(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Journal of Hymenoptera research"





Journal of 

Hymenoptera 




Volume 16, Number 1 April 2007 

ISSN #1070-9428 
CONTENTS 

CHIAPPINI, E. and S. V. TRIAPITSYN. Neotype designation for Anagrus atomus (Linnaeus) 

(Hymenoptera: Mymaridae) 1 

DEYRUP, L. D. and R. W. MATTHEWS. The effect of gland secretions on escape chewing in 
Melittobia (Hymenoptera: Eulophidae), including cross-species investigations 

GESS, F. W. Four new species of the wasp genus Celonites Latreille, 1802 (Hymenoptera: Vespi- 
dae: Masarinae) from south-western Africa, designation of neotype for C. michaelseni 
von Schulthess, 1923, species representation in Namibia, and key to species occurring 
in Namibia 11 

GESS, S. K. and F. W. GESS. Notes on nesting and flower visiting of some anthidiine bees (Hy- 
menoptera: Megachilidae: Megachilinae: Anthidiini) in southern Africa 30 

GIBSON, G. A. P., J. READ, and J. T HUBER. Diversity, classification and higher relationships 

of Mymarommatoidea (Hymenoptera) 51 

PAPP, J. Szepligeti's Cyclaulax types deposited in the Hungarian Natural History Museum 

(Hymenoptera: Braconidae: Braconinae) 147 

PITZ, K. M. and M. J. SHARKEY. Three new species of Cenocoeliinae (Hymenoptera: Braconi- 
dae) with novel morphological characteristics and habitat records 167 

SHEFFIELD, C. S. and S. M. WESTBY. The male oiMegachile nivalis Friese, with an updated key 
to members of the subgenus Megachile s. str. (Hymenoptera: Megachilidae) in North 
America 1 78 

VAN DAM, A. R., M. H. VAN DAM, and J M. HERATY A comparison of pyrethrum fogging 

and screen-sweep netting of micro-Hymenoptera in southern California chaparral . . 192 

OBITUARY: 

Marjorie Chapman Townes 28 March 1909-8 October 2006 206 



INTERNATIONAL SOCIETY OF HYMENOPTERISTS 

Organized 1982; Incorporated 1991 



OFFICERS FOR 2007 

Michael E. Schauff, President 

James Woolley, President-Elect 

Michael W. Gates, Secretary 

Justin O. Schmidt, Treasurer 

Gavin R. Broad, Editor 



Subject Editors 
Symphyta and Parasitica Aculeata 

Biology: Mark Shaw Biology: Sydney Cameron 

Systematics: Andrew Deans Systetnatics: Wojciech Pulawski 

All correspondence concerning Society business should be mailed to the appropriate officer at the 
following addresses: President, Plant Sciences Institute, Bldg. 003, Rm. 231 BARC-West, Beltsville, 
MD 20705, USA; Secretary, Southwestern Biological Institute, 1961 W. Brichta Dr., Tucson, AZ 85745, 
USA; Treasurer, PO Box 37012, c/o Smithsonian Institution, MNMH, MRC168, Washington, DC 
20013-7012, USA; Editor, Dept. of Entomology, The Natural History Museum, Cromwell Road, Lon- 
don SW7 5BD, UK. 

Membership. Members shall be persons who have demonstrated interest in the science of entomol- 
ogy. Annual dues for members are US$45.00 per year (US$40.00 if paid before 1 February), payable 
to The International Society of Hymenopterists. Requests for membership should be sent to the Trea- 
surer (address above). Information on membership and other details of the Society may be found on 
the World Wide Web at http://hymenoptera.tamn.edu/ish/. 

Journal. The Journal of Hymenoptera Research is published twice a year by the International Society of 
Hymenopterists, % Department of Entomology, Smithsonian Institution, Washington, D.C. 20560- 
0168, U.S.A. Members in good standing receive the Journal. Nonmember subscriptions are $60.00 
(U.S. currency) per year. 

The Society does not exchange its publications for those of other societies. 

Please see inside back cover of this issue for information regarding 
preparation of manuscripts. 



Statement of Ownership 

Title of Publication: Journal of Hymenoptera Research. 

Frequency of Issue: Twice a year. 

Location of Office of Publication, Business Office of Publisher and Owner: International Society of 

Hymenopterists, Department of Entomology, Smithsonian Institution, 10th and Constitution 

NW, Washington, D.C. 20560-0168, U.S.A. 
Editor: Gavin R. Broad, Department of Entomology, The Natural History Museum, Cromwell Road, 

London SW7 5BD, UK. 
Managing Editor and Known Bondholders or other Security Holders: none. 



This issue was mailed 25 April 2007 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 1-6 

Neotype Designation for Anagrus atomus (Linnaeus) 
(Hymenoptera: Mymaridae) 

Elisabetta Chiappini* and Serguei V. Triapitsyn 

(EC) Istituto di Entomologia e Patologia vegetale, Facolta di Agraria, Universita Cattolica del Sacro 

Cuore, Piacenza, Italy; email: elisabetta.chiappini@unicatt.it 

(SVT) Entomology Research Museum, Department of Entomology, University of California, 

Riverside, California 92521-0314, USA; email: serguei@ucr.edu 

*Author for correspondence. 



Abstract. — A neotype is designated for Anagrus atomus (Linnaeus), the type species of the 
common and widespread fairyfly genus Anagrus Haliday (Hymenoptera: Mymaridae). An 
illustrated description of the neotype specimen, collected at the type locality in Uppsala, Sweden, 
is provided. The taxonomic status of A. atomus is discussed, with particular reference to the closely 
related species A. ustulatus Haliday. 



Anagrus atomus (Linnaeus) (Hymenop- 
tera: Mymaridae) is an economically im- 
portant egg parasitoid of various crop- 
damaging leafhoppers (Hemiptera: Cica- 
dellidae) in the genera Arboridia Zachvat- 
kin, Edwardsiana Zachvatkin, Empoasca 
Walsh, Erythroneura Fitch, Neoaliturus Dis- 
tant, Zygina Fieber, and Zyginidia Haupt 
(Vidano and Arzone 1988, Triapitsyn 1998). 
Anagrus atomus has been recorded from 
numerous leafhopper species, sometimes 
due to misidentifications of both the host 
and parasitoid. It is a widely distributed 
species, present throughout Europe and 
also in Asia (China, Iran, Israel, Kyrgyz- 
stan, Pakistan, Republic of Korea, Eastern 
Russia, Turkey, Turkmenistan), America 
(Argentina, Canada, Chile, USA), Africa 
(Cape Verde Islands, Egypt), and Austra- 
lasia (New Zealand) (Triapitsyn and Bere- 
zovskiy 2004). Probably it was unintention- 
ally introduced into countries such as 
Argentina, Chile, and New Zealand. 

Linnaeus described Ichneumon atomus in 
1767. His very brief description (p. 941), in 
which he specified that the habitat is 
Uppsala, translated from Latin, is: "it is 
variegated pale and fuscous, it is smaller 
than Acarus sirene, so small that it is visible 



only when moving and it can be numbered 
among the smallest winged insects". This 
description of course could fit any small, 
pale microhymenopteran in several fami- 
lies. Therefore, a study of its type is needed 
but unfortunately, as Fitton (1978) and 
Graham (1982) stated, it is not present in 
the collection of Linnaeus owned by the 
Linnean Society of London, England. 

When Haliday (1833) defined the genus 
Anagrus he included two new species (A. 
ustulatus and A. incarnatus) and designated 
Ichneumon atomus as the type species of 
Anagrus but did not specify whether he 
had studied its type or not. His redescrip- 
tion of A. atomus is as brief as that of 
Linnaeus. Besides measurements of the 
body and the wings, he only stated that 
the head, the apex of the antennae, the 
prothorax and the "anus" are fuscous 
while the wings are hyaline and have 
a beautiful fringe. 

