost [4:1] in Orussidae, and of relationships in Apocrita, muscle 116 and the there is an associated reduction of the lobe are lost at least 9 times (10 if also lost in to a laterophragma small triangular lobe that forms Austroniidae) (Fig. 56). Of these, only in Diapriidae the attachment for muscle 150 [2:2]. Additional is the muscle lost within an entire family. autapomorphies in Orussidae include loss of the Groundplan states for the Apocrita are exem- pseudophragma [6:0] and the configuration of plified by Monomachidae (Fig. 14), Vanhorniidae muscle 180 which arises from the most and (fu 3 -sps2 ) spina (Fig. 15), Diapriidae (Fig. 29) Cynipoidea as a single narrow tendon that bifurcates about (Fig. 22), in which muscle 116 and the posterior half-way along its length to the metafurcal arms lobe of the laterophragma are retained (Daly 1963; (the plesiomorphic configuration is completely Gibson 1985). Plesiomorphic attributes of Apocrita its entire separated along length). include l)presence of a median furcal process (fp) forming the attachment for muscle 124, 2) furcal Apocrita bridge complete (fb) [1:4], 3) muscle 150a originat- Figs. 14-19, 22-46, 53, 54 ing laterally on the bridge or the lateral arm of the mesofurca, 4)laterophragma divided into an ante- loss of in The two muscles the MF-MPN com- rior apodeme and posterior lobe [2:1], 5)presence plex supports the monophyly of the Apocrita: the of the tergal depressor muscle (116) [4:0], 6) muscle fu and internal and metafurcal-spina (180, 3 -sps 2 ) [8:1] mesopostnotum [10:2] 7) prephragma the furcal-basalare muscle fu -ba have lost (fbl, 2 3 ) [9:0]. present [6:1]. Ceraphronoidea (Fig. 25) Muscle 150b is lost in all the mesofurcal but retain a divided (fu 2-pn2a ) [3:2] Apocrita bridge [1:7] those that retain the of the including posterior lobe laterophragma, and in some Diapriidae (Fig. 22) laterophragma. However, as discussed above, the the posterior lobe of the laterophragma is lost [4:1] homology of the remaining muscle 150 in Apart from changes in the laterophragma, Orussidae is not clear, and therefore loss of 150a is apomorphic changes of the MF-MPN complex in either a synapomorphy of Orussidae + Apocrita or Apocrita include loss of the mesofurcal bridge an autapomorphy of Apocrita. Loss of musclel37 [1:7] and changes in the structure of the anterior also may support the monophyly of Apocrita (see apodeme (=axillary lever) of the laterophragma discussion for Character 2). [2:3-6]. Changes in the shape of muscle 150a, lat- Shcherbakov (1981) stated that muscle 170 eral arms, furcal bridge and second phragma are

-fu , is connected to the coxal of and related to in (pl 2 2b s-cx2) process homoplastic probably changes the mesopleuron in all Hymenoptera. Based on thoracic shape. Notable exceptions in Aculeata our it is well in observations, developed all include development of an axillary lever [2:5] that 258 Journal of Hymenoptera Research is separated in Apiformes [2:6], fusion of the lat- the groundplan state of this character is in either eral arms of the mesofurca and metafurca [11:1], superfamily. Also, the axillary lever is deflected development of processes on the mesofurca for [2:3] in Platygastridae, some Scelionidae and most the furcal-trochanteral muscle [12:1,2], and a me- Chalcidoidea. Almost all Chalcidoidea including dial reduction of the attachment points (60 & 61) Mymaridae are unique in that muscle 150a at- between the second phragma and terga. taches to the entire length of the deflected axillary The cladogram presented in Figure 56 is con- lever (versus only apical attachment). sistent with the relationships proposed by Mymarommatidae have a medially inflected lever and Brothers and 1 and muscle 150a is conical and attaches Rasnitsyn (1988) Carpenter ( 993) [2:1] only for extant Apocrita, and is used as the model for to the apex of the lever (plesiomorphic). examination of character state change within Proctotrupidae (Fig. 32) have a similiar axillary Apocrita. Apart from the modifications in Aculeata, lever to that of Platygastridae (Fig. 28), although it the MF-MPN complex offers little support for is shorter and not as strongly deflected (Figs. 29- phylogenetic relationships proposed for Apocrita. 32). In Heloridae (Fig. 30) and Pelecinidae (Fig. 31), Within Apocrita a total of 22 steps are required to the lever is followed by a short spine (pn,?) that explain the distribution of character states as mod- has no muscle connection and may be the remnant elled on Figure 56 (excluding changes in the of the posterior lobe of the laterophragma. Some Chalcidoidea + Platygastroidea + Mymarom- Proctotrupidae have a blunt triangular lobe in a matoidea trichotomy, Austroniidae and homologous position. Heloridae, Pelecinidae and Peradeniidae). Given that the tergal depressor Proctotrupidae are the only Hymenoptera with a muscle and posterior lobe of the laterophragma vestigial posterior lobe of the mesopostnotum. are present [4:0] in what are assumed to be rela- Ropronndae were identical to Heloridae inter- tively derived groups (e.g., Cynipoidea, Diapriidae nally except for the lack of a vestigial posterior and Monomachidae), this complex must be lost a lobe. minimum of 9 times (or 10 times if it is absent in Most Ichneumonoidea (Fig. 17) have a Austroniidae). Only 8 losses are required if mesofurca consisting of a broad mesofurcal bridge Pelecinidae and Vanhorniidae are monophyletic and a strong medial projection. The furcal bridge of (Fig. 56). An additional 3 losses muscle 116 [4:1] of Hybrizontinae (Braconidae) (Fig. 33) lacks a are required using Konigsmann's (1978a) hypoth- median process but this could be autapomorphic. esis for Apocrita. Obviously, loss of this muscle The axillary lever in Ichneumonidae is similar to does not offer strong evidence for the monophyly Vespoidea and Apoidea [2:5]. This may be evi- of Megalyroidea, Stephanoidea and Trigo- dence for the close relationships to Aculeata that nalyoidea (Fig. 56). have been proposed (Rasnitsyn 1988, Sharkey and Among the parasitic families of Apocrita, the Wahl 1992), but a similar lever has not been found mesofurcal complex of Gasteruptiidae (Fig. 26) in Braconidae or Chrysidoidea. and Evaniidae (Fig. 27) were very different from Chrysidoidea do not have any defining each other. However, the mesofurcal bridge of apomorphies of the mesofurca or axillary lever. Aulacidae and Evaniidae are similar in shape The arched bridge of Plumariidae (Fig. 34), (broad and flattened with muscle 124 attaching Sclerogibbidae (Fig. 35) and Embolemidae (Fig. under the anterior margin of the bridge). In all 36) are similar but this feature may be three families the axillary lever is strongly in- plesiomorphic, as this shape is similar to flected and perpendicular to the lateral wall of the Stephanidae. The furcal bridge of the single fe- mesosoma (Figs. 26, 27), and the lever is connected male dryinid examined (Anteoninae) was similar a thin to by long tendon muscle 150a. Scelionidae to Bethylidae (Fig. 38) (relatively straight with a and Chalcidoidea both have a similar mesofurcal strong apical projection), and the bridge of the bridge (bowed with the median process virtually male dryinid (Fig. 35) was reduced to a straight absent) which may indicate a sister-group rela- bar. Obviously more taxa need to be surveyed to tionship between Chalcidoidea and Platy- understand the importance of furcal shape in gastroidea. However, given that the bridge is lost Chrysidoidea. Chrysididae have apomorphies that in Platygastridae, Mymarommatidae, Mymaridae distinguish taxa at the subfamily level: the fusion and some Chalcidoidea [1:7], it is not clear what of the lateral arms of the mesofurca and metafurca Volume 3, 1 994 259

or in Amiseginae (Fig. 39), and extensions of the Huber 1993) as the sister group of Spheciformes furca that support the furcal depressor of the + Apiformes (Carpenter and Brothers 1993). trochanter [12:1] in Chrysidinae (Fig. 40) and Apoidea (Figs. 45, 53), including Heterogynaidae, Cleptinae. These latter two subfamilies also share have the arms of the mesofurca and metafurca a reduction of the axillary lever [2:4]. Reduction of fused at the junction with the mesofurcal bridge the lever and of ventral and muscle 181 -fu is lost This feature axillary presence pro- (fu 3 2 ) [11:1]. cesses in Cleptinae and Chrysidinae do not sup- is unique in Hymenoptera. In Vespoidea (Figs. 41, port the hypothesis of relationships among 42, 44), Spheciformes (Fig. 46) and Heterogynaidae, Chrysididae proposed by Kimsey and Bohart the base of the axillary lever is broadly fused to the (1990) of Cleptinae + (Amiseginae + Chrysidinae). lateral arm of the mesopostnotum. All Apiformes Based on relationships for Chrysidoidea proposed have the axillary lever separated from the by Brothers and Carpenter (1993), Carpenter (1986) mesopostnotum as an independent sclerite [2:6] and Rasnitsyn (1988), the plumariid type of bridge (Fig. 45) (Snodgrass 1942). is plesiomorphic, Dry inidae have an independently derived bridge, and Chrysididae + Bethylidae have PARSIMONY ANALYSIS a relatively broad, straight furcal bridge with a strong medial projection. The mesofurcal-mesopostnotal complex rep- Monophyly of the Vespoidea and Apoidea is resents a single system of inter-related characters, supported by the development of an axillary lever and it should not be used alone to form new [2:6], which is subsequently reduced in hypotheses of relationships. However, it is of in- Bradynobaenidae [2:4] and modified in Formicidae terest to determine if our interpretations of charac- [2:1]. Most Aculeata also have a broad, flattened ter state evolution are in fact the most parsimoni- furcal bridge with muscle 124 attaching to the ous for these data. We tried two approaches to dorsal and lateral surfaces (Fig. 18), although the explore this question. First, we superimposed the bridge is not as well developed in Rhopalo- data in Appendix 2 on existing hypotheses of somatidae (Fig. 19), Sierolomorphidae and some relationships. PAUP Version 3.0s (Swofford 1985) Sphecidae (Fig. 46). In Tiphiidae (Fig. 44), the furca was used to optimize the character state data in is convoluted and broadly excavated dorsally. Appendix 2 on Figure 55 for Symphyta and Fig- The furcal bridge has several shapes in Vespoidea ures 55 and 56 for the entire Hymenoptera (using but their importance for assessing relationships one fully resolved tree for Apocrita based on will require a much broader survey of taxa. A Rasnitsyn (1988) and Brothers and Carpenter strong ventral process (vp) for the furcal depres- (1993)). Ibaliidae was used to represent the sor of the trochanter (174, fu,-tr,) confluent with groundplan of Apocrita in the former case, and the furcal is in bridge [12:2] present Sapygidae, each matrix included the outgroup (Appendix 2). Sierolomorphidae, Pompilidae (Fig. 42), Scoliidae All most parsimonious models of character state and Vespidae (Fig. 41). Under the most recent evolution were examined on these tree topologies. hypothesis of relationships for Vespoidea (Broth- Characters 1 and 4 are sensitive to assump- ers and Carpenter 1993), the ventral process is tions about transformation series. As discussed derived independently in Sapygidae, Pompilidae above (Character Analysis), it seems most reason- and Scoliidae + Vespidae, although several mod- able to assume that character 1 can evolve from els of state are Mutillidae change possible. (Fig. state to state 1 (unique for Hymenoptera), from 18), Rhopalosomatidae (Fig. 19) and Bradyno- state 1 to state 2 or state 3, from state 3 to state 4 and baenidae have apparently retained muscle fbl from state 4 to state 5; states 6 and 7 are indepen-

-ba ) and have a dent derivations from state 5. 1 (fu 2 3 [9:1] characteristically shaped Character was mesofurca, however these families are not regarded ordered using a character state tree (or "stepmatrix" as sister taxa (Brothers 1975, Carpenter and Broth- in PAUP) that specified the above transformation ers 1993). series. As discussed earlier, character 4 is treated Apoidea have been separated into two groups, as irreversible. The assumptions of additive cod- the Apiformes and the Spheciformes (Goulet and ing for characters 1 and 4 are referred to as the Huber 1993). Heterogynaidae have been included Ordered Characters. Other characters were al- either within the Spheciformes (see Goulet and ways treated as nonadditive (unordered). 260 Journal of Hymenoptera Research

For Symphyta, the models of character state different optimization for Character 4 (treated as change shown in Figure 55 are in fact the most irreversible in the ordered analysis) and reversals parsimonious for that tree topology. 27 steps are (4:1 to 4:0) are favoured. For this result, the loss of required to explain the data (indices of homoplasy the posterior lobe of the laterophragma and asso- are not possible for characters ordered using step ciated muscle 1 16 (4:1) would be a synapomorphy matrices). Different models of state change are of Orussidae and Apocrita and then the presence possible for character 3 (because it is scored as of both structures (4:0) would have to be regained missing in Pergidae and Orussidae) and character at least 4 times within Apocrita. Although this 5 (unknown polarity for ingroup). For character 5, may be a more parsimonious solution in terms of the presence of state 2 in fossil Xyelinae argues that the number of steps, we feel it is an unacceptable this condition is groundplan for Hymenoptera, model of character evolution and the assumption but it is equally parsimonious (at least for extant of irreversibility should be invoked. taxa) to assume that state 1 is groundplan and state Parsimony analysis using the branch and 2 arises in the interval below Macroxyelinae and bound algorithm (Symphyta) or branch-swapping remaining Hymenoptera (dashed line, Fig. 55). (Hymenoptera) was then performed on these data The latter solution would make Xyelidae to determine if another tree topology would result paraphyletic. Treating all characters as nonaddi- in a more parsimonious solution. When characters tive (unordered) does not change the tree length are ordered as above, for Symphyta, 24 trees of 27 (27 steps, consistency index 0.81, retention index steps result (after derooting, condensing and 0.77). In fact, the only effect is to allow for a number rerooting the resulting trees in PAUP, which is of alternate models of state change for character 1 . necessary when using directed characters). Cod- In this case, it is equally parsimonious to assume ing Anaxyelidae as state 1 (homoplastic) or state 3 any of states 0, 1, 2 or 3 in the nodes below (autapomorphic) had no effect on the number of Tenthredinoidea + remaining Hymenoptera and steps or resulting trees. As should be apparent Cephidae + remaining Hymenoptera, a result that from Figure 55, the basal taxa (Xyelinae, essentially discards the information in the trans- Macroxyelinae and Pamphiliidae) are unresolved formation series. Optimizing the ordered data in each of the 24 trees. Tenthredinoidea is always onto alternate hypotheses of Symphyta resulted in resolved as monophyletic but with little or no longer trees. Rasnitsyn's (1988) tree, with further resolution and Cephidae is consistently Macroxyelinae as sister group to Tenthredinoidea, placed as sister group to remaining Hymenoptera, is one step longer, because of the parallel loss of as in Figure 55. Anaxyelidae and Siricidae form a muscle 127 [7:1]. Konigsmann's (1977) tree, with trifurcation with Xiphydriidae + Orussidae + Symphyta monophyletic and Cephidae as sister Apocrita while the latter three taxa are essentially group to Apocrita, requires 37 steps (character 3 unresolved. In other words, there is no more par- optimized). simonious interpretation of these data and the For Hymenoptera, similar results were ob- most parsimonious solutions are congruent with tained for optimizing the Ordered Character data the data shown on Figure 55. When characters are onto one tree, which is summarized for Apocrita unordered for Symphyta, 138 trees of 26 steps in Figure 56. 58 steps were required to explain the result. The strict consensus solution for these trees data. Different optimizations were possible for is almost completely unresolved. For Hy- Character 1 forChalcidoidea + Mymarommatoidea menoptera, the lack of character support in + Platygastroidea (if parallel loss of mesofurcal Apocrita resulted in a considerably shorter tree bridge [1:7] is preferred), and Character 9 in (40 steps for the Ordered Characters) but also Vespoidea. In the former case, although it is equally provided no resolution of Apocritan relationships; parsimonious to assume loss of the mesofurcal the structure of the Symphyta portion remained bridge (1:7) followed by a gain [1:6], parallel loss of unchanged. the bridge seems more likely. Likewise, it seems more likely to postulate a parallel loss of the furcal- CONCLUSIONS basalare muscle [9:0] in Vespoidea. Treating all characters as nonadditive decreases the length of The mesofurcal-mesopostnotal complex is a the hymenopteran tree by 5steps because of a conservative character system that generally ex- Volume 3. 1 994 261 hibits few changes within families of Hymenoptera Xiphydriidae, Orussidae and Apocrita is supported but undergoes considerable change at higher taxo- by four synapomorphies. No evidence in the nomic levels. The most significant changes occur mesofurcal-mesopostnotal complex supports the in the Symphyta, but additional characteristics of inclusion of Xiphydriidae or Orussidae within the mesofurcal-mesopostnotal complex provide Siricoidea or the monophyly of Anaxyelidae + characters that are useful in inferring relation- Siricidae. ships within Apocrita. The monophyly of Apocrita is supported based

Of the for on losses of the muscles fu , fu , fu -ba , competing hypotheses relationships 2-pn, 3-sps 2 2 3 of those in 55 are the most -fu and t -ba . the evolution- Symphyta, Figure pl2 2b , probably Using parsimonious using the hypotheses of character ary scheme proposed by Rasnitsyn (1988), charac- state transformation discussed in this paper. Exact ters of the mesofurca and mesopostnorum were search methods using additive (ordered) charac- less useful for demonstrating relationships within ters reproduce these relationships with the same Apocrita. Taxa which have retained the posterior number of steps and identical indices of character lobe of the laterophragma and the associated homoplasy and fit. Siricoidea are recognized as a muscle 116 are treated by Rasnitsyn (1988) as paraphyletic group with respect to Apocrita. If relatively derived taxa (especially Diapriidae and Xyelidae are paraphyletic with the rest of Hy- Cynipoidea). Rasnitsyn's hypothesis requires at menoptera (Rasnitsyn 1980, 1988) and least nine independent losses of character 4, if loss Macroxyelinae + Tenthredinoidea are monophyl- of both lobe and muscle is irreversible. This ap- etic, then a minimum of one extra step is required pears to be a poor character for postulating rela- for character 7 of In (loss fUj-spSj). sharp contrast, tionships. Loss of the mesofurcal bridge [1:7] could a monophyletic grouping for Symphyta with indicate support for the monophyly of Cephidae as sister group to the Apocrita Ceraphronoidea, but it is homoplastic within both (Konigsmann 1977) requires 11 additional steps Chalcidoidea and Platygastroidea. Some changes for the same data. The phy logenetic hypothesis for in shape of the axillary lever or mesofurcal bridge Apocrita proposed by Rasnitsyn (1988; Fig. 56) within Apocrita are difficult to categorize or to was shorter than Konigsmann's by only three place into distinct transition series. Distinctive steps (for character 4). Hypotheses based on parsi- features such as the vestigial posterior mony are much shorter for Apocrita, even when laterophragmal lobe of Pelecinidae, Heloridae and characters are ordered, but these hypotheses Proctotrupidae are reductions and possibly con- should not be accepted until they can be included vergent. The development and separation of the in a larger data set with other character systems. axillary lever, fusion of the mesofurca and Monophyly of Orussidae + Apocrita (Vespina metafurca, and the retention (or redevelopment) sensu Rasnitsyn) is supported by 10 external char- of muscle fbl provide strong support for relation- acters of adults (Rasnitsyn 1988) and a shift of the ships within the Aculeata. furcal-coxal muscles from a posterior to an ante- Changes in the mesofurcal-mesopostnotal rior attachment on the discriminal lamella of the complex probably result from an increased em- mesosoma of (Johnson 1988). No features the phasis on the fore wings for flight, reduction of the mesofurcal-mesopostnotal complex discussed here metathorax and fusion of the first abdominal seg- for the of provide support monophyly Orussidae ment. Although the indirect flight muscles are the + Apocrita, without Xiphydriidae. The mesofurca main wing depressors, the direct muscles of the and of is laterophragma Orussidae highly derived mesothorax control flight through modifications and the only potential synapomorphy for Vespina of the pronation and rotation of the wings during 3: loss of muscle 150b, fu is the downstroke and the tension of (Character 2-pn, ) prob- by controlling not lack of in ably homologous. The 2ph-3ph the longitudinal flight muscles, thereby affecting Siricidae and Xiphydriidae has been proposed as the amplitude of the wing beat (Pringle 1957, 1960, a for + 1961 In potential synapomorphy Xiphydriidae ). Apocrita, the axillary lever acts to turn the Siricidae (Whitfield et al. 1992), but this muscle has apex of the mesophragma and the associated fourth also been lost independently in Goniozus axillary sclerite (Weber 1925). This is a shift in (Bethy lidae), Cleptes (Chrysididae), Formicidae and function from the lower Hymenoptera where Ampulex (Sphecidae) (Daly 1963). Monophyly of muscles 150a and 150b twist the posterior lobe of 262 Journal of Hymenoptera Research the laterophragma, initiating a greater force upon ACKNOWLEDGEMENTS the indirect flexor 116, t ). phragmal (muscle ;-pn 2 Gibson and two reviewers The excision of the laterophragma in Xiphydriidae Gary anonymous provided valuable comments and criticisms. We thank Henri Goulet, would reduce the effect of the anterior apodeme (= John Huber, Lubomir Masner and especially Gary Gibson axillary lever) on the posterior lobe. Pringle (1960, (all CNC) for providing specimens and guidance. Jeff 1961 that the and ) proposed pleurosternal axillary Cumming (CNC) and Dave Smith (USNM) provided speci- lever muscles control the power generated by the mens of miscellaneous families that were important addi- of tions to the material examined. This study was supported by indirect flight muscles and thus the amplitude TEXO-6705 of for initial work carried lever Hatch Project J.B.W. the wing beat. The flexibility of the axillary out at Texas A&M University and an NSERC grant to D.C.D. reaches a maximum in the Apiformes in which for subsequent work done in Ottawa. complete separation from the laterophragma re- sults in more powerful leverage upon the associ- LITERATURE CITED ated axillary sclerites of the forewing (Snodgrass 1942). Alam, S.M. 1951. The skeleto-muscular mechanism of clear in taxa from Homologies may not be Stenobracon deesae Cameron (Bracomdae, Hymenoptera) - and borers of India. opposite ends of a phylogenetic tree. For example, An ectoparasite of sugarcane juar Part I. Head and thorax. Muslim Pub- Weber (1925) discusses the mesofurcal ring Aligarh University + lications (Zoological Series) 3: 76 pp. 9 pis. (=mesofurcal bridge) of Siricidae and Vespidae as Brodskiy, A.K. 1992. Structure, function, and evolution of the do look dif- non-homologous. Superficially they II. terga of wing-bearing segments of insects. Organiza- but the intermediate ferent, by examining stages tional features of the terga in different orders of insects. in the 71: 8-28. the homologies can be verified. Changes Entomological Review D. 1975. and classification of the mesofurca and mesopostnotum provide a num- Brothers, J. Phvlogeny aculeate with reference to ber of characters that show informative transitions Hymenoptera, special Mutillidae. Kansas University Science Bulletin 50: 483- from the states in but plesiomorphic Xyelidae, 648. a number of taxa do these 1993. of only by comparing large Brothers, D. J. and J. M. Carpenter. Phylogeny transformation series become apparent. Nonaddi- Aculeata: Chrysidoidea and Vespoidea. Journal of Hy- tive multistate characters are not as useful in build- menoptera Research 2: 227-304. Bucher, G. E. 1948. The anatomy of Monodontomerus dentipes ing a classification because unlinked character Boh., an entomophagous chalcid. Canadian journal of states their inclusive members. How- support only Research (D) 26: 230-281. ever, a of transformation series are 1986. Cladistics of the priori ordering Carpenter, J. M Chrysidoidea (Hy- dependent on previous classifications or on the menoptera). journal of the New York Entomological Society presumed homology of sometimes very divergent 94: 303-330. R. F. 1992. General and function, 33- character states. For example, placement of Chapman, anatomy pp. 67. In The Insects of Australia, 2nd edition. Volume 1. as sister to the Apocrita led to a Cephidae group Melbourne University Press, Carlton. 1137 pp. of the misinterpretation hypostomal bridge by Daly, H. V. 1963. Close-packed and fibrillar muscles of the Ross (1937) and the mesofurcal bridge was not Hvmenoptera. Annals of the Entomological Society of discussed as an important character by Rasnitsyn America 56: 295-306. H. V. 1964. Skeleto-muscular morphogenesis of the (1969, 1980, 1988). Present interpretations of Daly, thorax and wings of the honey bee Apis mellifera (Hy- changes in the mesofurca and mesopostnotum menoptera: Apidae). University of California Publications would have made little sense without the direc- in Entomology 39: 77 pp. tion provided from new classifications for Duncan, C. D 1939. A contribution to the biology of North Symphyta proposed by Gibson (1985) and American vespine wasps. Stanford University Publica- tions Sciences) 8 (1): 272 pp. Rasnitsyn (1988). The current distribution of char- (Biological Gauld, I. and B. Bolton. 1988. The Hymenoptera Oxford Uni- acter states of the mesofurca and mesopostnotum versity Press, New York. 332 pp. within seems with other Apocrita incongruent Gibson, G. A. P. 1985. Some pro- and mesothoracic structures character information. Perhaps as more characters important for phylogenetic analysis of Hvmenoptera, are used to construct improved classifications of with a review of terms used for the structures. Canadian Entomologist 117: 1395-1443. Apocrita, this will also change. 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 Proctotrupoideasensu lata). Australian journal of Zoology of the pleuron in Hymenoptera based on a comparison 33: 761-783. S. 1957. Insect of the skeletomusculature of Xyelidae (Hymenoptera) Pnngle, J. W. Flight. University Press, Cam- and Rhaphidndae (Neuroptera). Canadian Entomologist bridge. 132 pp. S. 1960. of 165: 165-187. Pnngle, J. W. The function the direct flight muscles 1993. Families the in the bee. the 1 2th International Huber, J. T. and H. Goulet. of Hymenoptera of Proceedings of Congress of World. Research Branch, Agriculture Canada. Publication Entomology p. 660.

1894/E. 668 pp. Pnngle, J. W. S. 1961. The flight of the Bumblebee Natural James, H. C. 1926. The anatomy of a British phytophagous History 70 (7): 20-29. chalcidoid of the genus Harmolita {Isosoma). Proceedings Rasnitsyn, A. P. 1969. [Origin and evolution of the lower of the Zoological Society of London [1926] 75-182. Hymenoptera.) Trudy Paleontologicheskogo Instituta, Johnson, N.F. 1988. Midcoxal articulations and the phylog- Akademiya Nauk SSSR 123: 196 pp. of the order Annals the A. P. 1980. and evolution of eny Hymenoptera. of Entomologi- Rasnitsyn, [Origin Hymenoptera. ] cal Society of America 81: 870-881. Trudy Paleontologicheskogo Instituta, Akademiya Nauk SSSR Kelsey, L.P. 1957. The skeleto-motor mechanism of the Dob- 174: 190 pp. son Fly, Corydalus cornutus. Part II. Pterothorax. Mem- Rasnitsyn, A. P. 1988 An outline of evolution of the hy- oirs of the Cornell University Agricultural Experimental menopterous Insects (Order Vespida). Oriental Insects Station 346: 31 pp. + 55 figs. 22: 115-145. Kimsey, L. and R.M. Bohart 1990. The Chrysidid Wasps of the Ronquist, F. and G. Nordlander. 1989. Skeletal morphology World. Oxford University Press: Oxford. 652 pp. of an archaic cynipoid, Ibalia rufipes (Hymenoptera: Korn, W. 1916. Die muskulatur des Kopfes und des thorax Ibaliidae). Entomologica Scandinavica, Supplement 33: 60 von Myrmeleon europaes und ihr metamorphose. pp. Zoologische Jahrbucher 68: 273-330. Ross, H. H. 1937. A generic classification of the Nearctic Konigsmann, E. 1977. Das phylogenetische System der Hy- sawflies (Hymenoptera, Symphyta). Illinois Biological menoptera. Teil 2: "Symphyta." Deutche Entomologische Monographs 34: 173 pp. Zeitschrift (N.S ,) 24: 1-40. Saini, M.S., S.S. Dhillon, and R. Agarwal. 1982. Konigsmann, E. 1978a. Das phylogenetische System der Hy- Skeletomuscular differences in the thorax of winged menoptera. Teil 3 "Terebrantes" (Unterordung and non-winged forms of Camponotus camelinus (Smith) Apocrita). Deutche Entomologische Zeitschrift (N.S. )25: 1- (Hym., Formicidae). Deutche Entomologische Zeitschrifte 55. (N.S.) 29: 447-458. E. 1978b. Das D. B. Konigsmann, phylogenetische System der Hy- Sharkey, M. J. and Wahl. 1992. Cladistics of the menoptera. Teil 4 Aculeata (Unterordung Apocrita). Ichneumonoidea (Hymenoptera). Journal of Hymenoptera Deutche Entomologische Zeitschrift (N.S.) 25: 365-435. Research 1: 15-24. Knstensen, N.P. 1992. Phylogeny of extant Hexapods. pp. Shcherbakov, D. E. 1980. [Morphology of the pterothoracic 125-140. In The Insects of Australia, 2nd edition. Volume pleura in Hymenoptera. 1. Groundplan]. Zoologicheskn 1. Melbourne University Press, Carlton. 1137 pp. Zhurnal 59: 1644-1653. Lawrence, J.F., E.S. Nielson and I.M. Mackerras. 1992. Skel- Shcherbakov, D. E. 1981. [Modifications of the pterothoracic etal anatomy and key to orders, pp. 3-32. In The Insects of pleura in Hymenoptera. 2. Modtficat ions of the groundplan]. Australia, 2nd edition. Volume 1. Melbourne University Zoologicheskii Zhumal 60: 205-213. Press, Carlton. 1137 pp. Snodgrass, R. E. 1910. The thorax of Hymenoptera. Proceed- Maki, T. 1938. Studies on the thoracic musculature of insects. ings of the United States National Museum 39: 37-91 + 16 Memoirs the Science of Faculty of and Agriculture, Taihoku pis. + Imperial University 24: 343 pp. 17 pis. Snodgrass, R. E. 1927. Morphology and mechanism of the Markl, H. 1966. Peripheres Nervensystem und Muskulatur insect thorax. Smithsonian Miscellaneous Collections 80 lm der Arbeiterin Thorax von Apis mellifera L., Formica (1): 108 pp. polyctena Foerster und Vespa vulgaris L., und der Snodgrass, R. E. 1942. The skeleto-muscular mechanisms of Grundplan der Innervierung des Insektenthorax the honey bee. Smithsonian Miscellaneous Collections 103 Zoologische Jarbucher 83: 107-184. (2) 120 pp. Matsuda, R. 1970. Morphology and evolution of the insect Tulloch, G. S. 1935. Morphological studies of the thorax of the thorax. Memoirs of the Entomological Society of Canada 76: ant. Entomologica Americana 15: 93-130. 431 pp. Weber, H. 1925. Derthorax derHornisse. Zoologische jarbucher 1944. external + Michener, CD., Comparative morphology, 67: 1-100 4 pis. phylogeny, and a classification of the bees (Hy- Weber, H. 1927. Die Gliedrung der Sternalregion des menoptera). Bulletin of the American Museum of Natural Tenthredinidenthorax. Ein Beitrag zur vergleichenden 82: 157-326. History Morphologie des Insecktenthorax. Zeitschrift fur I.D. 1992. 916-1000. In The Naumann, Hymenoptera. pp. ivissenschaftliche Insektenbiologie 22: 161-198 + 6 pis. Insects Australia, 2nd edition. Volume 1. B. of Melbourne Whitfield, J. 1992. Phylogeny of the non-aculeate Apocrita University Press, Carlton. 1137 pp. and the evolution of parasitism in the Hymenoptera. Naumann, I.D. and L. Masner. 1985. Parasitic wasps of the journal of Hymenoptera Research 1: 3-14. a from Australia proctotrupoid complex: new family Whitfield, J. B„ N. F. Johnson and M. R. Hamerski. 1989. and a to key world families (Hymenoptera: Identity and phylogenetic significance of the 264 Journal of Hymenoptera Research

W. 1985. Neue oberliassiche und die metaposmotum in nonaculeate Hymenoptera. /birni/so/ Zessin, Apocnta der the Entomological Society of America 82: 663-673. Phylogenie Hymenoptera (Insecta, Hymenoptera). Deutsche Entomologische Zeitschrift, N.S. 32:129-142.