Bakkendorf (1926) synonymized almost 
all the previously described species of 
Anagrus under A. incarnatus. Debauche 
(1948), in contrast, re-established A. atomus 
as a valid species and redescribed it. He 
also synonymized A. ustulatus under A. 
atomus, unfortunately without mentioning 



2 Journal of Hymenoptera Research 

whether or not he had examined Haliday's ty), but were "unable to find anything like 

or Linnaeus' types (we suppose that he Mymaridae in their holdings" (M. Eriks- 

hadn't). son, pers. comm.). Third, specimens ac- 

Chiappini (1987) redescribed A. atomus cording to Haliday's (1833) brief redescrip- 
based on specimens from the Debauche tion as well as to Debauche's (1948) and 
collection and also on other specimens she Graham's (1982) concept of A. atomus were 
captured in traps and reared from grape captured in Uppsala, Sweden, the type 
leaves in Italy, all of which were identified locality of Ichneumon atomus, by Fredrik 
as A. atomus in accordance with the earlier Ronquist, formerly of the Department of 
concepts of this species (Debauche 1948, Systematic Zoology, Evolutionary Biology 
Viggiani 1970, Graham 1982). She did not Centre, Uppsala University. Several other 
designate a neotype, as, at that time, the Anagrus species were also captured at the 
case could not be included in the "circum- type locality (Triapitsyn and Berezovskiy 
stances admitted" specified in article 75 of 2004) but, of these, the only species 
the International Code of Zoological No- belonging to the atomus species group 
menclature (1985). Besides, her 1987 pub- (Chiappini 1989) was Anagrus ustulatus 
lication was not a "revisory work", and the (see Comments for the diagnosis), 
type of A. atomus could still be in Uppsala Therefore, considering that the identity 
(Graham 1982). In the same paper (Chiap- of A. atomus has long been in doubt, that no 
pini 1987), based on ecological as well as specimen(s) of Ichneumon atomus are pres- 
morphological features, she recognized e nt in either the Linnaeus collections at 
another distinct, then unnamed species Uppsala or London, that no neotype has 
which subsequently (Chiappini 1989) ever been designated for the type species of 
proved to correspond to A. ustulatus. By Anagrus, that all described species of this 
then Graham (1982) had already reinstated genus (for which type specimens exist) 
A. ustulatus as a valid taxon, designated have been carefully revised by us, and that 
a lectotype for it, and stated that it differed f res h material from the original type 
from A. atomus by its darker coloration, locality is available, it now seems appro- 
wider fore wings and, in females, by priate to designate here a neotype for A. 
different proportions of the funicle articles, atomus (Linnaeus). Its description follows; 

Lately, some doubts have been raised an abbreviation used in the text is: F = an 

whether A. ustulatus, the most closely antennal funicle article. 

related species to A. atomus, is really 

a different species because definitions of Anagrus (Anagrus) atomus (Linnaeus) 

both taxa seemed uncertain, largely due to ._,. „ „. 

(rigs 1—3) 

unavailability of the type material of A. . , , , . ' _ „,„ 

T , ,. . , . Ichneumon atomus Linnaeus, 1767:941. 

atomus. In addition, other circumstances , , , T . x TT ,., 100 ~ „.„ 

Aiiayus atomus (Linnaeus): Hahday, 1833: 347; 

have changed since Chiappini (1987) pub- Chiappini/ 1989: 102 -104 (diagnosis, syno- 

hshed the first paper on the subject. First, nyms and Hst of earlier cita tions); Triapitsyn 

a lot of revisory papers on Anagrus were and Berezovskiy, 2004 (distribution), 
published by Chiappini (1989), Chiappini 

et al. (1996), Chiappini and Lin (1998), Type material. — Neotype female of Ich- 

Triapitsyn (1997, 1998, 1999, 2001), and neumon atomus Linnaeus, 1767, here desig- 

Triapitsyn and Beardsley (2000). Second, nated in accordance with ICZN Article 75 

Mats Eriksson (curator of the Zoology (ICZN 1999), on slide, labelled: 1. "Ichneu- 

Section) and Hans Mejlon (curator of the mon atomus Linnaeus, 1767 = Anagrus 

entomological collections) thoroughly atomus (Linnaeus 1767) (Hymenoptera: 

searched the Linnaeus collection at the Mymaridae) NEOTYPE [female symbol] 

Museum of Evolution (Uppsala Universi- Des. by S. Triapitsyn & E. Chiappini 



Volume 16, Number 1, 2007 



2003"; 2. "SWEDEN: Uppsala, Hagadalen, 
26.viii-5.ix.1990, F. Ronquist, MT baited 
with rotten meat. Mounted at UCR/ERM 
by V. V. Berezovskiy 2002 in Canada 
balsam". The neotype was borrowed from 
the Canadian National Collection of In- 
sects, Ottawa (CNCI). By agreement with 
John Huber at the CNCI the neotype will 
be deposited in the Museum of Evolution, 
Uppsala University, Uppsala (UZIU). The 
neotype is in good condition, mounted 
in Canada balsam under two coverslips, 
one containing the wings (detached from 
the body), and the other the rest of the 
body (cleared in KOH prior to slide 
mounting). 

Other material studied. — Three other spe- 
cimens of A. atoimis were collected at or 
near the same locality as the neotype. Their 
collection data and depositories are as 
follows: 1 female on slide [CNCI]: SWE- 
DEN: Uppland Uppsala, Hagadalen, 17- 
26.viii.1990, F. Ronquist, MT. 1 female on 
card [CNCI]: SWEDEN: Uppsala, Hagada- 
len, 26.viii-5.ix. 1990, F. Ronquist, MT bai- 
ted with rotten meat (same data as the 
neotype). 1 female on card [Entomology 
Research Museum, University of Califor- 
nia, Riverside, California, USA (UCRC)]: 
SWEDEN: Uppland Uppsala, Eriksberg, 
30.vii-ll.viii.1986, F. Ronquist, MT/PT. 
Two females and a male in the Oxford, 
England, part of the Haliday collection, 
labelled respectively as W21 "Anagrus 
atomus Linn Haliday Coll.", W20, and 
W16 were also examined. 

Description. — Color: Head brown, except 
vertex mostly light brown (stemmaticum 
brown), eyes and ocelli red; scape and 
pedicel light brown, flagellum brown 
(apical flagellomeres slightly darker); pro- 
notum, posterior half of mesoscutum, 
anterior scutellum, metanotum and propo- 
deum light brown, anterior half of meso- 
scutum and axillae brown, posterior scutel- 
lum pale; wing venation brown; legs light 
brown (tarsi a little darker); gastral terga 
brown, with light brown membranous 
bands between them. 




Figs 1-2. Anagrus atomus (Linnaeus), neotype fe- 
male. 1. Antenna. 2. Fore wing. 

Head: About as wide as mesosoma. 
Antenna (Fig. 1) sparsely setose; scape 3.6 
x as long as wide and 2.2 x as long as 
pedicel; Fl oval, much shorter than pedicel 
and shortest of funicle articles; F2 a little 
longer than F3 and slightly shorter than F4 
or F5 which are equal in length, F6 longest 
and broadest of funicle articles; longitudi- 
nal sensilla on F4 (1), F5 (1) and F6 (2); 
clava a little longer than two preceding 
articles combined, with three longitudinal 
sensilla positioned subapically. 

Mesosoma: A little shorter than meta- 
soma. Mesoscutum finely longitudinally 
striate, without adnotaular setae. Fore 
wing (Fig. 2) 6.8 x as long as wide; distal 
macrochaeta about 2.5 x length of proximal 
macrochaeta; fore wing blade slightly 
infuscated behind venation but otherwise 
hyaline, with distinct bare area in broadest 
part next to posterior margin, discal micro- 
trichia arranged in 3 or 4 irregular rows; 
longest marginal cilia 2.9 x maximum fore 
wing width. Hind wing hyaline; disc with 
a few microtrichia at apex and a row of 
microtrichia along posterior margin. 

Metasoma: Ovipositor almost reaching 
mesophragma anteriorly and a little ex- 



Journal of Hymenoptera Research 



) 




Fig. 3. Anagrus atomus (Linnaeus), female (from egg 
of a grape leafhopper, Verago, Piacenza, Italy). 



serted beyond apex of gaster posteriorly 
(by about 1/15 of its total length). External 
plates of ovipositor with one seta each. 
Ovipositor length/ foretibia length 1.9:1. 

Measurements (in micrometers, urn). — 
Body length (taken before slide mounting) 
559; head length/width (length taken 
before slide mounting) 100:161; mesosoma 
209; metasoma 281; ovipositor 236. Anten- 
na: scape 75; pedicel 34; Fl 17; F2 44; F3 39; 
F4 48; F5 48; F6 52; clava 107. Fore wing 
length/width 546:80; longest marginal cilia 
233. Hind wing length/width 500:23; lon- 
gest marginal cilia 179. Legs (given as coxa, 
trochanter, femur, tibia, tarsus): fore 66, 42, 
130, 124, 155; middle 48, 39, 124, 173, 158; 
hind 70, 40, 120, 188, 164. 

Diagnosis. — Anagrus atomus can be dis- 
tinguished from all other species of the 
atomus species group, as defined by Chiap- 
pini et al. (1996), by the following combi- 
nation of features: F3 without longitudinal 
sensilla, F4 longer than the previous 
articles and bearing one longitudinal sen- 
sillum, F2 and F3 together much longer 
than F6, at least by half their combined 
length, mesoscutum without adnotaular 



setae, hairless area present only at broadest 
part of fore wing, and fore wing length/ 
width less than 10. 