Abbreviations used and Nordlander (1989) in brackets. Appendix 1 List of abbreviations used for figures. by Ronquist af anterior furcal arm [91] PN, mesopostnotum ap anterior process/axillary pph pseudophragma |pph]

lever of ) laterophragma (pn ; sps spina

cenchrus t, mesonorum

t metanotum ex coxa 3 dl discrimenal lamella [dl] ume upper mesepimeron ventral of lateral furcal arm fb mesofurcal bridge vp process mesofurcal first fp process [91] lph phragma second fu, metafurca 2ph phragma FWB fore wing base 60 attachment line of mesopostnotum horizontal of mesofurca (PN ) to scutellum (t ) [60) h P plate [90] 2 ; HWB hind wing base hi attachment line of mesopostnotum to metanotum If lateral furcal arm ] (PN,) (t,) [61] [fu ; of pn 2 laterophragma mesopostnotum (posterior face of lobe =t,) [pnap]

Characters and states are Appendix 2. Character states of the mesofurcal-mesopostnotal complex attributed to Hymenoptera. within a taxon. discussed in text. Superscript values refer to other observed states Volume 3, 1 994 265

dl

1-2. lateral habitus of Figs. Pamphiliidae. l,Pamphilius sp., mesosoma; 2, Acantholyda sp., sagittal section of mesosoma. Muscle 127 and removed location indicated by dashed line. Abbreviations as listed in Table 1 and Appendix 1. 266 Journal of Hymenoptera Research

3 XYELIDAE

sps 2

PERGIDAE 4 XYELIDAE 5

Figs. 3-5. 3, Xyela minor, section of mesothorax. skeletal oblique subsagittal 4, Xyela minor, components of Fig. 3, musculature removed. 5, Acordulecera frontolateral view of MF-MPN sp., complex. Abbreviations as listed in Table 1 and Appendix 1. Volume 3, 1 994 267

6 PAMPHILIIDAE

7 DIPRIONIDAE 8 CIMBICIDAE

6-8. 6, a, dorsal view of MF-MPN lateral of Figs. Pamphilius sp.: complex; b, view MF-MPN complex. 7, Diprion sp., supralateral view of MF-MPN Zaraea complex. 8, 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,

view. Tremex dorsal lateral 1 supralateral 11, sp.: a, view; b, view. Abbreviations as listed in Table 1 and Appendix Volume 3, 1 994 269

12 XIPHYDRIIDAE

1 3 ORUSSIDAE

14M0N0MACHIDAE 15 VANHORNIIDAE

12-15. abdommalis: Figs. MF-MPN complex. 12, Xiphydria 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

ICHNEUMONIDAE 16 SCELIONIDAE 17

18 MUTILLIDAE 19 RHOPALOSOMATIDAE

Figs. 16-19. MF-MPN complex. 16, Sparasion sp., supralateral view. 17, Megarhyssa sp. (Ichneumorudae): a, laterophragma view. Abbrevia- (inner view); b, supralateral view. 18, Mutilhdae (male), supralateral view. 19, Rhcpalosoma sp., supralateral tions as listed in Table 1 and Appendix 1. Volume 3, 1 994 271

20 BLASTICOTOMIDAE 21 TENTHREDINIDAE 22 IBALIIDAE

23 TRIGONALYIDAE 24STEPHANIDAE 25 MEGASPILIDAE

26GASTERUPTIIDAE 27 EVANIIDAE 28 PLATYGASTRIDAE

20-28. Figs. MF-MPN complex of Apocrita: right laterophragma (upper figure); mesofurca in dorsal view (lower figure). 20, Blasticotoma Nematus Ibalm sp.; 21, sp.; 22, 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 35 SCLEROGIBBIDAE HYBRIZONTINAE

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

47-50. Figs. 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 of area magnifications highlighted in upper figure. Volume 3, 1 994 275

51-54. in lateral view with of connection to 51 Figs. 51-52, Mesopostnotum magnification upper mesepimeron. , 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). 277 Volume 3. 1994

for based on with characters of the MF-MPN Fig. 56. Phylogenetic hypothesis Apocrita Rasnitsyn (1988), complex superim- in text. of Peradenndae and Vanhorniidae Masner comm., 1993). posed. Characters and states discussed Placement by (pers. in were not examined for internal characters Superfamily names follow Huber and Goulet (1993), those names parentheses are shaded circles indicate and question marks indicate unknown character states. Solid circles unique apomorphies; first followed derived state convergence; open circle indicates reversal. Multiple states indicate groundplan by proposed 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

of (YT) Biological Control Research Institute, Fujian Agricultural College, Fuzhou, Fujian, People's Republic 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- Approximately 1,200 specimens of Chinese conid genus whose species are parasitic upon Ascogaster were examined in this study, mostly microlepidoptera, predominantly Tortricidae. from the collections of the Fujian Agricultural Some species have been recorded frequently as College, the Zhejiang Agricultural University, the parasitoids of several economically important in- Zoology Research Institute of Academia Sinica, sect pests and have been considered for use in the Taiwan Agricultural Research Institute and biological control programs. About 110 species of the American Entomological Institute. The fol- Ascogaster have been recorded in the world. lowing acronyms are used to identify collections Huddleston (1984) revised the 30 Palaearctic spe- that provided specimens for this study and several cies; Tobias (1986a, 1986b, 1988) and Papp (1989) as the depositories for type material: added 12 species to this region. Shaw (1983) re- AEIG American viewed 11 species for the Nearctic Region and Entomological Institute, Gainesville, Florida, USA described a related new genus, Leptodrepana (4 BM Bishop Museum, Honolulu, Haiwaii, USA species). Walker and Huddleston (1987) recorded CNC Candian National Collection, Agriculture Canada, 12 of in Zealand. fauna species Ascogaster New The Ottawa, Canada of the Indo-Australian Region is largely unknown, ELKU Entomological Laboratory, Kyushu University, although Baker (1926) and Szepligeti (1905, 1908) Fukuoka, Japan FAC Control Research Institute, Fujian described a few species. The Ascogaster fauna of Biological Agricul- tural College, Fuzhou, China China is especially poorly studied with only five HNHM Hungarian Natural History Museum, Budapest, recorded, three from mainland China species Hungary (Fahringer 1934; He, et al 1989) and two from IRSNB Institut Royal des Sciences Naturelles de Belgique, Taiwan (Sonan 1932). In this work 23 species of Brussels, Belgium TARI Taiwan Research Tai- in China are treated, of which 13 Agricultural Institute, Wufeng, Ascogaster species wan are new to science and six are recorded for the first UEI Entomological Institute, Hokkaido University, time from China. Sapporo, Japan USNM National Museum of Natural History, Washington, D. C, USA "Throughout this paper, China includes mainland China and ZAU Department of Plant Protection, Zhejiang Agricul- Taiwan. tural University, Hungzhou, China 2S0 Journal of Hymenoptera Research

distance and the ZRJ Zoology Research Institute, Academia Simca, Beijing, respectively, to the ocellar-ocular China diameter of the posterior ocellus. The ratio CL/ CW is carapace length (CL) to carapace width also examined available Palaearctic and We (CW). Shaw (1983) and Huddleston (1984) dis- material in the U. S. National Indo-Australian cussed the important taxonomic characters of Museum of Natural and some mate- of the History type Ascogaster in detail and we have used most - rial from other institutions as follows: USNM for specific characters used by Huddleston (1984) abdominatorDahlbom,albitarsusRemhard,annulans the description of new species in this study (see Nees von atamiensis Ashmead (holo- Esenbeck, discussion under generic diagnosis). All dates from with bidentida Wesmael type, synonymized by labels on specimens examined have been stan- bicarinata Huddleston 1984), Herrich-Shaffer, dardized thus, 1-1-93, to avoid confusion. bidentida Wesmael, brevicornis Wesmael, canifrons Wesmael, cinctus Baker (holotype), coelioxoides BIOLOGY detectus Baker (Baker) (holotype), consobrnw Curtis, distinctus Baker (holotype), (holotype), fullawayi Members of Ascogaster, like all the chelonines, Baker (holo- (Baker) (type series), inconspicuus are internal egg-larval parasitoids of microlepi- Baker intensus Baker ( laeviventris their type), holotype), doptera, principally Tortricidae. They lay Baker luzonensis (holotype), longulus (holotype), eggs into the host egg, but their larval develop- Baker Baker (holotype), maculaticeps (holotype), ment is delayed at the first instar until the host modestus Baker malayanus (Baker) (holotype), (ho- larva is mature. Little detailed work has been done Baker series), lotype), philippinensis (type on the biology of Ascogaster. Huddleston (1984) reticulata Watanabe quadradentata Wesmael, (type and Shaw (1983) reviewed the hosts of Ascogaster vividus series), rufidens Wesmael, rufipes (Latreille), for the Palaearctic and Nearctic Regions respec- Baker BM - Fullaway and (holotype); argentea tively. Cox (1932), Rosenberg (1934) Boyce - (holotype); IRSNB of the (syntypes); rugosa Fullaway (1936) gave fairly detailed accounts biology armata Wesmael bidentida Wesmael (lectotype), of A. quadridentata Wesmael. A research group in brevicorinis Wesmael (lectotype), (lectotype), Japan has recently studied in some detail host- Wesmael limitatus Wesmael canifrons (lectotype); searching, host-preference, mating and oviposi- with scabricula (syntypes, synonymized (Dahlbom) tion behavior of A. reticulata W atanabe, an impor- Huddleston Wesmael by 1984), quadridentata (type- tant parasitoid of the smaller tea tortix, Adoxophyes series), rufidens Wesmael (type-series), varipes el al. 1982; sp., in Japan (Kainoh 1986, 1988; Karnoh, UEI - Watanabe Wesmael (type series); epinotiae Kainoh, et a/.1991; Kainoh and Tamaki 1982, 1988; with (type series, synonymized quadridentata Kainoh, et fl/,1990; Kainoh, et al. 1989; Kamano, et Huddleston reticulata Watanabe Wesmael by 1984); al. 1989; Kawakami 1985; Kawakami and Kainoh ELKU Huddleston (holo- (type series); -longicornis 1985, 1986). Little is known at present about the with Sonan, new type, synonymized formosensis biology of the Chinese Ascogaster species, how- - nilena synonymy); HNHM Papp (holotype). ever He et al. (1987) gave a full host list of A. termi- The general braconid morphological reticulata in mainland China. nology and measurements used in this study are mostly after van Achterberg (1988). The terminol- SYSTEMATICS ogy used for wing veins is illustrated and explaned The Cheloninae are from all in terms are based distinguished Sharkey ( 1988). Microsculprure other subfamlies of Braconidae the combina- on Harris (1979). Two character states of the ocel- by as tion of the first three metasomal terga lar triangle are defined after Huddleston (1984) having fused to form a 21-26), having a follows: where a straight line drawn between the carapace (Figs. carina and three submarginal anterior borders of posterior ocelli also touches complete postpectal cells in the Two tribes of the the anterior ocellus, the ocelli are referred to as forewing (Fig. 1). Cheloninae, Chelonini and Phanerotomini, are being 'on line'; where a line between the posterior in the Palaearctic and Indo- Austrialian ocelli passes behind the anterior ocellus without represented is di- to as 'not regions. In the Phanerotomini the carapace touching it, the ocelli are referred being vided into three two transverse su- on line'. The measurements CO and OD refer, tergites by Volume 3. 1 994 281 tures; in the Chelonini the carapace has no trace of examined and we have chosen to follow van sutures. The Chelonini are represented in the Achterberg's synonymy. This concept makes Palaearctic and Indo-Australian regions by three Ascogaster an apparent polyphyletic group and we genera, Ascogaster Wesmael, Megascogaster Baker feel that a thorough phylogenetic study of the and Chelonus Panzer. The former two genera can entire genus is needed before the correct position be separated from Chelonus by having the first of Leptodrepana and the many other unusual forms submarginal cell always separated from the discal from the Indo-Australian Region can be settled. cell by vein 1-Rs+M (Rs+M of some authors) (Fig. That is beyond the scope of this limited study aims to the 1). Many authors (e.g. Baker 1926; Shaw 1983; which only identify species from pre- Huddleston 1984; Walker and Huddleston 1987; viously unknown area of China. van Achterberg 1990b; Zettel 1990) state that Ascogaster is also characterized by having hairless Ascogaster Wesmael eyes. However, we have described one new spe- Wesmael 1835:226. cies, A. setula from Taiwan, with eyes densely and Ascogaster Type-species: Ascogaster mstabilis Wesmael \-abdonunator (Dahlbom)], subse- distinctly setose (Fig .8), but with most of the other quently desigated by Foerster (1862). characters of typical,4scogasfer. We include it within Cascogaster Baker 1926:482. Type-species: Cascogaster fullawayi until more with this unusal Ascogaster species Baker, original desigation, syn. by Watanabe (1937). character have been found. Megascogaster Baker is Leptodrepana Shaw 1983:37. Type-species: Leptodrepana van distinguished from Ascogaster only by having a opuntiae Shaw, original designation; syn. by Achterberg (1990b). very elongate carapace (distinctly longer than head and thorax) and a slender stigma in the forewing. Diagnosis.—Distinguished from all other bra- The taxonomic position of Megascogaster Baker conid the first three must wait until more material from the Indo- genera by having: (a) metasomal fused into a with- Australian region has been studied. Shaw (1983) terga rigid carapace out transverse sutures; and (b) the with cited thirteen characters for distinguishing forewing three submarginal cells and vein 1-Rs+M present, Ascogaster and Leptodrepana, however intermedi- the first and discal cells. ate forms of most characters occur in many unde- separating submarginal scribed Indo-Australian Ascogaster which we have

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 broadest at about middle in dorsal ventral of shorter 12(11 ) Carapace short, oval, point view; opening carapace 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 and new — Tang Marsh, species Carapace apically pointed, with a tubercle at apex; clypeus smooth, sparsely punctate, its apical border pointed with a medial tooth 18 face and acutus and Marsh new —18(17) Carapace oval; temple areolate-rugose Tang 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 clavate in dorsal not — Temple rounded, distinctly longer than eye in dorsal view; carapace longer, view, so deep in lateral view hei Tang and Marsh, new species

Ascogaster acutus Tang and Marsh, Paratypes.— BEIJING CITY: 1 female, T. L. 1 H. new species Lugouqiao, IX-57, Chen; male, IX-78, J. 1 Fig. 23 He. GUIZHOU PROVINCE: female, Guiyang, V-13-85, Q. H. Luo. FUJIAN PROVINCE: 1 female, Female.—Length of forewing 2.5-2.7 mm, of Shaowu, VII-20-45, H. F. Chao; 1 female, Shanhan, body 3.0-3.2 mm. VII-19-21-88, Y. Ma. TAIWAN: 1 female Duruanpi, Head. —Antenna with 28-32 flagellomeres, VIII-5-82, K. C. Chou and C. N. Lin; 1 female, 1 rV- slightly dilated medially, medial flagellomeres Wushe, IV-13-83, H. Townes; female, Wushe, H. 1 H. slightly broader than long, apical ones longer than 19-83, Townes; female, Wushe, V-3-83, 2 H. 1 broad. Temple slightly contracted behind eyes, Townes; females, Wushe, V-10-83, Townes; distinctly longer than length of eye in dorsal view. male, Wushe, V-15-83, H. Townes. Deposited in Occiput strongly concave. Ocelli small, on line, AEIG, FAC, TARI, ZAU and ZRI. 00= 4.0 OD. Frons Host. —Unknown. strongly depressed, finely — rugose or smooth, medial carina distinct, extend- Distribution. China (Beijing City, Guizhou ing from uper part of face betwen antennae to Province, Fujian— Province),Taiwan. anterior ocellus and strongly expanded between Remarks. This species is very close tofidlawayi antennae into an erect triangular lamella. Eyes (Baker) from which it can be distinguished by the mentioned in the protuberant, glabrous. Face slightly protuberant, characters — key. about twice as broad as high, strongly areolate- Etymology. The specific name is from the rugose. Clypeus punctate, strongly divided from Latin acutus meaning pointed in reference to the face, apical border produced medially into a small pointed tubercle at the apex of the carapace. pointed tooth. Mesosoma.—Pronotum slightly projecting in Ascogaster albitarsus Reinhard front of mesonotum, dorsolaterally rugose-punc- tate. Notauli not distinct; mesonotum completly Ascogaster albitarsus Reinhard 1867: 364. Lectotype male, POLAND: Gdansk by Huddleston 1984). coarsely rugose. Precoxal suture indistinguish- (designated Shenefelt 1973: 815; Huddleston 1984: 364; Papp 1989: able; mesopleuron strongly rugose. Propodeum 297. divided a transverse coarsely areolate-rugose, by Ascogaster leptopus Thomson 1874: 584. Lectotype female, carina which is rasied into a medial pair and a SWEDEN (designated by Huddleston 1984 and syn. by Hellen lateral pair of dentate flanges. Hind coxa strigate. 1953) Vein r of forewing about 1.0-1.5 times as long as 3- RS. Diagnosis. —Length of foewing 3.3-3.5 mm, of — 3.9-4.0 mm. Antenna with 36-38 Metasoma. Carapace (Fig. 23) short, with body dilated pointed tubercle at apex, areolate-rugose. Ventral flagellomeres, slightly medially (female); ocelli almost on face about 1 .5 times as broad opening of carapace longer, not distinctly in front line; as with its of apex. Ovipositor concealed. high, rugose-punctate; clypeus apical — border and reflexed Color. Black; carapace yellow at base; all legs produced medially slightly notauli yellow except hind coxa at base, tibia at apex and forwards; distinct, fovealate, coalescing tarsi black. posteriorly in a large areolate-rugose area, rest of mesonotum Male.—Same as females except antenna not densely areolate-punctate; precoxal suture rather dilated medially, smaller yellow spot at basal strongly areolate-rugose anteriorly, weak carapace. posteriorly, mesopleuron anterodorally rest of Holotype Female.—BEIJING CITY: Mentougou, strongly rugose, mesopleuron finely punc- hind dark with base of VIII-25-81, M. S. Shi. Deposited in ZAU. tate; leg predominately 284 Journal of Hymenoptera Research

close to but it tibia and tarsus pale-yellow; hind coxa finely punc- species is structurally very bidentula, be the more massive head tate; carapace long, clavate and widest in posterior can distinguished by third. and the more elongate carapace." However, our have the heads not so mas- Specimens Examined.—HEBEI PROVINCE: 2 specimens generally females. SICHUAN PROVINCE: 2 females. sive as described by Huddleston. In Chinese speci- Additional Specimens Examined.—WESTERN mens examined here, the temple is only 1.0-1.3 EUROPE: 1 male, det. by Huddleston, 1983. times as long as eye in dorsal view. Sonan (1932) Host.—A tortricid moth on pine tree. named this species as arisanicus, however, the Distribution. —China (Heibei Province, Sichuan holotype is labelled 'Ascogaster artsensis Sonan'. Province), Korea and some European countries. Ascogaster armatoides Tang and Marsh, Remarks.—Our specimens from China have new species the mesopleuron with a polished impuncate area Fig. 19 posteroventrally, but otherwise agreeing with the Female. — of 3.5-3.6 of redescription of Huddleston (1984) Length forewing mm, 4.2-4.4 mm. body — Ascogaster arisanica Sonan Head. Antenna with 42 flagellomeres, mod- dilated medial about Fig. 3 erately medially, segments as broad as long. Temple constricted behind eyes, 79. TAI- Ascogaster arisanicus Sonan 1932: Holotype male, slightly shorter than length of eye in dorsal view. WAN: Alishan Watanabe 1937: 77 (as of (=Arisan). syn. Occiput strongly concave. Ocelli very small, on

: the of rufipes (Latreille), Chou 1981 72 (notes on locality line, OO = 4.0 OD. Frons strongly excavate behind types). antennae, smooth, with a distinct medial carina Ascogaster arisanica Sonan: emended by Huddleston 1984: into erect 365. which is expanded between antennae an triangular flange (Fig. 19). Eyes glabrous, strongly — about 1.5 Diagnosis. Length of forewing 2.9-3.2 mm, of protuberant. Face slightly protuberant, body 3.3-3.7 mm. Antenna with 34-38 times as broad as high, strongly areolate-rugose. not flagellomeres, moderately dilated medially (fe- Clypeus dull, rugose-punctate, very distinctly male); ocellar triangle obtuse, ocelli almost on line; divided from face; apical border convex, not pro- and without distinct tooth or tu- face (Fig. 3) about 1.5 times as broad as high, duced medially densely punctate or rugose-punctate; apical bor- bercle. — der of clypeus produced, with a distinct medial Mesosma. Pronotum slightly projecting in excision flanked by two small tubercles; notauli front of mesonotum, areolate-rugose laterally shallow, coarsely rugose, coalescing posteriorly in Notauli indistinct; mesonotum strongly rugose. an areolate-rugose area, rest of mesontum punc- Precoxal suture indistinct; mesopleuron com- tate or rugose-punctate; precoxal suture broad, pletely areolate-rugose. Propodeum strongly ar- shallow, foveolate; rest of mesopleuron sparsely eolate-rugose, divided into dorsal and posterior punctate except dorsally rugose; hind coxa finely surfaces by a transverse carina which is raised into blunt teeth and a of punctate; carapace elongate, broadest in distal a pair of medial laterally pair half; ventral opening of carapace conspicuously in less stout teeth. Hind coxa strongly strigate. Vein front of apex. r of forewing about as long as 3-RS. — = Specimens Examined.—ZHEJIANG PROV- Metasotna. Carapace short, CL/CW 1.6, INCE: 16 females, 9 males. SICHUAN PROV- oval in dorsal view, somewhat pointed in INCE: 2 females. YUNNAN PROVINCE: 1 male. lateralview, areolate-rugose. Ventral opening of GUANGXI PROVINCE: 1 female. HAINAN carapace short, in front of apex. Ovipositor PROVINCE: 1 female. TAIWAN: 115 females, 35 convealed. males. Male. —Unknown. Host. —Unknown. Holotype Female.—GUANGXI PROVINCE: Distribution. —China V-18-82, H. He. in ZAU. (Zhejiang Province, Longzhou, — J. Deposited Sichuan Province, Yunnan Province, Guangxi Paratypes. INDIA: 4 females, Ammatti, S. Province, Hainan Province), Taiwan, Japan. Coorg, 11-52, P. S. Nathan. Deposited in CAN, Remarks.—Huddleston (1984) stated that "This USNM. Volume 3, 1994 285

C. 1 Host. —Unkown. (Wuyishan), VII-11-82, J. Huang; female, Distribution.—China (Guangxi Province), In- Wuyishan, VI-80, H. F. Chao; 2 females, Shanguan dia. (Wuyishan), VII-5-85, D. H. Huang and G. Zheng; S. Remarks.—This species is very close to arniata 1 female, Wuyishan, VII-15-86, J. Wang. Wesmael from which it can be distinguished by its GUANGXI PROVINCE: 1 female, Tainlin, V-30- H. He 1 H. longer antenna and less massive head. 82, J. ; female, Longsheng, VI-24-84, J. He. 5 Etymology.—The specific name is in reference ZHEJIANG PROVINCE: females, Songyang, arniata. H. and H. L. 1 to the similarity of this species to VII-15-89, J. He Chen; female, Tianmushan, VI-17-83, Y. Ma. ANHUI PROV- Ascogaster chaoi Tang and Marsh, new species INCE: 1 female, Huangshan, VII-30-86, S. C. Zhang. JILING PROVINCE: 1 female, Dongliao, VII-20- — X. in Female. Length of forewing 3.0-3.4 mmm, of 31-88, M.— Luo. Deposited FAC, USNM, ZAU. body 3.5-4.0 mm. Host. Unknown.— Head.—Antenna with 40-42 flagellomeres, Distribution. China (Jiling Province, Anhui strongly dilated medially, tapered at apex, medial Province, Zhejiang Province, Guangxi Province, segments slightly broader than long, apical seg- Fujian Province).— ments about as long as broad. Temple constricted Remarks. This species is close toquadridentata behind eyes, slightly longer than length of eyes in from which it can be easily distinguished by its dorsal view. Occiput deeply concave. Ocelli small, longer antenna, straight and long gena, and yel- OO =3.5 OD, on line. Frons moderately depressed, low hind coxa. —This is for Prof. H. finely punctate, with distinct medial carina. Eyes Etymology. species named in slightly protuberant, glabrous without distinct F. Chao, Fujian Agricultural College, apprecia- to the tax- setae. Gena long and straight in face view. Face tion for his outstanding contribution as of in China. slightly protuberant, about 1.5 times as broad onomy parasitic Hymenoptera high, areolate-rugose. Clypeus protuberant, punc- tate, distinctly divided from face; apical border Ascogaster consobrina Curtis 27 slightly impressed laterally, produced medially Figs. 20, into stout dentate flange. consobrmus Curtis 1837: 672. GREAT Mesosoma.—Pronotum projecting in front of Ascogaster Holotype male, BRITAIN: Shenefelt 1973: 819; Huddleston mesonotum, dorsolaterally areolate-rugose. England. 1984: 367; Papp 1989: 297. Notauli indistinguishable; mesonotum strongly — areolate-rugose. Mesopleuron completely rugose Diagnosis. Antenna with 32-34 flagellomeres, so that precoxal suture not distinct. Propodeum slightly dilated medially (female); ocellar triangle strongly areolate-rugose, with a medial transverse obtuse, ocelli almost on line; face (Fig. 20) about 1 .5 carina raised into a pair of medial stout dentate times as broad as high, areolate-punctate, hairs and a pair of lateral less stout dentate. Hind coxa downwards; apical border of clypeus not retracted, strigate. Vein r of forewing 1.0-1.5 times as long as medially produced into a broad, blunt point, with- 3-RS. out impression or tubercle; notauli distinct, Metasoma.—Carapace shorter, CL/CW = 1.5- foveolate, rest of mesonotum punctate except with 1.7, somewhat pointed in dorsal view, rounded a broad areolate-rugose area posteromedially; and deep in lateral view, areolate-rugose. Ventral precoxal suture distinct foveolate, mesopleuron opening of carapace short, distinct in front of apex. (Fig. 27) dorsally finely rugose-punctate except for Hypopygium short. Ovipositor sheath clavate. a polished and impunctate area just dorsal precoxal Color. —Black; carapace yellow at base; all legs suture, ventrally always sparsely punctate, shin- yellow except hind femur and tibia black at apex. ing; carapace rather long, clavate, widest in poste- Male. —Unknown. rior third, with a downwardly directed anterior Holotype Female.— FUJIAN PROVINCE: flange; hind coxa punctate, sometimes weakly Huanggangshan, VII-14-85, Y. Q. Tang. Deposited rugose dorsaly; hind leg yellow except coxa at inFAC. base black, apex of femur and sometimes of tarsus Paratypes.—FUJIAN PROVINCE: 1 female, infuscate. Length of forewing 3.3-3.6 mm, of body same as holotype; 6 females, Tongmuguan 3.8-4.3 mm. 286 Journal of Hymenoptera Research

Specimens Examined.—ZHEJIANG PROV- as 3-RS. = INCE: 1 female. TAIWAN: 89 females, 12 males. Metasoma.—Carapace long, CL/CW 2.2, ar- Additional Specimens Examined.—Western Eu- eolate-rugose, but finely and sparsely punctate ropean countries, 6 females, 10 males, det. by apically, with a distinct dorsal prominence at base. Huddleston,— 1983. Hypopygium large and broad. Ovipositor long Hosts. No reared material was examined. slender, upcurved.— Shenefelt (1973) recorded Chelonia caja and Gelechia Color. Black; legs yellow-brown except hind tarsis tibia vulgella as its hosts, but these records need to be coxa at base and black, and hind confirmed. infuscate with a pale-yelow band medially; cara- Distribution. —China (Zhejiang Province), Tai- pace with yellow-pale spots at anterolateral sides; wan., Japan, Korea, and several European coun- papli yellow-pale, tries (see Huddleston 1984: 367) Males. —Same as female except antenna longer, Remarks.—All specimens examined agree well not dilated medially, with 30-31 flagellomeres and = with the redescription given by Huddleston (1984) carapace more slender, CL/CW 2.3-2.5, more except for the slightly shorter medial flagellar rounded and deeper apically, without a dorsal segments of the female, which are generally about prominence at base. Length of forewing 3.6-3.9 as long as broad. mm, of body 4.3-4.5 mm. Holotype female.—ZHEJIANG PROVINCE: H. He. in Ascogaster dimorpha Tang and Marsh, Songyang, VII-18-31-89, J. Deposited new species ZALL Fig. 4 Paratypes.—ZHEJIANG PROVINCE: 3 males, H. 5 Songyang, VII-15-17-89, J. He. TAIWAN: Female. Length of forewing 3.8 mm, of body males, Meifeng 2150 m, -24-26-81, K. S. Lin and W. 4.4 mm. S. Tang (TARI); 1 male, Meifeng, VI-22-26-83, K. S. Lin and S. C. Lin. in ZAU. Head. Antenna, with 26 flagellomeres, flagel- — Deposited TARI, lum slightly dilated medially, all segments longer Hosts. Unknown. than broad. Temple more or less rounded behind Distribution. —China (Zhejiang Province), Tai- eyes,, slightly longer than eye in dorsal view. Oc- wan. = — ciput strongly concave. Ocelli large, OO 4.5 OD; Remarks. This species belongs to the caucasica- ocellar triangle acute, ocelli not on line. Frons group (sensu Huddleston 1984) by virtue of its behind antenna moderately excavated, smooth punctate face, hair on the upper part of the face anteriorly, rugose-punctate posteriorly. Eyes mod- pointing upwards and the bidentate clypeus. It erately protuberant, nearly glabrous with only appears to be related to caucasica Kokujev and few scattered minute setae. Gena contracted be- bicarinata Herrich-Schaffer, but is easily distin- low in face view. Face (Fig. 4) slightly protuberant, guished from the later two species by the charac- about twice as broad as high, densely and regu- teristics of the carapace.— larly rugose-punctate, moderately hairy, the hairs Etymology. The specific name refers to the on the upper part of the face pointing upwards. sexual dimorphism of the metasomal carapace. Clypeus moderately protuberant, rugose-punc- tate, but less densely than that of face, its apical Ascogaster formosensis Sonan border strongly incised medially. Fig. 5 Mesosoma.—Pronotum projecting in front of Sonan 1932:78. TAI- mesonotum, rugose-foveolate dorsolaterally. Ascogaster formosensis Holotype female, WAN: Arisan, Kunkiko. Watanabe 1937:76; Shenefelt Notauli not very distinct, mesonotum finely punc- 1973: 822; Chou 1981: 72 (notes on the locality of types); tate except with an areolate-rugose area Papp 1989: 297. Precoxal suture shallow posteromedially. Ascogaster formosanus (!): Watanabe 1934: 198. foveolate, rest of mesopleuron punctate. Ascogaster longicomis Huddleston 1984: 368. Holotype fe- Mt.Tachibana New Propodeum slightly impresed and punctate medi- male, JAPAN: (ELKU). synonymy. ally, strongly areolate-rugose laterally, but with- out distinct tubercles. Hind coxa smooth and Diagnosis. —Antenna with 45-49 flagellomeres, medial ocelli on line or shining. Vein r of forewing about 2.5 times as long segment weakly dilated; Volume 3, 1 994 287

on face 1.2- Examined.—TAIWAN: 2 females. almost line; (Fig. 5) rugose-punctate, Speimens — 1.5 times as broad as high; clypeus with its apical Additional Specimens Examined. Holotype of border rounded except its medial 1/4 weakly fullawayi (Baker), female, PHILIPPINES: Baguio, emarginate; notauli foveolate, rest of mesonotum Benguet (USNM). Holotype of fullawayi var. densely punctate except posteromedially areolate- maquilingensis (Baker), female, PHILIPPINES: Mt. rugose; mesopleuron coarsely rugose so that Makiling, Luzon (USNM). Paratype of fullawayi precoxal suture not distinct; propodeum areolate- (Baker), 1 female, PHILIPPINES: Imugin, N, rugose with 4 prominent tubercles; hind coxa finely Viscaya (USNM). at oval in Hosts.—Unknown. punctate; carapace always yellow base, — dorsal view, clavate in lateral view, ventral open- Distribution.— Taiwan, Philippines. ing of carapace distinctly in front of apex. Larger Remarks. Two Taiwanese speicmens exam- species, length of forewing 5.0-5.4 mm, of body ined agree well with the type of fullawayi (Baker) 5.8-6.4 mm. except that the antennae are shorter. This species Specimens Examined.—YUNNAN PROVINCE: is the only Philippine Ascogaster described by Baker 1 female. TAIWAN: 15 females, 7 males. (1926) found in China. It is distinguished from the Additional Specimens Examined.—NEPAL: 1 other Chinese Ascogaster by the characteristics of female, 1 male. INDIA: 6 females, 6 males. JAPAN: the carapace and sculpture of the head. Holotype of longicorinis Huddleston. Hosts. —Unknown. Ascogaster gibbosa Tang and Marsh, Distribution. —China (Yunnan Province), Tai- new species wan, India, Japan, Nepal. Fig. 28 Remarks.—This is species easily distinguished — from other species of Chinese Ascogaster by its Females.- Length of forewing 3.0-3.2 mm, of conspicuously long antennae and by the charac- body 3.5-3.8— mm. teristics of its face and clypeus. It is very similar to Head. Antenna with 37-40 flagellomeres, philippinensis Baker from which it differs only in slightly dilated medially, medial segments about the sclupture of the mesonotum. We treat it here as as long as broad. Temple rounded behind eye, a valid species given no intermediate forms are slightly shorter than eye in dorsal view. Occiput known. concave. Ocelli small, OO = 3.5-4.0 OD, on line. Frons moderately depressed, smooth, with a weak Ascogaster fullawayi (Baker) medial carina. Eyes protuberant, without distinct Figs. 7, 24 setae. Gena in face view contracted. Face about 1.5 times as broad as high, coarsely areolate-rugose. Baker 1926: 483. female Cascogaster fullawayi Holotype (cited Clypeus distinctly divided from face, punctate, as male), PHILIPPINES: Baguio, Benguet. apical border produced medially into a strongly Ascogaster fullawayi: Shenefelt 1973: 822. pointed large tooth. Mesosoma.—Pronotum in — slightly projecting Diagnosis. Antenna with 34-35 flagellomeres, front of mesonotum, dorsolaterally areolate-rug- slightly dilated medially; scrobes outwardly mar- ose. Notauli present, areolate-rugose; mesonotum gined by high, thin and complete carina middle of (Fig. 28) between notauli strongly protuberant, which is raised into sharply angulate teeth; ocelli rugose-punctate, rest of mesonotum areolate-rug- almost on line; face (Fig. 7) narrow, less 1.5 times ose. Precoxal suture indistinguishable; as broad as long, strongly and coasely rugose; mesopleuron completely areolate-rugose. clypeus smooth and shining, sparsely punctate, Propodeum strongly areolate-rugose, divided by apical border acute; mesonotum between notauli transverse carina which is raised medially and swollen, strongly areolate-rugose; mesopleuron laterally into prominent dentate flanges. Hind coarsely areolate-rugose, precoxal suture indis- coxa dorsally strigate. Vein r of forewing about as tinguishable; hind coxa rugose-punctate; carapace long as 3-RS. (Fig. 24) acute with a tubercle at apex; ventral Metasoma.—Carapace moderately long, CL/ opening of carapace short, distinctly in front of CW = 1.6-1.7, oval in dorsal view, somewhat apex. pointed in lateral view. Hypopygium short. 288 Journal of Hymenoptera Research