Comments. — Specimens of A. atomus 
from vineyards in southern Europe (e.g., 
Italy and France) may show a different 
color pattern on the gaster, with the terga 
from about fourth to seventh yellow 
(Fig. 3), whereas the northern forms ap- 
pear to be slightly darker or more uni- 
formly colored. 

The three specimens labeled as W21, 
W20, and W16 in the Oxford part of the 
Haliday collection clearly belong to A. 
atomus, as correctly stated by Graham 
(1982). 

We also re-examined the lectotype male 
of A. ustulatus Haliday (n 70), together 
with the two female specimens (n 72 and 
73) under this name in the Haliday 
collection at the National Museum of 
Ireland, in Dublin, in order to verify the 
possible synonymy of A. ustulatus under 
A. atomus. The lectotype agrees with what 
had already been stated by Graham (1982) 
and Chiappini (1989); namely, the ratio 
between the lengths of the macrochaetae 
on the fore wing marginal vein is greater 
than two, the fore wing has a hairless area 
on the disc, and it is very wide compared 
to that of A. atomus. The male genitalia, 
which had already been studied by 
Chiappini (1989) who stated (contrary to 
Graham) that they were typical of the 
atomus species group, were not checked 
again because to do so would require 
ungluing the type specimen. In contrast to 
the lectotype of A. ustulatus the ratio 
between the lengths of the macrochaetae 
is less than two in the females n 72 and 
73, as is typical of members of the 
incarnatus species group of Anagrus. 
Therefore, these two females cannot be 
conspecific with the lectotype of A. ustu- 
latus as they belong to a different species 
group. In addition, the fore wings of 
females n 72 and 73 are narrower and 
without a bare area on the disc and F2 is 
the longest, unlike either A. atomus or A. 



Volume 16, Number 1, 2007 



ustulatus. Specimens n 72 and 73 belong 
to A. incarnatus, according to the most 
recent concept of this species (Triapitsyn 
1997). 

Therefore, the species concept for A. 
ustulatus should be based only on the 
lectotype designated by Graham (1982). 
This male has fore wing proportions 
different from A. atomus males but equal 
to those of the males of the Anagrus species 
found on bramble and rose (Chiappini 
1987) and whose females differ from those 
of A. atomus by F4 being as long as F3 and 
without longitudinal sensilla (Chiappini 
1989), and by the fore wing being wider. 
Many other data, both ecological (Chiap- 
pini 1987) and chemical, support the 
separation of A. atomus from A. ustulatus. 
For example, the cuticular hydrocarbon 
patterns in these two species differ consid- 
erably, as the second species displays 
a notable amounts of alkenes not present 
in the first's pattern (Floreani et al. in 
prep.). On the basis of this knowledge, we 
treat A. ustulatus as specifically distinct 
from A. atomus, even though we know that 
more studies, particularly of field popula- 
tions of Anagrus, are needed to better 
characterize these two species. 

ACKNOWLEDGMENTS 

We thank John T. Huber (CNCI) for the loan of 
material and review of an earlier draft of the 
manuscript, and Mats Eriksson and Hans Mejlon 
(UZIU) for searching for the type of A. atomus and 
Mymaridae in general in the Linnean collection in 
Uppsala. Vladimir V. Berezovskiy (UCRC) helped 
with specimen preparation. 

LITERATURE CITED 

Bakkendorf, O. 1926. Recherches sur la biologie de 
YAnagrus incarnatus Haliday. Annates de Biologic 
Lacustre 14: 249-270. 

Chiappini, E. 1987(1986-1987). Ricerche sulla variabi- 
lita di Anagrus atomus (L.) (Hymenoptera Mymar- 
idae) e di una specie affine presente sul rovo. 
Bollettino di Zooktgia Agraria e di Bachicoltura, Serif 
II 19: 71-97. 

. 1989. Review of the European species of the 

genus Anagrus Haliday (Hymenoptera Chalcidoi- 



dea). Bollettino di Zoologia Agraria e di Bachicoltura, 
Serie II 21: 85-119. 

and N.-Q. Lin. 1998. Anagrus (Hymenoptera: 

Mymaridae) of China, with descriptions of nine 
new species. Annals of the Entomological Society of 
America 91: 549-571. 

, S. V. Triapitsyn, and A. Donev. 1996. Key to 

the Holarctic species of Anagrus Haliday (Hyme- 
noptera Mymaridae) with a review of the 
Nearctic and Palearctic (other than European) 
species and descriptions of new taxa. journal of 
Natural History 30: 551-595. 

Debauche, H. R. 1948. Etude sur les Mymarommidae 
et les Mymaridae de la Belgique. Memoires du 
Musee Royal d'Histoire Naturelle de Belgique 108: 
1-248. 

Fitton, M. G. 1978. The species of "Ichneumon" 
(Hymenoptera) described by Linnaeus. Biological 
Journal of the Linnean Society 10: 361-383. 

Floreani, C, F. Pavan, and F. Nazzi. 2006. Analysis of 
cuticular hydrocarbons in two Anagrus species 
(Hymenoptera: Mymaridae) as a tool to improve 
their correct identification. The Canadian Entomol- 
ogist 138: 348-356. 

Graham, M. W. R. de V. 1982. The Haliday collection 
of Mymaridae (Insecta, Hymenoptera, Chalcidoi- 
dea) with taxonomic notes on some material in 
other collections. Proceedings of the Royal Irish 
Academy B 82: 189-243. 

Haliday, A. H. 1833. An essay of the classification of 
the parasitic Hymenoptera of Britain, which 
correspond with the Ichneumones miuitti of Lin- 
naeus. Entomological Magazine 1: 333-350. 

ICZN (International Commission on Zoological No- 
menclature). 1985. International Code of Zoological 
Nomenclature. Third edition. International Trust for 
Zoological Nomenclature. Natural History Muse- 
um, London. 338 pp. 

ICZN (International Commission on Zoological No- 
menclature). 1999. International Code of Zoological 
Nomenclature. Fourth edition. International Trust 
for Zoological Nomenclature. Natural History 
Museum, London. 306 pp. 

Linnaeus, C. 1767. Systema Naturae. Editio duodecimo 
reformata 1 (2). Holmiae. 533-1327 + 32 pp. 

Triapitsyn, S. V. 1997. The genus Anagrus (Hymenop- 
tera: Mymaridae) in America south of the United 
States: a review. CEIBA 38: 1-12. 

. 1998. Anagrus (Hymenoptera: Mymaridae) egg 

parasitoids of Erythroneura spp. and other leaf- 
hoppers (Homoptera: Cicadellidae) in North 
American vineyards and orchards: a taxonomic 
review. Transactions of the American Entomological 
Society 124: 77-112. 

. 1999 (2000). A review of the species of Anagrus 

Haliday, 1833 (Hymenoptera: Mymaridae) col- 
lected by A. A. Ogloblin in Argentina. Russian 
Entomological journal 8: 213-222. 



Journal of Hymenoptera Research 



— . 2001. Review of the Australasian species of 
Anagrus (Hymenoptera Mymaridae). Belgian Jour- 
mil of Entomology 3 2: 267-289. 

— and J. W. Beardsley. 2000. A review of the 
Hawaiian species of Anagrus (Hymenoptera: 
Mymaridae). Proceedings of the Hawaiian Entomo- 
logical Society 34: 23-48. 

— and V. V. Berezovskiy. 2004. Review of the 
genus Anagrus Haliday, 1833 (Hymenoptera: My- 
maridae) in Russia, with notes on some extralim- 
ital species. Far Eastern Entomologist 139: 1-36. 



Vidano, C. and A. Arzone. 1988. Natural enemies of 
Zyginidia pallida (Rhynchota Auchenorrhyncha). 
Pp. 581-590 in: Vidano, C, and A. Arzone, eds. 
Proceedings of the 6th Auchenorrhyncha Meeting, 
Turin, Italy, September 7-21, 1987. 

Viggiani, G. 1970. Ricerche sugli Hymenoptera 
Chalcidoidea XXIV. Sul valore tassinomico 
dell'organo copulatore nei Mimaridi del gen- 
ere Anagrus Hal. Bollettino del Laboratorio di 
Entomologia Agraria "Filippo Silvestri", Portici 25: 
10-18. 