into medial and of Ovipostor short, straight. produced postero-lateral pairs Color. —Almost completely black except ex- stout teeth. Hind coxa strongly strigate or ar- Vein r of about as as treme apex— of hind coxa and of femur yellow. eolate-rugose. forewing long Male. Unknown. 3-RS. — Holotype Female. —TAIWAN: Tsuifeng, VI-3- Metasoma. Carapace (Fig. 21) very long, CL/ = 80, L. Y. Chou and C. C. Chen. Deposited in TARI. CW 2.3-2.5, areolate-rugose. Ventral opening of in middle of Paratypes. —TAIWAN: 2 females, same as ho- carapace very short, ending almost VII-16-82. S. C. Lin and concealed. lotype; 2 females, Tsuifeng, carapace. —Ovipositor C. N. Lin; 1 female, Tsuifeng, VI-21-79, K. S. Lin Color. Black; fore and middle legs yellow and B. H. Chen; 1 female, Meifeng, VII-31-IX-2-82, except coxae and femora black basally, hind leg L. Y. Chou and K. C. Chou; 1 female, Tayuling, IX- black except trochanters reddish-brown; papli 12-15-80, K. S. Lin and C. H. Wang. Deposited in pale yellow. TARI, USNM. Male. —Same as female except antenna not Host. —Unknown. dilated medially. Distribution. —Taiwan. Holotype Female.—ZHEJIANG PROVINCE: — distin- X. Wu. in Remarks. This species can be easily Xitianmushan, VII-29-84, J. Deposited guished from other species of the quadridentata- ZAU. 1984:371 the character- PROVINCE: 1 group (see Huddleston ) by Paratypes.—ZHEJIANG male, istics of the mesonotum and by the strongly pointed same as holotype; 1 female, Xitianmushan, VII-27- black X. 1 X. clypeus tooth and completely body. 84, J. Wu; female, Xitianmushan, VII-22-87, The name is from the M. Lou. in USNM, ZAU. Etymology.— specific Deposited— Latin gibbosus meaning hunched or humped in Host. Unknown.— reference to the strongly humped mesonotum. Distribution. China (Zhejiang Province). Remarks.—Both this species and macrogaster, Ascogaster grandis Tang and Marsh, new species, run into the quadridentata-group in new species the key of Huddleston (1984), but both of them are of this the Fig. 6, 21 distinguished from species group by — characteristics of head shape and their elongate Females. Length of forewing 4.0-4.1 mm, of carapaces. — body 4.8-5.0 mm. Etymologi/. The specific name is from the the Head. —Antenna with 37 flagellomeres; flagel- Latin grandis meaning large in reference to lum dilated medially, strongly tapered at apex, large expanded temple. medial segments about as long as broad. Temple slightly constricted behind eyes, at least 1.5 times Ascogaster hei Tang and Marsh, new species as long as eye in dorsal view. Occiput deep, con- Figs. 15, 16 cave. Ocelli on OO = 4.0 OD. Frons very small, line, — strongly depressed, rugose-punctate. Eyes protu- Females. Length of forewing 4.0-4.2 mm, of distinct setae. swol- 4.6-5.0 mm. berant, glabrous, without Gena body — len in lateral view. Face (Fig. 6) slightly protuber- Head. Antenna with 33-37 flagellomeres, ant, about 1.5 times as broad as high, areolate- moderately dilated medially, medial flagellomeres rugose, with medial carina dorsally. Clypeus slightly broader than long. Temple rounded be- of in slightly protuberant, punctate, distinctly divided hind eyes, distinctly longer than length eye Ocelli from face, apical border impressed laterally, pro- dorsal view. Occiput deeply concave. small, = duced medially into a very small pointed tooth. OO 3.5-4.0 OD, almost on line. Frons (Fig. 15) Mesosoma.—Pronotum strongly projecting in moderately depressed behind antennae, rugose- front of mesonotum, dorsolaterally rugose-punc- punctate, with a weak medial carina. Eyes protu- tate, smooth ventrally. Notauli not distinct, berant, nearly glabrous but with a few scattered mesonotum strongly areolate-rugose. Precoxal setae. Gena in face view contracted. Face (Fig. 16) suture indistinguishable, mesopleuron strongly about twice as broad as high, strongly areolate- areolate-rugose. Propodeum completely coarsely rugose. Clypeus distinctly divided from face, punc- areolate-rugose, divided by a transverse carina tate; apical border produced medially into a small Volume 3. 1 994 289 pointed tooth. Ascogaster lini Tang and Marsh, new species Mesosoma.—Pronotum projecting in front of Fig. 9 mesonotum, dorsolaterally areolate-rugose. — Notauli not distinct; mesonotum strongly areolate- Females. Length of forewing 3.9-4.3 mm, of 4.4-4.8 mm. rugose; precoxal suture indistinguishable; body — mesopleuron coarsely areolate-rugose. Head. Antenna with 33-36 flagellomeres, Propodeum strongly areolate-rugose, divided by weakly dilated medially, slightly tapered at apex, a transverse carina which is raised medially and medial segments about as broad as long, basal than laterally into stout dentate flanges. Hind coxa segments distinctly longer broad, apical seg- than broad. strongly strigate. Vein r of forewing about as long ments slightly longer Temple strongly as 3-RS. constricted behind eyes, slightly shorter than eye Metasoma.—Carapace longer, CL/CW =1.6- in dorsal view. Occiput deeply concave. Ocelli = 1.7, clavate in dorsal view. Ventral opening of small, on line, OO 4.0 OD. Frons with two carapace longer, not very distinctly in front of depressed, polished impunctate areas behind an- apex. Hypopygium short. Ovipositor sheaths clav- tenna, median carina distinct. Eyes moderately ate. protuberant, glabrous. Face (Fig. 9) slightly protu- Color. —Almost completely black except fore berant, about 1.5 times as broad as high, finely tibia, and sometimes middle and hind coxae testa- punctate, with a median carina in upper part; ceous apically. Clypeus weakly divided from face, protuberant, Male. —Same as females except antennae not sparsely punctate, smoother than face; apical bor- dilated medially, with 34-35 flagellomeres. der of clypeus strongly impressed laterally, me- Holotype Female.—ZHEJIANG PROVINCE: dial area raised and produced forwards into a Fengyangshan, VIII-12-84, L. R. Sheng. Deposited distinct pointed tooth. inZAU. Mesosoma.—Pronotum projecting in front of

Paratypes.—JILING PROVINCE : 2 females, 3 mesonotum, laterally shining and smooth, dor- males, Dongjiang, VII-20-31-88, X. M. Luo. sally rugose-foveolate. Notauli strong, foveolate, HEILONHJIANG PROVINCE: 1 female, coalescing posteriorly in an areolate-rugose area; Qingdinzhi, V-77, Y. Y. Hun. ZHEJIANG PROV- rest of mesonotum rugose-punctate. Precoxal su- INCE: 1 male, same as holotype; 1 female, ture shallow, foveolate; mesopleuron above Xitainmushan, VIII-3-84, L. K. Sheng; 1 female, precoxal suture rugose-foveolate anteriorly, 1 VI- rest of Songyang, VII-15-17-89, J. H. He; male, Zhuji, smooth and punctate posteriorly, 10-85, X. X. Chen. FUJAIN PROVINCE: 1 female, mesopleuron ventrally densely punctate. 1 male, Guadun (Wuyishan), VI-11-82, J H. Xiu; 1 Propodeum completely areolate-rugose divided female, Wuyishan, VI-20-80, N. Q. Lin; 1 female, medially by a transverse carina with medial and 1 lateral the broad and Wuyishan, VI-30-80, J. H. Xiu; female, Wuyishan, tubercles, medial pair stout, Y. 1 S. lateral VIII-3-83, Ma; male, Wuyishan, VII-30-81, J. tubercles more prominent. Hind coxa finely Weng. Deposited in FAC, USNM, ZAU. punctate. Vein r of forewing about 1.5-2.0 times as Host. —Unknown. long as 3-RS. Distribution. —China (Heilongjiang Province, Metasoma.—Carapace long, CL/CW =1.8-2.0, Tiling Province, Zhejiang Province, Fujian Prov- clavate in dorsal view, finely areolate-rugose, with ince). — two weak medial carinae at base. Ventral opening Remarks. This species can be distinguished of carapace longer, almost at apex of carapace. from other species of the quadridentata-group by Hypopygium short. Ovipositor sheaths clavate, the more massive head, strongly strigate-rugose short. hind coxa, and larger and stout body. Color. —Black except carapace yellow at base; —This is for Prof. Etymology. species named J. antenna infuscate except scape yellow; legs yel- H. He, Zheijiang Agricultural University, in ap- low except hind coxa at base, mid and hind femur preciation for his helpfulness in this work. 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, Etymology. It is a pleasure to name this spe- malaise trap, VI-84, K. S. Lin and K. C. Chou. cies in honor of its collector, Mr. K. S. Lin, who also Deposited in TARI. collected a large amount of Braconidae in Taiwan. 1 3 Paratypes. TAIWAN: female, males , Tsuifeng," VI-21-79, K. S. Lin and B. H. Chen; 3 Ascogaster macrogaster Tang and Marsh, males, Tsuifeng, VI-3-80, L. Y. Chou and C. C. new species 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. Females. Length of forewing 3.5-4.2 mm, of 12 4 K. S. Lin 4.3-4.9 mm. Tang; females, males, VI-25-27-81, body — and W. S. Tang; 2 females, 1 male, Tsuifeng, V-23- Head. Antenna with 33 flagellomeres, slightly 82, L. Y. Chou; 9 females, 1 male, VII-16-82, S. C. dilated medially, medial segments about as long Lin and C. N. Lin; 4 females, 5 males, Tsuifeng, VI- as broad. Temple slightly contracted behind eyes, 23-25-83, K. S. Lin and S.C. Lin; 13 females, 14 slightly longer than eyes in dorsal view. Occiput male, same as holotype; 20 females, 8 males, concave. Ocelli larger, OO =3.0 OD, on line. Frons Tsuifeng, VII-XI-84, Malaise trap, K. S. Lin and K. slightly excavated, rugose-punctate; medial ca- C. Chou; 1 male, Tsuifeng, IX-85, Malaise trap, K. rina distinct. Eyes not protuberant, glabrous with- S. Lin; 1 female, 1 male, Meifeng, V-10-79, K. C. out distinct setae. Gena expanded in lateral view. Chou; 2 females, Meifeng, VI-20-22-79, K. S. Lin Face slightly protuberant, about 1 .5 times as broad and B. H. Chen; 2 females, 1 male, Meifeng, VI-2- as high, areolate-rugose. Clypeus punctate; apical 4-80, L. Y. Chou and C. C. Chen; 1 female, Meifeng, border impressed laterally, produced medially VI-5-8-80, C. C. Chen; 1 female, Meifeng, VI-8-80, into a small pointed— tooth. K.S. Lin and B. H. Chen; 1 female, Meifeng, X-5-9- Mesosma. Pronotum strongly projecting in 80, C. C. Chen and C. C. Chien; 1 female, 10 males, front of mesonotum, dorsolaterally areolate-rug- Meifeng, V-7-9-81, K. S. Lin and S. C. Lin; 2 fe- ose. Notauli not distinct; mesonotum rugose-punc- males, Meifeng, VI-24-26-81, K. S. Lin and W. S. tate anteriorly, rest of mesonotum foveate. 2 L. Y. 1 Tang ; males, Meifeng, V-22-82, Chou; Mesopleuron deep punctate, but precoxal suture male, Meifeng, V-3-83, H. Townes; 5 males, not distinguishable. Propodeum strongly areolate- Meifeng, V-10-83, H. Townes; 2 males, Meifeng, V- rugose, not distinctly divided into dorsal and pos- 15-83, H. Townes; 1 female, 4 males, V-22-83, H. terolateral surfaces. Hind coxa smooth, shining. Townes; 4 females, Meifeng,VI-22-26-83, K. S. Lin Vein r of forewing about 1.5 times as long as 3-RS. = and S. C. Lin; 1 male, Meifeng, V-8-11-84, K. C. Metasoma.—Carapace very long, CL/CW Chou and C. C. Pan; 1 female, 3 males, Meifeng, 2.5, areolate-rugose, rounded apically in lateral VI-84, Malaise trap, K. S. Lin and K. C. Chou; 1 view. Ventral opening of carapace at 1/4 apex of female, Meifeng, VIII-84, Malaise trap, K. S. Lin carapace. Hypopygium broad. Ovipositor short, and K. C. 1 clavate. Chou; female, Meifeng, X-84, Malaise pointed apically;— ovipostor sheath trap, K. S. Lin and K. C. Chou; 1 female, Meifeng, Color. Black; carapace sometimes yellow at X-85, Malaise trap, K. S. Lin; 1 female, Sungkang base; all legs yellow except hind femur and tibia 2100 m, X-84, Malaise trap, K. S. Lin and K. C. brownish apically; palpi pale yellow. Chou; 1 female, 2 males, Sungkang, X-XI-85, Mal- Male. —Same as female except antenna not aise trap, K. S. Lin; 8 females, 3 males, Tayuling dilated medially, all flagellomeres longer than 2560 m, VI-9-16-80, Malaise trap, K. S. Lin and B. broad, and smaller yellow spot at basal carapace. H. Chen; 1 female, 2 males, Alishan 2400 m, VI-12- Holotype Female.— FUJIAN PROVINCE: T. K. S. 4 16-65, Maa and Lin; females, Anmashan Tongmugun (Wuyishan), VI-23-82, J. C. Huang. 2275m, VII-6-9-79, L. Y. Chou. Deposited in AEIG, Deposited in FAC. FAC, TARI, USNM. Paratypes.—FUJIAN PROVINCE: 1 female, Host.—Unknown. same as holotype; 1 female, Wuyishan, VI-80, H. F. Distribution. —Taiwan. 1 VI-80. C. 1 Chao; female, Wuyishan, J. Huang; Remarks.—This species is very similar to male, Huanxi, V-30-86, N. Q. Lin. Deposited in perkinsi in the characteristics of the face and clypeus, FAC. but it is rather different in the shape of the carapace Host. —Unknown. and color patterns of both legs and antennae. Distribution. —China (Fujian Province). Volume 3, 1 994 291

Thomson 1892: 1719. female, Remarks.—This species is very close to grandis Ascogaster nigricornis Lectotype SWEDEN (designated and syn. by Huddleston 1984). from which it can be distinguished not only by the Shenefelt 1973: 826. characters mentioned in the key but also by the Ascogaster cynipum Thomson 1892: 1720. Holotype male, lack of distinct tubercles on the propodeum. SWEDEN Huddleston 1984). Shenefelt,1973: — (syn. by Etymology. The specific name is from the 820. 1895: 83. USSR Greek makros meaning long and gaster meaning Ascogaster egregius Kokujev Holotype male, (syn. Huddleston 1984). Fahnnger 1934: 525; Shenefelt stomach in reference to the long metasomal cara- by 1973: 821. pace. Chelonus nigrator Szepligeti 1896: 303. Holotype female, YU- GOSLAVIA: Buccari (syn. by Huddleston 1984). Ascogaster perkinsi Huddleston Shenefelt 1973: 859. Chelonus Viereck 1909: 43. female, USA: Figs. 10, 22 carpocapsae Holotype Michigan (syn. by Rosenberg 1934). Ascogaster epinotiae Watanabe, 1937: 76. Holotype female, Ascogaster perkmsi Huddleston 1984: 368. Holotype female, JAPAN: Hokkaido, Sapporo (syn. by Huddleston 1984). JAPAN: Mt.Tachibana, Fukuoka. Shenefelt 1973: 821. Shenefelt 1973: Shaw 1983: — Ascogaster quadridentata; 828; 32; Diagnosis. Length of forewing 2.6-2.9 mmm, Huddleston 1984: 376 Walker and Huddleston 1987: of body 3.0-3.5 mm. Antenna with 34-36 343; Papp 1989: 258. flagellomeres, slightly dilated medially (female); — 27-32 ocellar triangle obtuse, ocelli almost on line; face Diag)iosis. Antenna with flagellomeres, dialted ocelli on line or almost (Fig. 10) about 1.5 times as broad as high, rugose- slightly medially; face twice as broad punctate; apical border of clypeus strongly im- on line (Fig. 17); (Fig. 18) about pressed laterally, medial area projecting and pro- as high, genearlly finely rugose; clypeus finely duced forwards into a distinct pointed tooth; punctate, apical border produced medially into a notauli weak, foveolate, rest mesonotum rugose- pointed tooth; notauli and precoxal suture indis- punctate; precoxal suture foveolate, mesopleuron tinguishable, mesonotum and mesopleuron com- above precoxal suture areolate-foveolate anteri- pletely areolate-rugose; hind coxa always mostly orly, sparsely punctate posteriorly, rest of black and strongly strigate-rugose; carapace oval, mesopleuron densely punctate; carapace short, short, genearlly deep in lateral view, sometimes oval in dorsal view, clavate in lateral view, gener- yellow at base; ventral opening of carapace short, in front of ally yellow at base; ventral opening of carapace usually distinctly apex. — fe- (Fig. 22) short, distinctly in front of apex; hind coxa Specimens Examined. BEIJING CITY: 1, finely punctate; all legs yellow except middle tibia male, 1 male. JILFNG PROVINCE: 1 male. HUNAN dark at base, hind femur and tibia dark at base, PROVINCE: 2 females. JIANGXI PROVINCE: 1 mid and hind tarsi infuscate. male. GUANGXI PROVINCE: 1 female, 1 male. Specimens Examined.—ZHEJIANG PROV- YUNNAN PROVINCE: 2 females, 2 males. INCE: 41 females, 24 males. HUNAN PROVINCE: GUIZHOU PROVINCE: 3 females, 2 males. 1 male. FUJIAN PROVINCE: 4 females, 3 males. ZHEJIANG PROVINCE: 29 females, 30 males. TAIWAN: 11 females, 6 males. FUJIAN PROVINCE: 18 females, 11, males. TAI- Host. —Unknown. WAN: 26 females, 7 males. Distribution. —China (Zhejiang Province, Additional Specimens Examined.—Lectotype of Hunan Province, Fujian Province), Taiwan, Japan. quadridentatus Wesmael, female, BELGIUM: Brus- Remarks.—All specimens examined agree well sels (IRSNB). Lectoparatypes of quadridentatus with the description of Huddleston (1984). This is Wesmael, 1 female, 5 males, same as lectotype a new species record for the fauna of China. (IRSNB). USA: 1 female, 1 male, WA, Pullman, August 30, 1989, ex. Cydia pomonella, det. by Shaw, Ascogaster quadridentata Wesmael 1990 (FAC). Figs. 17, 18 Hosts.—Huddleston (1984) and Shaw (1983) recorded the following hosts in the Palaearctic and Wesmael 1835: 237. fe- Ascogaster quadridentatus Lectotype Nearctic Regions: Cydiafunebrana (Treitsche), Cydia male, BELGIUM: Brussels (designated by Shaw 1983). nigricana Steph., Cydia pallifrontana (Lienig & Chelonus impressus Herrich-Schaffer 1838: 153. Syntypes, Zeller), Cydia pomonella (L.), Epiblema uddmanniana GERMANY ( syn. by Reinhard 1867) 292 Journal of Hymenoptera Research

(L.), Grapholitlui molesta (Busck), Grapholitapnmivora YUNNAN PROVINCE: 3 females. GUANHXI (Walsh), Spilonota ocellana (Denis & Schiffermuller) PROVINCE: 1 female. FUJIAN PROVTNCE: 1 fe- viteana 1 male. TAIWAN: 31 36 males. (Lepidoptera: Tortricidae); Endopiza male, females,— (Clem.); Yponomeuta padella (L.) (Lepidoptera: Additional Specimens Examined. Holotype of Yponomeutidae). reticulatus Watanabe, male, JAPAN: Hokkaido, Distribution. —China (Beijing City, Jiling Prov- Asahigawa, June 15, 1966 (UEI). Paratypes of ince, Jiangsu Province, Zhejiang Province, Yunnan reticulatus Watanabe: 3 females, JAPAN, May 16- Province, Guizhou Province, Guangxi Province, July 20, 1966 (UEI); 1 female, 1 male, JAPAN, July Fujian Province), Taiwan, Japan, Korea, New 20-23, 1966 (USNM). Zealand, West Palaearctic Region, Nearctic Re- Hosts. —Acleris fimbriana (Thunberg), orana von gion. — Adoxophyes (Fischer Roslerstamm), Remarks. This species has been studied Archips issikii Kodama, Archips oporana (L.), Archips widely as a parasitoid of several economically pulchra (Butler), Carposiatw nipponensis Walsingham important insect pests in Europe and North (Lepidoptera: Tortricidae). America. Its hosts and biology were reviewed by Distribution. —China (Beijing City, Henan Huddleston (1984) and Shaw (1983). There is very Province, Shandong Province, Shanxi Province, little published work on this species in China. Shannxi Province, Anhui Province, Jiangshu Prov- Fahringer (1934) recorded it in Jiangsu Province ince, Jiangxi Province, Zhejiang Province, Yunnan (=Kiangfu). This species is characterized by its Province, Guangxi Province, Fujian Province), coarse sculpture, short oval and deep carapace, Taiwan, Japan,— Korea, Czechoslovakia. short temple and antenna, highly strigate and Remarks. This species is easily distinguished black hind coxa, and the single tooth on its cly peus. from other species of the quadridentata-gToup by the matt and no tooth clypeus and by the yellow Ascogaster reticulata Watanabe and smooth hind coxa and the hind tibia with a yellow-pale band medialy.

Ascogaster reticulatus Watanabe 1967: 41 . Holotype male (cited as Hokkaido, Chen and female), JAPAN: Ashigawa. He, Ascogaster rugulosa Tang and Marsh, new Ma 1989: 438. species Ascogaster reticulata: Huddleston 1984: 377; Papp 1989: 298. Fig. 13 Diagnosis. —Length of forewing 2.8-3.2 mm, of 3.4-3.8 m. Antenna with 34-36 body flagellomeres, Females. —Length of forewing 3.0-3.2 mm, of dilated ocelli on face weakly medially; large, line; body 3.4-3.7 mm. about 1.5 times as broad as high, coarsely irregu- Head.—Antenna with 31-35 flagellomeres, matt and larly rugose; clypeus strongly punctate, slightly dilated medially, flagellomeres 1-11 longer its border almost flat without trace of apical any than broad, rest of flagellum about as broad as tooth or tubercle; mesonotum areolate-rugose, long. Temple rounded behind eyes, about equal to notauli suture indistinct; precoxal indistinguish- length of eye in dorsal view. Occiput concave. cara- able, mesopleuron strongly areolate-rugose; Ocelli small, OO =3.5-4.0 OD, on line. Eyes protu- sometimes at ven- pace short, deep, yellow base; berant, glabrous. Frons moderately concave, tral of short, in front of opening carapace distinctly smooth, with a weak medial carina. Face (Fig. 13) hind coxa smooth or sometimes apex; yellow, about 1.5 times as broad as high, finely areolate- hind tibia black with a finely punctate; pale yellow rugose, generally more finely scluptured ventrally medial band. than dorsally. Clypeus punctate, apical border Examined.— CITY: 2 fe- Specimens BEIJING produced medially into a pointed tooth. males. HENAN PROVINCE: 13 females. Mesosoma.—Pronotum little projecting in front SHANDONG PROVINCE: 8 females, 5 males. of mesonotum, rugose-punctate dorsolaterally. SHANXI PROVINCE: 8 females. SHAANXI Notauli present but not very distinct; mesonotum PROVINCE: 3 females, 1 male. ANHUI PROV- rugose-punctate except an areolate-rugose area INCE: 1 male. PROVINCE: 9 JIANGSHU females, posteromedially. Precoxal suture very weak, shal- 2 males. PROVINCE: 2 females. JIANGXI low foveolate; rest of mesopleuron sparsely but PROVINCE: 7 females, 4 males. ZHEJIANG coarsely punctate. Propodeum coarsely areolate- Volume 3. 1 994 293

Malaise K. S. 4 rugose, divided by a tranverse carina which is Tsuifeng, IX-85, trap, Lin; males, K. S. Lin S. C. 3 raised into a pair of medial dentate flanges and a Tsuiofeng, IX-12-14-84, and Lin; H. pair of lateral dentate flanges. Hind coxa finely males, Wushe, IV-26-83, Townes;l male, Wusha, K. C. 1 VIII-6- punctate. Vein r of forewing about as long as or VII-25-78, Chou; female, Sungkang, K. S. 1 1 slightly longer than 3-RS. 84, Lin; female, male, Sungkang, X-84, Metasoma.—Carapace short, oval in dorsal Malaise trap, K. S. Lin and K. C. Chou; 2 males, view, clavate in lateral view, areolate-rugose. Ven- Sungkang, Malaise trap, IX-X-85, K. S. Lin; 1 fe- tral opening of carapace less distinctly in front of male, 2 males, Shengkuang, IX-20-68, K. S. Lin; 1 apex. Ovipositor short, tapered apically. female, Wuling, VI-27-29-79, K. S. Lin and L. Y. Color.—Black; fore and middle legs yellow Chou; 1 female, Tayuling, VI-9-16-80, K. S. Lin and except middle tibia at apex and all tarsi infuscate, B. H. Chen; 1 male, Lishan, IX-12-68, K. C. Chou. hind leg black except apex of coxa and base of Deposited in AEIG, FAC, TARI, USNM, ZAU, femur yellow and with a pale yellow medial band ZRI. at Host.—Unknown. on tibia; carapace always yellow laterally base, — but black medially.— Distribution. China (Zhejiang Province, Males. Same as females except antenna not Hainan Province),— Taiwan. dilated medially. Remarks. This species is close to reticulata Holotype Female.—TAIWAN: Meifeng, V-7-9- from which it can be distinguished not only by the 81, K. S. Lin and S. C. Lin. Deposited in TARI. characters mentioned in the key but also by its Paratypes.—ZHEJIANG PROVINCE: 1 male, smooth mesopleuron, less deep carapace and Xitainmushan, VI-2-4-90, L. G. Weng. HAINAN mostly black hind coxa. PROVINCE: 1 male, Shuiman, V-26-60, X. F. Li. Etymologic—The specific name is in reference TAIWAN: 7 females, 1 male, same as holotype; 1 to the rugulose sculpturing on the clypeus. female, Meifeng, VII-26-78, K. C. Chou; 1 female, Meifeng, IV-8-9-78, K. S. Lin; 1 male, Meifeng, VII- Ascogaster semenovi Telenga 18-79, K. C. Chou; 1 female, Meifeng, V-15-22-79; Figs. 11, 25 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 Ascogaster semenovitelenga 1941: 310, 453. Holotype female, MONGOLIA: Alashan, Shenefelt 1973: 825; C. C. Chien; 1 male, Meifeng, VI-24-26-81, K. S. Lin Dyn-juan-in. Huddleston 1984: 352. and W. S. Tang; 3 males, Meifeng, VIII-28-29-81, L. Ascogaster kyushuensis Yoneda 1978: 291. Holotype female, Y. Chou and S. C. Lin; 1 female, 1 male, VII-31-IX- JAPAN: Kyushu, Fukuoka Pref., Fukuoka City, Hakozaki L. Y. Chou and S. C. 2 2 1984 2-82, Lin; females, males, ( syn. by Huddleston ). Meifeng, V-8-11-82, K. C. Chou and C. C. Pan; 1 — male, Meifeng, VII-30-83, L. Y. Chou; 1 female, Diagnosis. Length of forewing 4.3-4.7 mm, of Meifeng, VI-22-26-83, K. S. Lin and S. C. Lin; 1 body 5.2-5.6 mm. Antenna of female with 20 female, Meifeng, X-4-7-82, K. C Chou; 1 male, V- flagellomeres and medial segments dilated, 8-11-84, K. C. Chou and C. C Pan; 1 female, genearlly broader than long, male with 25-26 Meifeng, VII-84, Malaise trap, K. S. Lin and K. C. flagellomeres, medially not dilated, all segments Chou; 1 female, Tungpu, X-18-21-82, K. C. Chou longer than broad; ocellar triangle acute, ocelli not and S. C. Lin; 1 female, Tungpu, VI-20-24-83, K. C. on line; face (Fig. 11) about twice as broad as high, Chou and C. Y. Wong; 1 female, Tungpu, VII-85, rugose-punctate, the hairs on the upper part of Malaise trap, K. S. Lin; 1 female, Tsuifeng, VI-21- face pointing upwards; clypeus with its apical 79, K. S. Lin and B. H. Chen; 1 female, Tsuifeng, VI- border rounded medially, produced laterally into 3-80, L. Y. Chou and C. C. Chen; 1 female, Tsuifeng, broad dentate flanges, without tooth or incision; V-8-81, K. S. Lin and S. C. Lin; 4 females, 1 male, notauli distinct, foveolate, rest of mesonotum punc- Tsuifeng, VI-25-27-81, K. S. Lin and W. S. Tang; 1 tate with an areolate-rugose area posteromedially; male, Tsuifeng, VIII-27-81, L. Y. Chou and S. C. precoxal suture deep foveolate, rest of mesopleuron Lin; 1 male, Tsuifeng, IX-1-3-82, L. Y. Chou and K. punctate with a deep foveolate groove C. Chou; 3 females, Tsuifeng, VI-23-25-83, K. S. Lin anterodorsally; hind coxa finely and sparsely punc- and S. C. 2 VI- = Lin; females, Tsuifeng, VII-84, Mal- tate; carapace (Fig. 25) long, CL/CW 2.2-2.4, aise trap, K. S. Lin and K. C. Chou; 3 males, pointed in dorsal and lateral views; ovipositor 294 Journal of Hymenoptera Research sheaths broad, knife-like. Almost completely black Color.—Black; carapace always yellow at base; except sometimes fore tibia reddish-brown. all legs yellow except hind femur and tibia apically Specimens Examined.—JIANGSU PROVINCE: in fu scare. 2 females, 6 males. SHANGHAI CITY: 8 females, Males. —Same as female except antennae not 3 males. ZHEJIANG PROVINCE: 2 females, 8 dilated medially and carapace black at base. males. Holotype Female.—TAIWAN: Tungpu, XI-18- — S. C. Lin. in Hosts. Unkown. 21-82, K. C. Chou— and Deposited TARI. Distribution. —China (Jiangsu Province, Shang- Paratopes. TAIWAN: 4 females, 1 male, same 2 IX-25-29- hai City, Zhejiang Province), Mongolia, Japan. as holotype; 3 females, males, Tungpu, Remarks. This remarkable speices is easily dis- 80, L. Y. Chou and T. Lin; 1 female, Tungpu, IV-28- the Lin and C. 2 1 tinguished from other species of Ascogaster by V-2-81, T J. Lee; females, male, lateral dentate flanges on its clypeus and by the Tungpu, X-5-8-81, T. Lin and W. S. Tang; 1 female, broad and knife-like ovipositor sheaths. It is new Tungpu, XI-18-23-81, T. Lin and W. S. Tang; 6 to the fauna of China. females, Tungpu, X-XI-85, Malaise trap, K. S. Lin; 4 males, Tungpu, IV-16-VII-27-84, K. C. Chou and Ascogaster setula Tang and Marsh, new species C. H. Yung; 2 females, 2 males, Lienhuachi, III-VII- K. S. K. Fig. 8 84, Malaise trap, Lin and C. Chou. Depos- ited in USNM. — FAC,— TARI, Females. Length of forewing 2.4-2.7 mm, of Hosts. Unknown.— body 2.9-3.1 mm. Distribution.— Taiwan. Head. —Antenna with 26-29 flagellomeres; fla- Remarks. Previous authors (Baker 1926, Shaw gellum moderately dilated medially, medial seg- 1983, Zettel 1990) treated glabrous eyes as a char- ments slightly broader than long, apical segments acter to distinguish Ascogaster. Strikingly this spe- about as broad as long. Temple rounded behind cies has distinct and dense setae on its eyes, but eyes, about equal to or slightly shorter than eye in except for this character it agrees well with typical = dorsal view. Occiput concave. Ocelli small, OO Ascogaster. Its very small body and the character- 4.0 OD, almost on line. Frons moderately depressed istics of the clypeus are also useful to distinguish this from other Chinese behind antenna. Eyes slight protuberant, with species — Ascogaster. dense and distinct hairs. Malar space shorter, about Etymology. The specific name is the diminu- 0.5 time as eye high. Gena strongly constricted in tive form of the Latin seta meaning bristle in to the face view. Face (Fig. 8) protuberant, coarsely rug- reference unusually hairy eyes. ose-punctate, about 1.5 times as broad as high. Clypeus not distinctly divided from face, slightly Ascogaster townesi Tang and Marsh, convex, its apical border more or less straight, new species 26 with a very small— tooth medially. Figs. 12, Mesosoma. Pronotum protecting in front of — mesonotum, dorsolaterally foveate. Notauli shal- Females. Length of forewing 4.3-4.5 mm, of low foveolate, rest of mesonotum punctate except body 4.6-4.9 mm.