J. HYM. RES. 

Vol. 16(1), 2007, pp. 7-10 

The Effect of Gland Secretions on Escape Chewing in Melittobia 
(Hymenoptera: Eulophidae), Including Cross-species Investigations 

Leif D. Deyrup and Robert W. Matthews* 

Department of Entomology University of Georgia, 413 Biological Sciences Building, Athens, GA 

30602-2603 USA; email: lsdeyrup@hotmail.com; rniatthew@uga.edu 

^Address for correspondence: Robert W. Matthews, Department of Entomology, University of 

Georgia, 413 Biological Sciences Building, Athens, GA 30602-2603; tel (706) 542-2311; fax (706) 542- 

2640; rmatthew@uga.edu 



Abstract. — Melittobia is a genus of small, gregarious idiobiont parasitoids in the family 
Eulophidae. Following emergence as adults, females form circles in which they cooperate to chew 
an escape hole from the host cells in which they developed. Dry milked crude venom, which could 
contain constituents from the alkaline gland as well as the venom reservoir, has been shown to elicit 
chewing in M. digitata. Here we investigated whether a related species (M. femorata) chewed in 
response to compounds in its dissected venom reservoir plus alkaline-gland, and whether crude 
venom milked from a member of another species group (M. australica) would also elicit chewing in 
M. digitata. Melittobia femorata chewed significantly more at combined gland and reservoir extract- 
marked spots than at controls. To examine the crude venom's effect across species we marked spots 
with milked M. australica venom, and introduced female M. digitata wasps. These milked crude 
venom spots elicited chewing similar to that elicited by that of M. digitata marked spots, and the 
response to either's venom was significantly different from blank controls. Possible reasons for the 
lack of a high level of specificity in the chewing response to a pheromone are discussed. 



Melittobia Westwood is a cosmopolitan 
genus of small gregarious parasitic wasps 
(Balfour-Browne 1922, Buckell 1928, 
Dahms 1984b). They are commonly found 
attacking mud dauber (Hymenoptera: 
Sphecidae) prepupae and their associates 
(Matthews 1997), but also attack a wide 
range of solitary bees and wasps and their 
associates (Balfour-Browne 1922, Krom- 
bein 1967). 

When attacking a mud dauber wasp 
they have to escape from the thick-walled 
mud nest, yet females do not have notice- 
ably well developed mandibles. Donovan 
(1976) observed M. haivaiiensis Perkins 
females circled around another female that 
had started chewing a pit in the mud wall, 
and speculated that they then cooperated 
in chewing their way out. Subsequently, 
such cooperative chewing has been ob- 
served in several Melittobia species (L.D. 
Deyrup unpublished). 



Deyrup et al. (2005) reported that 
chewed pits invariably had associated 
sting marks and showed that a putative 
pheromone in the milked crude venom, 
which most likely contains constituents of 
the alkaline gland as well as the venom 
reservoir, of M. digitata Dahms elicited 
chewing from conspecific females. Because 
similar chew pits made by other species of 
Melittobia also typically show sting marks 
in their centers (Deyrup unpublished), we 
decided to investigate whether extracted 
venom components would elicit chewing 
in a closely related species, M. femorata 
Dahms (Dahms 1984a). 

Such chewing, if demonstrated, could be 
in response to a normal constituent of 
crude venom, or blend of odors. Regard- 
less, it is difficult to envision selection 
pressure sufficient to cause evolutionary 
divergence in such a cue, since there 
appear to be no negative effects of co- 



Journal of Hymenoptera Research 



operative escape chewing, even among 
unrelated females. 

METHODS 

In general the methods follow those 
described by Deyrup et al. (2005). Melitto- 
bia australica Girault responded to the 
venom-milking procedures described in 
Deyrup and Matthews (2003), yielding 
adequate amounts of crude venom for the 
experiment. However, M. femorata does not 
respond to this venom-milking technique. 
Therefore, as an alternative we dissected 
the lower reproductive tract of females in 
insect saline [10 mM sodium phosphate, 
0.9% (w/v) NaCl, pH 8.0]. While there are 
many possible pheromone sources in the 
female reproductive system, the two most 
likely are the alkaline gland and venom 
reservoir. These were separated from the 
ovipositor and combined for use in the 
experiment. Since milked crude venom 
used in previous work could contain 
a combination of the fluids contained in 
both organs we decided to combine them 
for this experiment. 

As described in Deyrup et al. (2005), 20 
plastic box lids were prepared for the first 
set of experimental treatments by making 
four pin indentations, one in each corner of 
the inner side. We then smeared the 
combined alkaline gland and venom reser- 
voir dissected from a single female of M. 
femorata into one pin indentation and 
repeated this using a fresh female applied 
to the pit on the opposite corner. The other 
two pits served as controls for chewing 
stimulated by the pit alone as in Deyrup et 
al. (2005). Treated lids were then placed on 
20 boxes of 250-300 1-3 day old mated M. 
femorata females and left for 12 hours in 
complete darkness at 25 C, after which 
they were examined for evidence of chew- 
ing at each of the four pits. 

To determine if M. australica or M. 
digitata would be stimulated to chew by 
M. australica crude venom, we set up a two 
more series of boxes. Three corner circles 
were drawn on the lids as in Deyrup et al. 



(2005) and randomly assigned one of three 
treatments. One circle received 1 FED 
(female equivalent dose) of milked M. 
australica venom. In another circle a clean 
pin rub served as a negative control, and 
the third circle was 1 FED of milked venom 
from a M. digitata. Fifteen of these lids were 
prepared for each series, and placed on 
boxes of 250-300 females as before. Boxes 
were then placed in absolute darkness at 
25 C, and scored for signs of chewing 
12 hours later. 

Cochran Q tests were used to analyze 
chewing frequencies (Statistica 6.0). This 
test was chosen because the treatments 
were paired, and the results were scored as 
chewing presence or absence (1 or re- 
spectively). 

RESULTS 

The experimental group containing 
smeared M. femorata venom reservoir and 
alkaline gland contents elicited chewing 
from M. femorata in at least one of the two 
treated pits in 19 of the 20 replicates. In 
contrast, both control pits were chewed on 
only two occasions out of 20. These 
differences were highly significant (P 
<0.0001, Q =17.0000, 1 df). In the series 
to determine if M. australica chewed at their 
own milked crude venom or the milked 
crude venom from M. digitata, there was no 
chewing what-so-ever at any treatment or 
control. 

In the experiment to examine if chewing 
was elicited in M. digitata by milked crude 
venom from M. australica, chewing oc- 
curred in 9 of the 15 replicates (Table 1). 
The overall Cochran test was significant (P 
<0.0031, Q =11.5556, 2 df). Therefore, 
using Fisher's test for multiple analyses, 
we ran pairwise Cochran tests that re- 
vealed a significant difference between the 
blank and M. australica venom (P <0.0047, 
Q =8.0, 1 df), and the blank and M. digitata 
venom (P <0.0143, Q =6.0, 1 df). There was 
no significant difference between chewing 
at the positive control, M. digitata venom, 



Volume 16, Number 1, 2007 



Table 1. Chewing response by M. digitata females 
after 12 hours exposure of treatment circles containing 
venom milked from either M. digitata or M. australica. 
Different letters in the "Significance" column indicate 
significant differences using a Cochran Q test p< 0.05, 
df = 1 (Statistica 6.0). 



Treatment 



Chewed Replicates Significance 



M. australica venom 9 

M. digitata venom 7 

Control 1 



15 

15 
15 



a 
a 
b 



and M. australica venom (P <0.3173, Q 
=1.0, 1 df). 

DISCUSSION 

The M. femorata chewing results in re- 
sponse to dissected M. femorata reproduc- 
tive tract organs suggest that M. femorata 
has a pheromone in its crude venom that 
stimulates chewing at a particular spot. 
This adds support to the idea that chewing 
in response to crude venom components 
evolved before the speciation event that 
separated M. digitata and M. femorata. 

The negative results for M. australica 
chewing are hard to interpret since the 
design does not allow us to test a "lack of 
stimulus". The species has been observed 
to cooperatively chew. There could be 
many reasons for the crude venom not to 
be attractive such as the possibility that 
other factors are necessary or that chewing 
only occurs during a particular unestab- 
lished window of opportunity. More rig- 
orous experimentation would be required 
to establish that the crude venom is not at 
least a part of the chewing stimulus. 

The positive results for the attraction of 
M. australica crude venom for M. digitata 
females (Table 1) might seem surprising 
since the two species belong to different 
species groups (Dahms 1984a). Especially 
since we were unable to elicit chewing in 
response to milked crude venom for M. 
australica. However, there is little reason 
to expect that such a pheromone, if there 
is a pheromone for chewing in M. aus- 
tralica, would not be conserved, since 



a mutation could leave carriers trapped 
in the host's cell. Even if there is no such 
pheromone present in crude venom for 
chewing in M. australica, the chemical that 
stimulates chewing for M. digitata could 
be one that is stable and under selection 
for another purpose (e.g., perhaps con- 
taining a constituent causing develop- 
mental delay in the host [Deyrup et al. 
2003]). Components of other pheromones 
appear to have been conserved in Melitto- 
bia. Matthews et al. (1985) found that 
females of M. digitata, M. femorata, and M. 
australica were attracted to non-conspecif- 
ic as well as conspecific males in choice 
tests. Further work should be done on 
investigating the source of the pheromone 
in which either the venom reservoir or the 
alkaline gland is presented alone and 
together. 