. — an areolate-rugose area posteromedially Precoxal Head. Antenna with 23 flagellomeres, me- suture not very distinct, mesopleuron generally dial segments not dilated, all segments longer coarsely rugose-punctate, sometimes areolate- than broad. Temple rounded behind eyes, slightly rugose. Propodeum coarsely areolate-rugose, di- longer than eye in dorsal view. Occiput deeply vided by a transverse carina which is raised into a concave. Ocelli small, 00= 4.0-4.5 OD, ocellar medial pair and a lateral pair of stout dentate triangle acute, ocelli not on line. Frons behind flanges. Hind coxa smooth. Vein r of forewing 1.0- antenna slightly depressed, sparsely punctate. Eyes 1.3 times as long as 3-RS. moderately protuberant, glabrous with few scat- Metasoma.—Carapace shorter, CL/CW = 1.5- tered minute setae. Malar space short. Gena in face 1.8, areolate-rugose, oval, deeper in lateral view. view strongly constricted. Face (Fig. 12) slightly Ventral opening of carapace distinctly in front of convex, about twice as broad as high, less hairy, apex. Hypopygium short. Ovipositor short, its the hairs on the upper part of face pointing up- sheath clavate. wards, finely punctate, with a medial carina on the Volume 3, 1994 295

upper part. Clypeus slightly protuberant, more Ascogaster varipes Wesmael scatteredly punctate than face; apical border Fig. 14 straight, without medial teeth or flanges. Wesmael 1835: 234. BEL- Mesosoma.—Pronotum projecting little in front Ascogaster imnpes Lectotype female, GIUM: Brussels (designated by Huddleston 1984). of mesonotum, dorsolaterally smooth, shallow Shenefelt 1973: 837; Huddleston 1984: 370; Tobias 1986: fovea te. Notauli deep, foveolate, rest of mesonotum 305; Papp 1989:298. areolate- except with a depressed and Thomson 1874: 585. female punctate Ascogaster cavifrons Lectotype ,

. shal- rugose area posteromedially Precoxal suture SWEDEN: Skane, Torekov (desiganted and syn. by Huddleston Shenefelt 1973: 818. low foveate; rest of mesopleuron sparsely punc- 1984). sternalis Thomson 1874: 587. female, tate, smooth except anterodorsally foveolate. Ascogaster Lectotype SWEDEN: Smaland (designated by Huddleston 1984 Propodeum not distinctly divided into dorsal and and syn. by Telenga 1941). surfaces; with no posterlateral strongly rugose Ascogaster jaroslawensis Kokujev 1895: 86. Holotype female, tubercles. Hind coxa smooth. Vein r of USSR: 1984 forewing 'Jaroslaw' (syn. by Huddleston ). Shenefelt 1.0-1.2 times as long as 3-RS. 1973: 824; Tobias 1986: 304 (as a valid species). — Ascogaster variipes; Telenga 1941: 322. Metasoma. Carapace (Fig. 26) very long, CL/ CW = 2.3-2.5, clavate in dorsal view, flat in lateral — Diagnosis. Antenna with 32-36 flagellomeres, view, areolate-rugose, apically polished. Ventral slightly dilated medially (females); ocellar opening of carapace at apex of carapace. triangle obtuse, ocelli almost on line; face (Fig. 14) gener- Hypopyium large and broad. Ovipositor long, all hairs downwards; slender and upcurved. ally rugose-punctate, clypeus with its apical border transversely impressed, Color.—Black; legs yellow-brown except hind without medial tooth or excision; mandibles with coxa basally, hind and middle femora and tibia a deep semicircular depression at base; notauli apically and tarsi brownish; antenna infuscate distinct, foveolate, rest of mesonotum punctate except scape reddish brown; palpi yellowish brown. except posteriorly areolate-rugose; mesopleuron coarsely rugose so that suture not Males. —Same as females except antennae precoxal easily distinguished; propodeum completely areolate- longer with 32 flagellomeres, flater carapace and hind coxa yellow hind coxa. rugose; strongly transversely strigate; short, oval, in lateral view, ventral Holotype Female. —TAIWAN: Meifeng 2150 m, carapace deep of short, in front of V-10-83, H. Townes. Deposited in AEIG. opening carapace distinctly Paratypes.—TAIWAN: 1 male, Meifeng 2150 apex. Examined.—SHANDONG PROV- m, IV-19-21-83, K. C. Chou and P. Huang; 4 males, Specimens INCE: 2 females, Lao-shan, 800 m. Meifeng.2150 m, IV-26-83, H. Townes; 1 female, 30 Additional Examined.— of males, Meifeng 2150 m, V-3-83, H. Townes; 2 Specimens Lectotype varipes Wesmael, female, BELGIUM: Brussels females, 17 males, same as holotype. Deposited in and of AEIG, FAC, TARI, USNM. (IRSNB). Lectotype paralectotypes varipes 2 Host.—Unknown. Wesmael, BELGIUM: females, 3 males (IRSNB). Shenefelt Distribution.—Taiwan. Hosts.— (1973) listed many host — records. However, much of this information is of Remarks. This species is easily distinguished little value because of doubt about the of from other species of Ascogaster by the hairs on the accuracy identification of the parasite species involved (see upper part of the face, very flat carapace, no tu- Huddleston 1984). bercle on its clypeus, and long, upcurved oviposi- — tor. Distribution. China (Shandong Province), — Korea, "USSR" and other coun- Etymology. It is a pleasure to name this spe- many European tries Huddleston cies in honor the late Henry Townes who collected (see 1984). Remarks.—This species is characterized on many of the type series and who contributed greatly account of the semicircular at the base to our knowledge of the World Ichneumonoidea depression of the mandibles, the hind during his lifetime. strongly strigate 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 Etymology. The species name is in reference of a finely areolate-rugose face as in the type series. to the type locality.

Ascogaster yunnanica Tang and Marsh, new ACKNOWLEDGEMENTS species We thank the following people for loan of Male. —Length of forewing 2.7 mm, of body types and much of unidentified specimens or for 3.3 mm. providing valuable information: C. van Achterberg Head.—Antenna incomplete, 1st flagellomere (The Netherlands), A. Austin (Australia), H. F. about 3.0 times as long as broad, 2-6 th flagellomeres Chao (FAC), X. X. Chen (ZAU), L. Y. Chou (TARI), 2.0-2.5 times as as broad. constricted P. Dessart H. T. long Temple (IRSNB), J. He (ZAU), Huddleston behind eyes, slightly shorter than eye in dorsal (BMNH), K. Maeto (UEI), I. Nauman (Australia), view. concave. Ocelli small = D. Preston Occiput deeply OO J. Papp (HNHM), J. (BM), M. J. Sharkey 4.0 OD; ocellar triangle acute, ocelli not on line. (CNC), S. R. Shaw (Wyoming), G. N. Shida (BM), Frons behind antenna M. D. Y. moderately depressed, Suwa (UEI), Wahl (AEIG), J. Wang (ZRI). smooth. Eyes slightly protuberant, glabrous with- We also thank C. M. Liu (FAC) for help with out distinct setae. Malar space short. Gena in face translations of some Russian papers. The senior view strongly constricted. Face slightly convex, author offers special thanks to Prof. H. F. Chao for finely and sparsely punctate, about twice as broad his continued helpfulness and encouragement and as high, the hairs on the upper part of face pointing to Chunying for her patience and understanding. upwards. Clypeus slightly protuberant, more This work was partly supported by a grant to Y. Q. sparsely punctate than face, apical border straight Tang by the National Educational Commission of without flange—or tooth. the People's Republic of China. Mesosoma. Pronotum projecting little in front of mesonotum, deep foveate dorsolateraly . Notauli LITERATURE CITED foveolate, coalescing posteriorly in a fine areolate- rugose area; rest of mesonotum punctate. Precoxal Achterberg, C. van. 1976. A preliminary key to the subfami- lies of the Braconidae voor suture shallow foveate anterodorsally, indistinct (Hymenoptera). Tijdschrift Entomologie 119: 33-78. posteroventrally; rest of mesopleuron smooth Achterberg, C van. 1988. Revision of the subfamilv Blacmae posteroventraly, punctate anterodorsaly. Foerster (Hymenoptera: Braconidae). Zoologische not Propodeum distinctly divided into dorsal and Verhandelingen 249: 1-324. posterolateral surfaces, strongly rugose but with Achterberg, C. van. 1990a. Illustrated kev to the subfamilies of the Holarctic Braconidae no dentates. Hind coxa smooth. Vein r of forewing (Hvmenoptera: Ichneumonoidea). Zoologische Mededelmgen 64: 1-20 about twice as long as 3-RS. Achterberg, C. van. 1990b. Revision of the western Palaearctic Metasoma.— CL/CW = Carapace very long, Phanerotomini (Hymenoptera: Braconidae). Zoologiche in dorsal 2.5, oval view, deeper and not so strongly Verliandelingen 255: 1-106. flat in lateral view, areolate-rugose, but sparsely Baker, C. F. 1926. Braconidae-Cheloninae of the Philippines, Malaya, and Australia. The Science punctate apically. Ventral opening of carapace Philippine Journal of 31: 451-489. long, at apex of carapace. Bovce, H R. 1936. Laboratory breeding of Ascogaster Color. —Black; all hind coxa legs yellow except carpocapsae Viereck with notes on biology and larval hind tibia mostly black, femur and apically and morphology. Canadian Entomologist 68: 241-246. A. 1932. an tarsi infuscate; palpi yellow-brown. Cox, J. Ascogaster carpocapsae Viereck, important Female. —Unknown. parasite of the codling moth and the oriental fruit moth. Technical Bulletin of the New York State Agricultural Ex- Holotype Male.—YUNNAN PROVINCE: periment Station 188: 1-26. Kunming, 111-30-81, J. H. He. Deposited in ZAU. Fahringer, J. 1934. Opuscula Braconologica, Palaearktische Re- Host.— Unknown. gion 2: 321-594. Distribution. —China (Yunnan Province). Fullaway, D. T. 1919. New genera and species of Braconidae — mostly journal of Straits Branch, Royal Asiatic Remarks. Morphologically this species is simi- Malayan, Society 80: 39-61. lar to townesi from which it differs not only in the Harris, R. A. 1979. A glossary of surface sculpturing. Occa- characters mentioned in the but aslo in the key sional Papers in Entomology, California Department ofFood smaller and the less flat body carapace. and Agriculture 28: 1-31. Volume 3. 1 994 297

Y. 1989. Braconid one noctuid and 21 : He, J. H., X. X Chen and Ma. The (Hy- species. Applied Entomology Zoology 8-14. menoptera) parasites of Adoxophyes orana Fischer von 1989. Braconidae from Korea XI. Acta Rosterstamm (Lepidoptera: Tortncidae) from China Papp, ]. (Hymenoptera) 15 437-439. 295-326. Acta Agricultae Universitatis Zhejiangensis (4): Zoologica Hunganca 35(3-4): D. L. and C. van 1990. of the Huddleston, T. 1984. The Palaearctic species of Ascogaster Quicke, J. Achterberg. Phy logeny subfamilies of the Braconidae (Hymenoptera: Braconidae). Bulletin of the British Mu- family (Hymenoptera: 258: 1-95. seum (Natural Htston/) (Entomology) 49: 341-392. Ichneumonoidea). Zoologische Verhandehngen F. and distribution in France Kainoh, Y. 1986. Mating behavior of Ascogaster reticulatus Rosenberg, H. 1934. The biology L. Bulletin Watanbe (Hymenoptera: Braconidae, an egg-larval para- of the larval parasites of Cydia pomonella of 25: 201-256. sitoid of the smaller tea tortrix moth, Adoxophyes sp. Entomological Research, M. R. and T. Huddleston. 1991. Classification and (Lepidoptera: Tortncidae) I. Diel patterns of emergence Shaw, and mating, and some condintions for mating. Applied biology of braconid wasps (Hymenoptera: Braconidae). 7 Entomology and Zoology 21: 1-7. Handbooks for the Identification of British Insects (11): 1- Kainoh, Y. 1988. Some factors influencing sex ratio in 126. S.R. 1983. taxonomic of Nearctic Ascogaster reticulatus Watanabe (Hymenoptera: Shaw, A study Ascogaster Braconidae). Applied Entomology and Zoology 23:35-40. and a description of a new genus Leptodrepana (Hy- 2: 1-54. Kainoh, Y., T. Hiyori and Y. Tamaki. 1982. Kairomone of the menoptera: Braconidae). Entomography R. D. 1973. (Noi<. ed. egg-larval parasitoid, Ascogaster reticulatus Watanabe Shenefelt, Catalogus Hymenopterorum part (Hymenoptera: Braconidae). Applied Entomology and Zo- 10 Braconidae, 6:813-936. s'Gravenhage. 1932. Notes on some Braconidae and Ichneumomdae ology \7: 102-110. Sonan, J. of Kainoh, Y., T. Nemoto, K. Shimizu, S. Tatsuki, T. Kusano and from Formosa with descriptions 18 new species. Y. Kuwahara. 1991. Mating behavior of Ascogaster Transactions of the Natural History Society of Formosa 22: reticulatus Watanabe (Hymenoptera: Braconidae), an 66-86. G. V. 1905. Exotische Braconiden aus den egg-larval parasitoid of the smaller tea tortrix, Adoxophyes Szepligeti, II. Identification of a sex Dnentalischen und Austtralischen sp. (Lepidoptera: Tortricidae) Aethiopischen, Annates Musei Nationalis 3: 25-55. pheromone. Applied Entomology and Zoology 26: 543-549. Regionen. Hungarici Die Kainoh, Y. and Y. Tamaki. 1982. Searching behavior and Szepligeti, G. V. 1908. Braconidae und Ichneumomdae. Australiens 317-324. oviposition of the egg- larval parasitoid, Ascogaster Fauna Sudtoest- 1(9): reticulatus Watanabe (Hymenoptera: Braconidae). Ap- Telenga, N. A. 1941. Insects Hymenoptera, Family Braconidae, Braconinae and plied Entomology and Zoology 17: 194-206. Subfamily (continued) Sigalphinae Kainoh, Y. and S. Tatsuki. 1988. Host egg kairomones essen- Fauna SSSR 5 (3): 1-466.

I. of Cheloninae tial for egg-larval parasitoid, Ascogaster reticulatus Tobias, V. 1986a. New species Subfamily from the Far East of the Watanabe (Hymenoptera: Braconidea I, Internal and (Hymenoptera: Braconidae) external kairomones. journal of Chemical Ecology 14: 1475- USSR. Trudy Zoologicheskogo Instituta Akademiya Nauk 1484. SSSR 159: 3-17 (in Russian).

Kainoh, Y., S. Tatsuki. and T. Kusano. 1990. Host moth scales; Tobias, V. I. 1986b. Guide to the insect of European part of the a cue for host location for Ascogaster reticulatus Watanabe USSR Hymenoptera. Opredeliteli Faune SSSR, 3 (4): 1-500 (Hymenoptera: Braconidae). Applied Entomology and (in Russsian). of the Zoology 25: 17-25. Tobias, V. I. 1988. Two new species of Braconidae from the Kainoh, Y., S. Tatsuki, H. Sugie and Y. Tamaki. 1989. Host egg subfamlily Cheloninae (Hymenoptera) pro- Acta kairomones essential for egg-larval parasitoid, Ascogaster tected territoties of the Liruanian SSR. Entomology reticulatus Watanabe (Hymenoptera: Braconidae) II. Lituaman 9: 89-94. Identification of internal kairomone. Journal of Chemical Walker, A. and T. Huddleston. 1987. New Zealand chelonine Ecology 15: 1219-1229. braconid wasps (Hymenoptera: Braconidae). journal of Kamano, Y., K. Shimizu, Y. Kainoh and S. Tatsuki. 1989. Natural History 21: 339-361. Mating behavior of Ascogaster reticulatus Watanabe (Hy- Watanabe, C. 1937. A contribution to the knowledge of the menoptera: Braconidae), an egg-larval parasitoid of the Braconid fauna of the Empire of Japan, journal of the smaller tea tortix, Adoxophyes sp. (Lepidoptera: Faculty of Agriculture Hokkaido Imperial University 42(1): Tortricidae). Applied Entomology and Zoology 24: 372-378. 1-188." Kawakami, T. 1985. Development of the immature stages of Watanabe, C. 1967. Description of a new species of the genus

Ascogaster reticulatus Watanabe (Hymenoptera : Ascogaster Wesmael and notes on synonym of Apanleles Braconidae), an egg-larval parasitois of the smaller tea species (Hymenoptera: Braconidae). Insecta 41-44. tortrix moth, Adoxophyes sp. (Lepidopatera: Tortricidae). Matsumurana 29(2): Applied Entomology and Zoology 20: 380-386. Wesmael, C. 1835. Monographic des Braconides de Belgique. Kawakami, T. and Y. Kainoh. 1985. Host discrimination and Noui'eaux Memoires de VAcademic Royale des sciences et competition in the egg larval parasitoids, Ascogaster belles-lettres de Bruxelles 9: 1-252. reticulatus Watanabe (Hymenoptera: Braconidae). Ap- Yoneda, Y. 1978. A new species of the genus Ascogaster plied Entomology and Zoology 20: 362-364. Wesmael (Hymenoptera, Braconidae) from Japan. Kawakami, T. and Y. Kainoh. 1986. Host specificity of the Kontyu 46: 291-296. egg-larval parasitoid, Ascogaster reticulatus Watanabe Zettel, V.H.I 990. Eine Revision der Gattungen der Cheloninae (Hymenoptera: Braconidae) among five tortricid and (Hymenoptera, Braconidae) mit Beschreibungen neuer 298 Journal of Hymenoptera Research

Gattungen und Arten. Annates Naturlustorisches Mu- Most of the material examined by Chen et al. of bidentula seum Wien 91(B): 147-196. 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 NOTE ADDED IN PROOF the fact that atamiensis was recorded from China by Fahringer (1938, Ark. Zool. 30:3) and Shenefelt (1973). this needs to be studied The following paper came to our attention after Obviously species complex further our manuscript was sent to the printer: Chen Ascogaster consobrina Curtis and Ascogaster inf acetus Chen and Wu Zhishan. 1994. Jiahua, Huang Juchang and Huang. The new species described in Chen et al., six records of Notes on two new species and new mfacetus, will run to consobrina in our key. Fig. IB in Chen et al. which is an illustration of the is the genus Ascogaster Wesmael from China (Hy- mesopleuron not clear. Thus, we must reserve comment on this menoptera: Braconidae: Cheloninae). Journal ofthe spe- cies until the type can be observed Fujian Agricultural College (Natural Sciences Edi- Ascogaster longicornis Huddleston. We treated this species tion) 23(1)51-57. we were not able to Although as a synonym olformosensis based on type examination. et study the specimens used by Chen al., we have Ascogaster perkinsi Huddleston. See notes under wuyiensis below. made comments below on each species mentioned reticulatus Watanabe. this was by them based on a translated version of their Ascogaster Although species mentioned in their key, it was not discussed in the text. paper. Thus, we must see their specimens before deciding upon the relationship of this species and albitarsis. albitarsis Sonan. The characters mentioned in Ascogaster Ascogaster rufidens Wesmael. We did not find this species in their are not to this from key significant separate species any of the material that we studied. Again, their mate- reticulatus based on our discussion under study (see rial should be studied before we confirm this species is reticulatus). in China. arisanica Sonan and bidentula Ascogaster Ascogaster Ascogaster wuyiensis Chen and Huang. This species will Wesmael. our we examined the ma- During study type run to perkinsi in our key based on the description in terial of bidentula and atamiensis of arisanica, (synonym Chen etal, and the color of the clypeus and theshapeof All of the material we examined under the bidentula). the yellow mark at the base of the carapace may be only name arasanica with the but none agreed type agreed variation in perkinsi. We described a new species, lini, with of bidentula or atamiensis. Furthermore, we types which is related to wuyiensis and perkinsi. The validity of found that the characters used Chen et al. and by these three species must wait until specimens of wuyiensis Huddelston (1984) are not realiable in distinguishing are examined. bidentula and arisanica (see our discussion of arasanica).