ACKNOWLEDGEMENTS 

Mark Deyrup, Jan Matthews, and Jorge Gonzalez 
provided valuable discussion and editorial help. This 
work was supported in part by a grant from the 
National Science Foundation to R. W. Matthews. 

LITERATURE CITED 

Balfour-Browne, M. A. 1922. On the life-history of 
Melittobia acasta Walker, a chalcid parasite of bees 
and wasps. Parasitology 14: 349-369. 

Buckell, E. R. 1928. Notes on the life-history and habits 
of Melittobia chalybii Ashmead (Chalcidoidea: 
Elachertidae). Pan-Pacific Entomologist 5: 14-22. 

Dahms, E. C. 1984a. Revision of the genus Melittobia 
(Chalcidoidea: Eulophidae) with the description 
of seven new species. Memoirs of the Queensland 
Museum 21: 271-336. 

. 1984b. A review of the biology of species in 

the genus Melittobia (Hymenoptera: Eulophidae) 
with interpretations and additions using observa- 
tions on Melittobia australica. Memoirs of the 
Queensland Museum 21: 337-360. 

Deyrup, L. D. and R. VV. Matthews. 2003. A simple 
technique for milking the venom of a small 
parasitic wasp, Melittobia digitata (Hymenoptera: 
Eulophidae). Toxicon 42: 217-218. 
— , M. Deyrup, and R. VV. Matthews. 2003. 
Paralyzation and developmental delay of a facti- 
tious host by Melittobia digitata (Hymenopter: 
Eulophidae) journal of Entomological Science 38: 
703-705. 



10 



Journal of Hymenoptera Research 



— R. W. Matthews, and J. M. Gonzalez. 2005. 
Cooperative chewing in a gregariously develop- 
ing parasitoid wasp, Melittobia digitata Dahms, is 
stimulated by structural cues and a pheromone in 
crude venom extract. Journal of Insect Behavior 18: 
293-304. 
Donovan, B. J. 1976. Co-operative material penetration 
by Melittobia hawaiiensis (Hymenoptera: Eulophi- 
dae) and its adaptive significance. New Zealand 
Entomologist 6: 192-193. 



Krombein, K. V. 1967. Trap-nesting Wasps and Bees: Life 
Histories, Nests, and Associates. Smithsonian Press, 
Washington, DC. 

Matthews, R. W. 1997. Teaching ecological interac- 
tions with mud dauber nests. The American 
Biology Teacher 59: 152-158. 

, J. Yukawa, and J. M. Gonzalez. 1985. Sex 

pheromones in Melittobia parasitic wasps: Female 
response to conspecific and congeneric males of 3 
species, journal of Ethology 3: 59-62. 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 11-29 

Four New Species of the Wasp Genus Celonites Latreille, 1802 

(Hymenoptera: Vespidae: Masarinae) from South-western Africa, 

Designation of Neotype for C. michaelseni von Schulthess, 1923, Species 

Representation in Namibia, and Key to Species Occurring in Namibia 

Friedrich W. Gess 

Albany Museum, Grahamstown, 6139 South Africa 
email: f.gess@ru.ac.za; tel 0466222312 



Abstract. — Four new species of the genus Celonites Latreille, 1802 (Hymenoptera: Vespidae: 
Masarinae) are described from south-western Africa: heliotropii and pulcher from Namibia, 
kalahariensis from Namibia and the adjacent trans-Orange part (Gordonia) of the Northern Cape of 
South Africa, and arenarius from the north-western corner (Richtersveld) of the Northern Cape. A 
neotype is designated for the widespread, chiefly Namibian Celonites michaelseni von Schulthess, 
1923, with which C. gariepensis Gess, 1997 is sunk into synonymy. Namibian records are given for C. 
andrei Brauns, C. capensis Brauns, C. clypeatus Brauns, and C. tiiniidiscutellatus Gess, all better known 
from South Africa. Distribution maps are given for all nine species and forage plant records are 
included for eight. A key to the species of Celonites occurring in Namibia is given. 

Key words. — Hymenoptera, Vespidae, Masarinae, Celonites, new species, Namibia, southern Africa 



The genus Celonites Latreille was revised 
by Richards (1962) as part of his study of 
the Masarinae of the world. He dealt with 
a total of 26 species from the Palaearctic 
and Afrotropical regions, eight species 
being from southern Africa. Amongst 
a number of species mentioned but not 
examined by Richards was one additional 
southern African species, C. michaelseni von 
Schulthess from the present day Namibia, 
known only from the holotype which he 
rightly believed to have been destroyed in 
Hamburg during World War 2. 

Since 1962 seven additional species have 
been recognised from southern Africa, 
three described by Gess (1997) and four 
described in the present paper. The dis- 
covery of these species resulted from 
purposeful collecting in under-collected 
parts of the Western and Northern Cape, 
South Africa and particularly in Namibia. 
The overall known distribution of the 
genus in southern Africa has been given 
by Gess and Gess (2004b: Fig. 7). 



As no Celonites, other than the single 
specimen of C. michaelseni, have previously 
been recorded from Namibia, particular 
attention is paid to the fauna of that 
country from which eight species are now 
known: the here recognized and wide- 
spread C. michaelseni von Schulthess, C. 
heliotropii sp. nov., C. pulcher sp. nov., C. 
kalahariensis sp. nov. and four species 
previously known from South Africa, C. 
andrei Brauns, C. capensis Brauns, C. clypea- 
tus Brauns, and C. tumidiscutellatus Gess. It 
is highly probable that C. arenarius sp. nov., 
described from the southern bank of the 
Orange River, will also be found in 
Namibia. 

The key is restricted to those species 
occurring in Namibia. A key to all the 
southern African species was attempted 
but was found to be impracticable at the 
present time due to the paucity of material 
of some species. 

The notation used for expressing geo- 
graphic co-ordinates is as in the gazetteer 



12 Journal of Hymenoptera Research 

of The Times Atlas of the World (1981). The ventrally produced in proximal half 

figures before the stop are degrees, those (Fig. 10). Male: most of mandibles, entire 

after the stop are minutes; the stop is not labrum and clypeus; variably developed 

a decimal point. facial markings and in some specimens 

For purposes of plotting distributions, underside of proximal flagellomeres, lem- 

co-ordinates have been given in square on yellow. Male genitalia as in Fig. 1. 
brackets in the text for those localities for Description. — Female: Black. The follow- 

which none are given on the data labels. ing are dark reddish-brown: distal half of 

On a few data labels from collections mandible; underside of antennal club; 

other than that of the Albany Museum the pronotum; tegula; scutellum and median 

collecting locality is followed by degree part of metanotum; in some specimens 

latitude and degree longitude and by half- large spot anteriorly on mesopleuron, 

and quarter-degree reference letters (e.g., propodeal lamella laterally; in great major- 

321 8BB) according to the Degree Reference ity of specimens terga I - VI (except for 

System of Leistner and Morris (1976). As black bases); sterna I, II and VI and lateral 

this system is not universally understood and posterior margins of III - V; femur 

an attempt has been made here to find on (distally), tibia and tarsi of all legs. Wings 

a map the localities concerned and to add browned. 

in square brackets the co-ordinates ex- Length 5.8-7.3 mm (average of 4: 

pressed in the manner adopted in this 6.6 mm), length of front wing 4.4-5 mm 

paper (e.g., 3218BB [32.11S 18.54E]). (average of 4: 4.7 mm), hamuli 7 or 8; 

In listing the material examined, the length of extended tongue 3.9 mm. 
localities have been arranged, as far as Head 1.3X as wide as long (measured 

practicable, in north to south order within across eyes and from vertex to bottom of 

Namibia or, in the case of South Africa, emargination of clypeus respectively), 

within provinces. frons and clypeus not carinate. Clypeal 

Acronyms here used for institutions in disc markedly raised, with surface finely 

which material is housed are: AMG = reticulate punctate and with microscul- 

Albany Museum, Grahamstown, South tured interstices; frons, vertex and occiput 

Africa; CAS = California Academy of more coarsely reticulate punctate, with 

Sciences, San Francisco, United States of smooth and shiny interstices. Frons slightly 

America; NCP = National Collection of raised above and between antennal sockets 

Insects, Pretoria, South Africa; NNIC = arid very slightly depressed medially 

Namibian National Insect Collection, above swollen area. 

Windhoek, Namibia; ZMH = Zoologisches Pronotum, mesoscutum, mesopleuron, 

Museum Hamburg, Hamburg, Germany. scutellum and dorso-lateral part of propo- 

deum similarly punctured to frons and 

DESCRIPTIONS OF SPECIES AND vertex, shiny. Scutellum gently convex, 

COLLECTION DATA anteriorly not raised above level of meso- 
scutum. Tegula circa 1.9X as long as 

Celonites heliotropii Gess, new species maximum width, posteriorly narrowed 

Celonites sp. (undescribed): Gess and Gess, witn outer margin curving inwards to- 

2004a: 39 (flower visiting). wards rounded but acute posterior angle. 

Propodeal declivity markedly longitudi- 

Diagnosis.— Both sexes: relatively small nally rugoso-punctate. Lateral lamella of 

(5.6-7.3 mm); black, with pronotum, te- propodeum broad, with surface in same 

gula, scutellum and terga reddish brown; plane as adjacent median part, its outer 

clypeus and frons without carinae. Fore margin minimally curved, its apex trun- 

femur, particularly in female, postero- cate, separated from median part by 



Volume 16, Number 1, 2007 



13 



narrow incurved slit. Terga more finely 
and closely punctured than thorax; in- 
terstices microsculptured; postero-lateral 
angles produced, acutely pointed; hind 
margins entire (non-crenulate). 

Fore femur (Fig. 10, with for comparison 
the unmodified front leg of C. michaelseni, 
Fig. 11) postero-ventrally produced in 
proximal half; end of tibia when folded 
against femur coinciding with produced 
region; tarsus short (only 1.2X tibial 
length); underside of tibia with moderately 
dense, short setae, tarsus setose throughout 
but with setae particularly dense on un- 
derside of tarsomeres I - IV where forming 
stiff brush. 

Male: Black. The following are lemon 
yellow: most of mandible; entire labrum 
and clypeus; variably developed supracly- 
peal spot medially on frons; usually small 
to minute spot in lower half of ocular sinus; 
in some specimens underside of proximal 
flagellomeres. The following are various 
shades of reddish-brown: flagellomeres; 
pronotum (colour grading almost to yellow 
on humeral angle; tegula (colour grading 
almost to yellow anteriorly); scutellum and 
median part of metanotum; large spot 
anteriorly on mesopleuron; propodeal la- 
mella laterally; transverse bands on terga I 
- VI (colour of each band dark adjacent to 
black base, lighter posteriorly, grading 
almost to yellow on postero-lateral angles); 
sterna I - VI (partially) and VII (totally); 
femur (distally), tibia and tarsi of all legs 
(streaks on tibiae almost yellow). Wings 
lightly browned (paler than those of 
female). 

Length circa 5.6 mm; length of front 
wing circa 3.9 mm; hamuli 6; length of 
extended tongue circa 3 mm. 

More gracile than female but structurally 
similar, apart from usual more markedly 
swollen antennal club and more pro- 
nounced postero-lateral angles of terga. 

Genitalia in ventral view as in Fig. 1; in 
dorsal view with parameres distally sub- 
truncate, posterior margin of each para- 
mere gently concavely curved from round- 



ed inner posterior angle to protruding but 
rounded outer (lateral) posterior angle. 

Etymology. — The name heliotropii, geni- 
tive singular, is formed from the generic 
name of the plant Heliotropium tubulosum 
(Boraginaceae) to the flowers of which the 
wasp appears to be restricted for purposes 
of foraging for nectar or nectar and pollen. 

Material examined. — Holotype: 9, NAMIBIA: 
19 km SSW of Uis on road to Henties Bay 
(21.27S 14.45E), 18.iv.2002 (F. W. and S. K. Gess) 
(visiting white flowers of Heliotropium tubulosum 
E. Mey. ex DC, Boraginaceae) [AMG]. Para- 
types: NAMIBIA: near Palmwag (19.53S 
13.55E), 26.iii.2004, 1 9 (visiting white flowers 
of Heliotropium tubulosum E. Mey. ex DC, 
Boraginaceae); Two Palms, near Palmwag 
(19.53S 13.54E), 27.iii.2004, 8 99 (visiting white 
flowers of Heliotropium tubulosum); same local- 
ity, 28.iii.2004, 13 99 (visiting white flowers of 
Heliotropium tubulosum); Uis to Omaruru (21.14S 
15.00E), 14.iii.2004, 13 99 (12 99 visiting white 
flowers of Heliotropium tubulosum); 19 km SSW 
of Uis on road to Henties Bay (21.27S 14.45E), 
17.iv.2002, 6 99, (3 99 visiting white flowers of 
Heliotropium tubulosum; 1 9 associated with hole 
in ground; 1 9 on ground); same locality, 
18.iv.2002 , 8 99, 1 ; (7 99, 1 3 visiting white 
flowers of Heliotropium tubulosum; 1 9 on 
ground); Gross Spitzkuppe (21.51S 15.12E), 
19. iv. 2002, 6 99 (visiting white flowers of 
Heliotropium tubulosum); 77 km E of Henties 
Bay on road to Klein Spitzkuppe (21.54S 
14.58E), 19.iv.2002, 13 99 (visiting white flowers 
of Heliotropium tubulosum); 58 km SW of Usakos 
on road to Swakopmund (22.12S 15.10E), 
23.iv.2002, 9 99, 1 J (7 99, 1 $ visiting white 
flowers of Heliotropium tubulosum; 2 99 on 
ground next to Heliotropium tubulosum); 33 km 
by road from Swakopmund to Usakos, near 
Rossing Mtn. (22.34S 14.49E), 22.iv.2002, 1 9; 
same locality, 28. iv. 2002, 1 9 (at Heliotropium 
tubulosum) - (all F. W. and S. K. Gess) -- [all 
AMG]; Karibib District, 15 km W Karibib 
[21.56S 15.43E], 26.ii.1990, 1 £ Karibib District, 
55 km SW Usakos [22.16S 15.08E], l.iii.1990, 1 9; 
Karibib District, 65 km SW Usakos [22.20S 
15.05E], 24.ii.1990, 2 99, 1 J - (all W. J. Pulawski) 
[all CAS]; Damaraland, 6 km N Arandis (22.22S 
14.59E), 3-31.vii.1984, 1 9, 8.v.-5.vi.l984, 2 99, 
5.vi.-3.vii.l984, 2 99 (all J. Irish; H. Liessner); 
same locality, 9.iv. - 6.V.1985 (J. Irish, H. Rust), 1 



14 



Journal of Hymenoptera Research 



9; Swakopmund Dist, Rossing Mine (22.28S 
15.02E), 13.iii.-10.iv.1984, 2 99, 10.iv.-8.v.l984, 1 
9 (both J. Irish; H. Liessner); Swakopmund Dist., 
Upper Panner Gorge (22.29S 15.01E), lO.iv.- 
8.V.1984 (J. Irish; H. Liessner), 5 99; Swakop- 
mund Dist., Upper Ostrich Gorge (22.29S 
14.59E), 13.iii.-10.iv.1984, 2 99, lO.iv.- 8.V.1984, 
2 99 (both (J. Irish; H. Liessner); same locality, 
12.ii.-ll.iii.1985, 1 9, 1 $, 9.iv.- 6.V.1985, 1 9 
(both J. Irish; H. Rust); Swakopmund Dist., 
Lower Ostrich Gorge (22.30S 14.58E) 13.iii.- 
10. iv. 1984 (J. Irish; H. Liessner), 1 9; same 
locality, 12.ii.- ll.iii.1985, 1 3, 9.iv.-6.v.l985, 1 
9 (both J. Irish; H. Rust) - [all NNIC]. 

Geographic distribution. — (Fig. 12): Known 
only from Namibia, collection localities 
being north, northeast and east of Swa- 
kopmund in the Mopane Savanna, the 
Central Namib, and the Semi-desert and 
Savanna Transition of Giess (1971). 

Floral associations. — Boraginaceae (Helio- 
tropium tubulosum E. Mey. ex DC). 

Discussion. — At four localities the species 
has been found foraging on the flowers of 
Hcliotropium tubulosum in company with 
the masarine Jugurtia namibicola Gess 
which similarly appears restricted to this 
plant (Gess and Gess 2004: 39, Gess 2004: 
709). Celonites heliotropii is the only south- 
ern African Celonites known to forage on 
Hcliotropium, however, C. jousseaumei du 
Buysson has been recorded on flowers of 
this genus in the Sudan and in Cyprus. G. 
A. Mavromoustakis found C. cyprius de 
Saussure and C. rugiceps Bischoff to be 
confined to H. ?villosum Willd. and to H. 
europaeum L. respectively (Richards 1962: 
224). 