Figs. 1-2. Fore wings, somewhat diagrammatic. 1, Ascogaster. 2, Chelonus. SM=submarginal cells; D=discal cell. 299 Volume 3. 1994

n. 5 Sonan. n. Sonan. , 6, Figs. 3-8. Faces of Ascogaster species. 3, arisanica 4, dimorpha sp. formosensis grandis 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 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 lems. We suspect that only about one-third of the species of Pseudomethoca in the Neotropics have Mutillids are solitary parasitoid wasps that been described. Previous taxonomic work on Mickel exhibit great sexual dimorphism, making sex as- Pseudomethoca was done by (1924, 1935, sociations difficult. The New World genus revision of North American species; 1952, key Pseudomethoca Ashmead illustrates this problem; to females of Guyanan species), Schuster (1945, only one-fifth of its 103 described species are known key to Caribbean species), and Krombein (1992). from both sexes. Distinctly fewer sex associations Gynandromorphy is a developmental phe- have been obtained for Neotropical than for Ne- nomenon useful for associating the sexes in some arctic species of Pseudomethoca. Seventeen out of extremely dimorphic animals, including mutillids 45 Nearctic species known (37.8%) have both sexes (Mickel 1928, 1936, 1952; Bischoff 1931). Unfortu- recognized (Krombein, 1992). Of the remaining 28 nately, gynandromorphs are rare in Mutillidae. species, 20 are known only from females, and After examining more than 15,000 mutillid speci- eight only from males. In contrast, only four of mens, we have discovered only two gynandro- some 58 Neotropical species of Pseudomethoca are morphs. A review of the literature revealed only known from both sexes (6.9%) (Nonveiller, 1990; six previously published cases (Table 1). We re- Cambra & Quintero, 1992). Of the remaining 54, 46 port here the second known Neotropical mutillid are known from females only, and eight from gynandromorph. We recently discovered a gy- males only. Success in associating the sexes will nandromorph of Timulla labdace Mickel, from facilitate future biological work on the group and Panama, that will be described in a separate pub- will solve some of the annoying taxonomic prob- lication. 304 Journal of Hymenoptera Research

evidence of a carina. Propleura and mesopleura Pseudomethoca areta (Cameron) with close smooth, with- - punctures; metapleura Figs. 1 4 out punctures; sides of propodeum with only a few scattered Pronotum, scutellum, 1895: punctures. Sphaerophthalma areta Cameron, 332, pi. 14, fig. 12, metanotum and dorsal face of clothed female. Bugaba, Chinqui Province, Panama, Champion propodeum, erect col., BM(NH), London, Type 15.822, examined; with sparse, long white pubescence; Pseudomethoca areta: Cambra & Quintero, 1992: 474. mesonotum clothed with sparse black pubescence; female. panamensis Cameron, 1895: 334-35, white Sphaerophthalma tegula clothed with intermixed, black and Bugaba, Chinqui Province, Panama, Champion col., pubescence; metapleura and sides of propodeum BM(NH), London, Type 1 5.833, examined; Pseudomethoca almost bare, with white panamensis: Cambra & Quintero, 1992: 475. NEW SYN- only sparse ONYMY. micropubescence. Anterior and intermediate coxae without teeth

— revised it a keel on inner mar- Diagnosis. In Mickel's key (1935) or keels; posterior coxa with five. The white runs to P. vanduzei Bradley in couplet gin. Legs clothed with sparse pubescence. as male of P. areta differs from P. vanduzei follows: Calcaria pale. so as to one sessile posterior half of tegula bent downward Abdomen with segment completely the form a posterior face at a sharp angle with with second. Terga one and two with small, sepa- dorsal surface (in P. vanduzei the tegula is uni- rated punctures, except the apical margins with formly convex, without a posterior face); anterior close punctures; terga three to six with small, close lacks x as margin of clypeus in P. areta is bidentate (it punctures. Pygidium rugose. Felt line 0.6 long teeth in P. vanduzei); mandibles tridentate at the tip as lateral margin of tergum two. Sternum one in P. areta (bidentate in P. vanduzei); integument of almost smooth, with only a few, sparse punctures, abdomen mostly ferruginous in P. areta (totally and with a low, median longitudinal carina on black in P. vanduzei). Pseudomethoca areta is en- anterior two-fifths. Sternum two with sparse, mod- demic to Panama, and P. vanduzei is present in the erate punctures. Sterna three to six with close, southeastern United States (Krombein, 1979). moderate punctures. Posterior margin of — clothed Description. Integument black, except apex hypopygium evenly convex. Tergum one of tergum one and abdominal segments two to six, with sparse, long erect, white pubescence. Terga orange. Head large, subrectangular in dorsal view, two to six with sparse long erect, orange pubes- as wide as thorax, clothed with long, erect and cence, the apical margins with a band of dense, recumbent white pubescence; row of six to eight recumbent, orange pubescence. Last tergum long, erect, dark hairs near inner margin of eyes; clothed with black pubescence. Sterna one to six posterolateral angles of head not dentate. Man- clothed with white pubescence, except lateral dibles tridentate distally; clypeus strongly margins of sterna two to four, with orange pubes- bidentate medially on the cephalic margin; disk of cence. Last sternum clothed with white pubes- clypeus densely punctate. Scape with a strong cence and a few intermixed fuscous hairs. Wings longitudinal carina beneath; first flagellomere infuscated, especially apically; forewing with two equal in length to second. Front, vertex and genae well defined submarginal cells and traces of a coarsely and confluently punctate. Antennal third. Body length; 10.6 mm. scrobes and genae not carinate. Ocelli small, dis- ALLOTYPE Male Information.—PANAMA: tance between eye margin and lateral ocelli equal Darien Province, Cruce de Mono, Estacion diameter to approximately five times the greatest INRENARE, 8 Feb 1993 (yellow trap). R. Cambra latter. B. of the T., J. Coronado, Museo de Invertebrados "G. Pronotum, mesonotum and scutellum with Fairchild", Universidad de Panama (MIUP). close, more or less confluent punctures, punctures Additional Material Examined. —PANAMA: about the size of those on head. Propodeum Darien Province, Cruce de Mono, Estacion reticulate and 144 strongly and coarsely dorsally pos- INRENARE, R. Cambra, J. Coronado, females - teriorly. Tegula punctate throughout; posterior and 27 males, 8 Feb 4 Mar 1993, deposited in part of tegula bent downward so as to form a MIUP, NMNH-Smithsonian Institution, Univer- face at a with the dorsal posterior sharp angle sity of Minnesota Insect Collection, and BM(NH). without surface. Humeral angles of pronotum any 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. Gynandromorph Individual of Comments— on Sex Association and Nezo Syn- Psendotnethoca areta onymy. Sex association is based on observations Fig. 2 of males courting and mating with females in the — field, and has been corroborated in the laboratory Description. Head identical to that of a nor- with mating experiments. Courtship is very brief, mal female, without a trace of male characters. lasting 15-40 seconds, and consists of bursts of Thorax and legs identical to those of a normal rapid vibrations of the wings and antennae, inter- male, without a recognizable trace of female char- spersed by short hopping flights. The male climbs acters. The anterior wings have abnormally thin, onto the female and grasps her neck with his translucent venation; they are torn along their mandibles and attempts to mate with her. We posterior half, and we suspect were never func- have not seen heterospecific courtship or mating tional. Abdomen with six segments, as in the in Psendomethoca. Erroneous heterospecific sex as- female. First tergum completely male. Second ter- sociations may be made if containers or outdoor gum a mosaic: right half with coloration and pu- sites become contaminated by pheromones re- bescence of male and female; left half is com- leased from a female of a different species that pletely male-like. Third tergum, right half with recently occupied that site (personal observations). female characteristics only; left half is a mosaic Males of Psendomethoca areta, like those of the with integument coloration and pubescence both genotype and males of numerous other species of of male and female. Abdominal segments four to Pseudomethoca we have examined, have a strong six are female only. Second sternum with right longitudinal carina beneath the scape. Therefore, half female and left half male, same as other sterna, we consider erroneous Casal's (1965) observation except for sternum three, identical to that of a that males of the ill-defined genus Darditilla Casal female. Body length: 11.1 mm. (genotype is only male known, and 35 other spe- Data and comments on the gynandromorph. — cies, known only from females) differ from those The gynandromorph individual was collected on of Psendomethoca in having a carina beneath the 26 February 1993, at 10:00 AM, in the general scape, said to be absent in the latter. The male locality of the allotype. When we first noticed it, genitalia of P. areta are symmetrical (Fig. 3), as are the individual was walking over dry leaves on the those of all other Neotropical sphaeropthalmine ground. Shortly afterward, we watched a male males we have examined, and phoretic mating is arrive, flying upwind, attracted by what we absent in this group. In contrast, the Neotropical thought was a normal female. The male quickly mutilline males of Timulla (Timnlla) present attempted to mate but encountered indifference strongly asymmetrical genitalia and phoretic mat- on the part of the female-like individual. Female ings (Cambra & Quintero, 1992). The asymmetry mutillids are known to attract winged males by in the male genitalia possibly functions to provide means of wind-dispersed pheromones (see Cambra a better hold, or grasp, of the female while air- & Quintero, 1993). The upwind arrival of the male borne. suggests that the gynandromorph individual was Cameron's types of areta and panamensis are secreting female pheromones. The abnormally thin both from Bugaba, and we found them to be and quite battered forewings, suggest that the the areta has it identical; name page precedence over was unable to fly, although had perhaps that of panamensis. Psendomethoca areta is closely attempted to. The specimen exhibits anterior/ related to P. hecate (Gerstaecker, 1874), from Costa posterior division of male and female compo- Rica, differing only in the integumental coloration nents, as well as mosaic segments, a type of gynan- of the vertex and dorsolateral areas of the thorax. dromorphy not previously described for mutillids We suspect that they are the same species. To (see Table 1). 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. Volume 3. 1994 307

ACKNOWLEDGMENTS Gerstaecker, C. E. 1874. Mutillarum Americae mendionalis indigenarun synopsis systematica et synonymica. Archil' fur Naturgeschiechte 40: 41-77. We thank the Smithsonian Tropical Research Institute Krombein, K. V. 1979. Family Mutillidae, pp. 1276-1314. In: (STRI), particularly Ira Rubinoff, for providing research fa- Krombein, K. V. et al., eds. Catalog of Hymenoptera m cilities. We are grateful to the British Embassy in Panama, in America North of Mexico, Vol. 2. Smithsonian Institution particular Ambassador Thomas H. Malcomson, for securing Press, Washington, D.C travel funds to England for R.A.C.; to the Entomology De- Krombein, K. V. 1992. Host relationships, ethology and sys- partment personnel of the Natural History Museum, Lon- tematics of Pseudomethoca Ashmead (Hymenoptera: don, BM(NH) for providing R.A.C. with working facilities Mutillidae, Andrenidae, Hahctidae and Anthophondae). and assistance during his visit. We also thank Indra Proceedings of the Entomological Society of Washington Candanedo and Roberto Arango, Instituto Nacional de 94(1): 91-106. Recursos Naturales Renovables, who aided us with permits Maeklin, F. W. 1856. Om hermafroditism bland insekterna, and logistics in the Darien National Park. This project was samt beskrifning ofver en i Helsingfors funnen financed in by Vicerectoria de Investigacion y Postgrado, part hermafrodit af Mutilla obscura Finska Universidad de Panama, Fund No. 1-4500-91-12. Our thanks Nyl. ofver. af Vetenskaps-Societens Forliandlingar 3: 106-112. to Karl V. Krombein, Smithsonian Institution, Annette Aiello Mann, W. M. 1915. A gynandromorphous mutillid from and David W. Roubik, STRI, and an anonymous reviewer, for Montana. Psyche 22: 178-180. reading and making comments on the manuscript, James Mickel, C. E. 1924. A revision of the mutillid wasps of the Coronado, now at STRI, for providing very valuable help in genera Myrmilloides and Pseudomethoca ocurring in the field, and to Angel Aguirre, STRI, for bibliographic assis- America North of Mexico. Proceedings of the United States tance and location of references on gynandromorphs. National Museum 64(15): 1-51. Mickel, C. E. 1928. Biological and taxonomic investigations LITERATURE CITED on the mutillid wasps. United States National Museum Bulletin 143: 1-351. Mickel, C. E. 1935. and records of mutillid Bischoff, H. 1931. Der typus der Mutilla dubia ¥.—ein Descriptions of the and Pseudomethoca gynander. Mitteilungen deutschen entomologischen wasps genera Myrmilloides Mutillidae). Transactions of the American gesellschaft 2(4): 54-56." (Hymenoptera: Cambra, R. A. and D. Quintero A. 1992. Velvet ants of Entomological Society 61: 383-398. Mickel, C. E. 1936. New and records of Nearctic Panama: distribution and systematics (Hymenoptera: species mutillid of the Annals the Mutillidae), pp. 459-478. In: Quintero A., D. & A. Aiello wasps genus Dasymutilla. of America 29: 29-60. (eds.) Insects of Panama and Mesoamerica: Selected studies. Entomological Society of E. 1952. The Mutillidae of British Guiana. Oxford University Press, Oxford. Mickel, C. (wasps) 105-150. Cambra, R. A. and D. Quintero A. 1993. Studies on Timulla Zoologica 37(3): G. 1990. the Mutillidae, and Ashmead New distribu- Nonveiller, Catalog of Myrmosidae (Hymenoptera: Mutillidae): Mexico Bradynobaenidae of the Neotropical region including tion records and synonymies, and descriptions of previ- (Insecta: Hymenoptera). SPB Academic Publishing bv, unknown Pan-Pacific Entomologist 69(4): ously allotypes. The Netherlands, 150 pp. 296-310. Schuster, R. M. 1945. A new species of Pseudomethoca P. 1895-1896. vol. 2. Centrali Cameron, Hymenoptera, Biologia (Mutillidae) from the West Indies. Bulletin of the Americana, pp. 262-360, plates 13-14. Brooklyn Entomological Society 40: 7-8. Casal, O. H. 1965. Darditilla nuevo genero Neotropical de Wheeler, W. M. 1910. A gynandromorphous mutillid. Psyclie Sphaeropthalminae (Hym. Mutillidae). Eos, Madrid, 41: 17: 186-190. 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. 308 Journal of Hymenoptera Research

Table 1. Previously published cases of gynandromorphy in Mutillidae

Species Type Wing Country Ref. & tegula

Maeklin Mutilla europea obscura bilateral, right f, left m Finland 1856 DallaTorre & Friese 1899

Mickel 1928 Dasi/nmtilla cypris head, thorax, abdom. segm. USA [=hora] 5-7 m; 1-4 f

Mickel 1936 Dasymutilla gloriosa not described USA [=reperticia]

Dasymutilla vestita decussated, USA Mann 1915 [=euchroa head half:

=fulvohirta] right m, left f; thorax & abdomen:

right f; left m

Pseudomethoca frigida bilateral, USA Wheeler 1910 [=canadensis] right m, left f

TraumatomutiUa dubia head, thorax, 1st Am. Mer. Bischoff 1931 abdom. segment m, [Guyana] Mickel 1952 rest abdomen mosaic

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." b " P. 70, Appendix 1: entry for Barteria should read ". . . HFOWE Y Z . . . P. 74, Appendix 1: entry for Maieta should read "... B;D; Vasconcelos 1990, 1991, IP; Herre et al. 1986" uf R P. 78, Appendix 1: entry for Cordia should read ". . . Iy 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

research The journal of Hymenoptera Research invites papers of high scientific quality reporting comprehensive on all and Taxonomic aspects of Hymenoptera, including biology, behavior, ecology, systematics, taxonomy, genetics, morphology. a case is such as in papers describing single species are unlikely to be accepted unless strong evident, importance 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 of accompanied by the name 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 6962 uccTFR INDIAN^" !