Celonites pnlcher Gess, new species 

Celonites sp. nov. F: Gess and Gess, 2003: 41 
(flower visiting). 

Diagnosis. — Both sexes: (7.2-7.9 mm); 
black or in some specimens with ground 
colour of pronotum, mesopleuron, scutel- 
lum, propodeum and gaster largely red- 
dish brown; head, pronotum, meso- 
pleuron, tegula, propodeum and gaster 
with yellowish-white markings. Clypeus 



and frons carinate, shiny, with small, well 
separated punctures and smooth interstic- 
es. Propodeum laterally with long, anteri- 
orly directed, narrow, sinuous slit; median 
part of propodeum postero-laterally mark- 
edly produced into lamella. Male genitalia 
as in Fig. 2. 

Description. — Female: Black. The follow- 
ing are yellowish-white: spot on upper 
half of clypeus between converging arms 
of M-shaped carina; streak margining inner 
orbit from immediately above end of 
frontal carina to level of lateral ocellus; 
spot on humeral angle and narrow band 
(widened medially) along hind margin of 
pronotum; elongate spot anteriorly on 
mesopleuron; tegula anteriorly and poster- 
iorly; in some specimens small spot medi- 
ally on scutellum and narrow streak on 
metanotum laterally; posterior two-thirds 
of lateral propodeal lamella; lateral and 
medial transverse markings posteriorly on 
terga I - V; medial round spot on tergum 
VI. 

Specimens from the north of the species' 
range differ markedly from those from the 
south in having the black largely replaced 
by reddish brown, all specimens, however, 
having the mandible and antenna light 
reddish-brown. 

In the southern, melanistic specimens 
the following are dark reddish brown: ill- 
defined area anterior to yellowish-white 
posterior band on pronotum, tegula medi- 
ally; extreme apex of scutellum, metano- 
tum medially; diffuse patches between 
pale markings posteriorly on terga I and 
II; tibiae and tarsi. 

In northern specimens the following are 
light reddish-brown: labrum; in some 
specimens diffuse area on clypeus sur- 
rounding pale spot, diffuse area on frons 
medially between arms of V-shaped carina 
and diffuse area on vertex behind eye; 
pronotum (other than for pale markings); 
in some specimens an ill-defined, poster- 
iorly directed, V-shaped marking medially 
and a lateral marking posteriorly on 
mesonotum; upper half of mesopleuron 



Volume 16, Number 1, 2007 



15 




Figs 1-6. Celonites species. Ventral view of male genitalia (x50). 1, C. heliotropii. 2, C. pulcher. 3, C. michaelseni. 
4, C. kalahariensis. 5, C. tumidiscutellatus. 6, C. arenarius. 



(other than for pale anterior spot); tegula clivity of tergum I and narrow black 

medially; entire scutellum and metanotum anterior transverse bands (usually hidden) 

(other than for pale markings sometimes on terga II - VI; femora, tibiae and tarsi of 

present); median part of propodeum en- all legs. 

tirely or partially (sometimes only poste- Wings browned in all specimens. 

rior lamella); metasoma (other than for Length 7.2-7.9 mm (average of 6: 

pale markings listed above), black de- 7.5 mm); length of front wing 5.3-6.0 mm 



16 



Journal of Hymenoptera Research 



(average of 6: 5.5 mm); hamuli 7-9. Length 
of extended tongue 5.6-5.8 mm; tongue 
length:body length = 0.74. 

Head 1.4X as wide as long (measured 
across eyes and from vertex to bottom of 
emargination of clypeus respectively). 
Clypeus and frons shiny, with small well 
separated punctures and smooth interstic- 
es; vertex dull, rugoso-punctate. Clypeus at 
mid-height with well-defined, smooth, 
widely and shallowly M-shaped carina 
and below it on each side with subtrans- 
verse, unpunctured, subcarinate swelling; 
surface of clypeal disc above M-shaped 
carina raised, especially laterally, below M- 
shaped carina (that is between it and 
subcarinate swelling) concave. Frons with 
conspicuous, smooth, widely and shallow- 
ly V- shaped carina (its arms somewhat 
sinuous) arising on each side opposite but 
outside middle of ocular sinus and meeting 
medially at obtuse angle at level of upper 
margin of antennal sockets; surface of frons 
falling very steeply from carina to antennal 
sockets and medially overhanging clypeal 
base. 

Pronotum, mesopleuron, mesoscutum, 
scutellum and mesodorsal part of propo- 
deum more coarsely sculptured than head, 
markedly longitudinally reticulate-punc- 
tate. Scutellum low, gently convex, gradu- 
ally rising from mesoscutum. Tegula un- 
usually long (circa 2.3 X as long as maxi- 
mum width), posteriorly hardly narrowed, 
evenly curved. Propodeum with declivity 
finely punctured; with postero-lateral 
flange of median part finely imbricate, 
lateral lamella on each side with a few 
large punctures. Lateral lamella of propo- 
deum at an angle to adjacent median part, 
with its outer margin gently convex, its 
inner margin emarginate in distal half, its 
apex rounded, separated from expanded 
postero-lateral flange of median part by 
narrow, sinuous, anteriorly directed slit 
(Fig. 22). 

Terga with punctures smaller than those 
on thorax and with interstices micro- 



sculptured yet moderately shiny; postero- 
lateral angles minimally produced; hind 
margins entire (non-crenulate). 

Male: Both males examined are dark and 
similar in coloration to the southern 
females. The yellowish-white facial mark- 
ings are slightly different from those of the 
female: that on the clypeus is larger, there 
is a marking on each arm of the frontal 
carina, and the marking near the eye fills 
the sinus rather than margining the upper 
inner orbit. 

Length 7.5-7.7 mm; length of front wing 
5.1-5.2 mm; hamuli 7 or 8. Length of 
extended tongue 5.3 mm. 

More gracile than female but structurally 
similar, differing most noticeably in the 
partial effacement of the clypeal carina, the 
reduction of the frontal carina, the more 
markedly swollen antennal club and the 
more pronounced postero-lateral angles of 
the terga. 

Genitalia in ventral view as in Fig. 2; in 
dorsal view with posterior margin of each 
paramere concave from rounded inner 
posterior angle to protruding, slightly 
incurved and pointed outer (lateral) poste- 
rior angle. 

Etymology. — The name pulcher, a Latin 
adjective meaning beautiful, refers to the 
strikingly colourful appearance of the 
species. 

Material examined. — Holotype: 9, NAMIBIA: 
57 km W of Keetmanshoop on road to Aus 
(26.46S 17.43E), 4.iii.2000 (F. W. and S. K. Gess) 
(visiting purple flowers of Anticharis scoparia (E. 
Mey. ex Benth.) Hiern ex Schinz, Scrophular- 
iaceae) [AMG]. Paratypes: NAMIBIA: Two 
Palms, near Palmwag (19.53S 13.54E), 
28.iii.2004, 1 9, 1 j 1 (visiting purple/violet 
flowers of Anticharis inflata Marloth & Engl., 
Scrophulariaceae); 120 km from Khorixas on 
road to Palm (20.17S 14.05E), 8.iv.l998, 6 99 
(visiting purple/violet flowers of Anticharis 
inflata); 57 km W of Keetmanshoop on road to 
Aus (26.46S 17.43E), 4.iii.2000, 1 9, 1 3 (visiting 
purple flowers of Anticharis scoparia (E. Mey. ex 
Benth.) Hiern ex Schinz) - (all F.W. and S. K. 
Gess) [AMG]. 



Volume 16, Number 1, 2007 



17 




Figs 7-11. Celonites species. Figs. 7-9. Ventral view of male genitalia (X50). 7, C. andrei. 8, C. capensis. 9, C. 
clypeatus. Figs. 10-11 Left fore leg of female. 10, C. heliotropii (x50). 11, C. michaelseni (x40). 



18 



Journal of Hymenoptera Research 



Geographic distribution. — (Fig. 13): Known 
from northern Namibia from two localities 
in the Mopane Savanna of Giess (1971) and 
from southern Namibia from one locality 
in the Dwarf Desert Savanna. 

Floral associations. — Scrophulariaceae: 
Aptosimeae (Anticharis spp.) 

Celonites kalahariensis Gess, new species 

Diagnosis. — Both sexes relatively small 
(6.3-8.3 mm); black, with pronotum, me- 
sopleuron, tegula, axilla, scutellum, meta- 
notum, propodeum and gaster largely 
light reddish brown (male generally and 
in part more melanistic); clypeus and frons 
with carinae (that of clypeus incomplete 
medially and in male less pronounced than 
in female); sculpture (particularly in fe- 
male) of frons below carina. and of raised 
disc of clypeus (particularly below carina 
of each side) markedly subcostulate-punc- 
tate with raised lineations on each side 
running obliquely ventro-medially. Meso- 
and metapleura with pronounced, postero- 
ventrally directed, apically rounded, pro- 
cesses (situated below base of lateral 
lamella of propodeum). Male genitalia as 
in Fig. 4. 

Description. — Female: Black. The follow- 
ing are yellowish white: in most specimens 
dorsal surface of propodeal lamella pos- 
tero-laterally; ill defined and diffuse 
patches postero-laterally on terga I - III. 
The following are light reddish brown: all 
but extreme base of mandible; in some 
specimens and to a varying extent labrum, 
distal margin of clypeus and supracarinal 
marking on same, supracarinal marking on 
frons, narrow streak behind eyes dorsally; 
proximal flagellomeres and underside of 
club; entire pronotum; extensive area of 
mesopleuron; tegula; small area postero- 
medially on mesoscutum (in some speci- 
mens expanded to cover most of mesoscu- 
tum, in others not present); axilla; entire 
scutellum; metanotum; most of dorsal 
surface and declivity of propodeum; most 
of terga; underside of front femur; apex of 



mid and hind femora; entire tibia and tarsi 
of all legs. Wings dark. 

Length 6.7-8.3 mm (average of 10: 
7.3 mm); length of front wing 4.3-5.3 mm 
(average of 10: 5.1 mm); length of extended 
tongue 4.4-5.8 mm (average of 3: 4.8 mm); 
hamuli 8 to 10 (most commonly 9). 

Head 1.3X as wide as long (measured 
across eyes and from vertex to bottom of 
emargination of clypeus respectively). 
Clypeus below each antennal socket 
raised and with an inwardly curved 
carina (carinae not produced medially 
and therefore not meeting each other). 
Frons with well-developed, widely and 
shallowly V-shaped carina arising on each 
side opposite but outside middle of ocular 
sinus and meeting medially at acute angle 
at level of upper margin of antennal 
sockets. Frons above carina and vertex 
coarsely reticulate-punctate; frons below 
carina and raised disc of clypeus (partic- 
ularly below carina of each side) marked- 
ly subcostulate-punctate with raised 
lineations on each side running obliquely 
ventro-medially. 

Pronotum, mesopleuron, mesoscutum, 
scutellum and propodeum mesodorsally 
less coarsely sculptured than vertex, longi- 
tudinally reticulate-punctate. Scutellum 
rising abruptly from mesoscutum, medial- 
ly somewhat swollen, slightly anteriorly 
produced and laterally subcarinate. Tegula 
circa 1.8 X as long as maximum width, 
posteriorly narrowed with outer margin 
curving inwards towards rounded but 
acute posterior angle. Meso- and meta- 
pleura with pronounced, postero-ventrally 
directed, apically rounded, processes (sit- 
uated below base of lateral lamella of 
propodeum). 

Propodeum with declivity subcostu- 
late and with lateral lamella of each side 
finely, closely and deeply punctured. 
Lateral lamella of propodeum broad, in 
same plane as adjacent median part, its 
outer margin gently curved and apex 
truncate, separated from median part by 



Volume 16, Number 1, 2007 



19 



narrow incurved slit (Fig. 24). Terga 
more finely and closely punctured than 
thorax; interstices microsculptured; pos- 
tero-lateral angles minimally produced; 
hind margins entire (that is non- 
crenulate). 

Male: Similar in coloration to female but 
generally more melanistic. Antennae en- 
tirely black; markings on head reduced, at 
most consisting of transverse band at base 
of clypeus. 

Length 6.3-7.6 mm (average of 6: 

6.7 mm); length of front wing 4.7-5.3 mm: 

4.8 mm); length of extended tongue 4.2- 
5.4 mm (average of 5: 4.7 mm); hamuli 8-9 
(most commonly 8). 

More gracile than female but structurally 
similar, apart from usual more markedly 
swollen antennal club, less pronounced 
clypeal carinae, and differently formed 
postero-lateral angles of the terga (more 
produced, upwardly bent, laterally curved 
and apically roundly acute). 

Genitalia in ventral view as in Fig. 4. 

Etymology. — The name, a Neolatin adjec- 
tive, is derived from Kalahari and is 
intended to indicate the provenance of 
the species. 

Material examined. — Holotype: 9, NAMIBIA: 
71 km E of Stampriet on road to Aranos (24.09S 
19.00E), 27.iii.2000 (F. W. and S. K. Gess) 
(visiting purple/violet flowers of Aptosimum 
procumbens (Lehm.) Steud., Scrophulariaceae) 
[AMG]. Paratypes: NAMIBIA: Gobabis [22.27S 
18.58E], v.1973 (R. Bayliss), 1 $ [AMG]; Onse 
Rust 192 (24.09S 18.02E), 17-18.V.1973 [C. F.] 
Jacot-Guillarmod), 1 9, 1 J [AMG]; 71 km E of 
Stampriet on road to Aranos (24.09S 19.00E), 
27.iii.2000, 1 9, 4 J6" (visiting purple/violet 
flowers of Aptosimum procumbens (Lehm.) 
Steud., Scrophulariaceae); same locality, 
28.iii.2000, 3 99 (1 9 visiting purple/violet 
flowers of Aptosimum procumbens; 1 9 on ground 
next to Aptosimum; 1 9 at hole in sand near 
Aptosimum); 24 km E of Stampriet on road to 
Aranos (24.14S 18.35E), l.iv.2000, 1 9 (visiting 
purple/violet flowers of Aptosimum procum- 
bens); 19 km E of Stampriet on road to Aranos 
(24.15S 18.33E), l.iv.2000, 3 99, 1 J (visiting 
purple/violet flowers of Aptosimum procum- 



bens); 2 km from C17 on R511 road to Mata 
Mata (25.37S 19.25E), 7.iii.2000, 1 9 (on ground 
near Aptosimum procumbens); same locality, 
8.iii.2000, 2 L 1 (on ground near Aptosimum 
procumbens) - (all F. W. and S. K. Gess) [all 
AMG]. SOUTH AFRICA: NORTHERN CAPE: 
123 km N of turnoff from NIO on R360 from 
Upington to Kgalagadi Park (27.30S 20.48E), 
5.iv.2000 (F. W. and S. K. Gess), l 9, (visiting 
purple/violet flowers of Aptosimum procumbens) 
[AMG]. 

Geographic distribution. — (Fig. 14): Known 
only from eastern and south-eastern Na- 
mibia in the Camelthorn Savanna (Central 
Kalahari) and the Mixed Tree and Shrub 
Savanna (Southern Kalahari) of Giess 
(1971) and from the trans-Orange part 
(Gordonia) of the Northern Cape of South 
Africa. All collecting sites were on the red 
dunes of the Kalahari. 

Floral associations. — Scrophulariaceae: 
Aptosimeae (Aptosimum procumbens). 

Discussion. — C. kalaharicnsis is similar to 
C. tumidiscutellatus but may be distin- 
guished by the following characters. In 
the female: the sculpture on the clypeus is 
more noticeably obliquely orientated to- 
wards the midline; the frons, midway 
between the V-shaped carina and the 
anterior ocellus is less noticeably raised; 
the scutellum is more abruptly and steeply 
raised anteriorly above the level of the 
mesoscutum but is less swollen medially. 
The distribution of reddish-brown is more 
extensive: notable are the frequent pres- 
ence of red facial markings; the extensive 
red area on the mesopleuron; the entirely 
red scutellum; the presence of red on the 
propodeum; the mostly red terga; and the 
more extensive red on the legs. 

In the male the produced postero-lateral 
angles of the terga are wider, more 
rounded and less sharply pointed apically, 
in consequence the hind margin of tergum 
VII is bi-sinuate rather than rounded. The 
most notable difference between the two 
species is in the male genitalia - compare 
those of C. kalaharicnsis (Fig. 4) with those 
of C. tumidiscutellatus (Fig. 5). 



Journal of Hymenoptera Research 
20 

Celonites arenarius Gess, new species Propodeal declivity finely longitudinally 

rugoso-punctate. Lateral lamella ot propo- 
Diagnosis— Both sexes relatively large deum broad, its surface with large punc- 
(7.6-10.0 mm long); black, with pronotum, tureg and ghiny i nte rstices, its outer margin 
tegula, terga dark reddish brown; clypeus mm i ma iiy curved and its apex truncate, 
and frons without carinae. Scutellum se p arate d from median part by an in- 
steeply raised anteriorly, markedly longi- curvec i s Ht. Terga coarsely and closely 
tudinally depressed medially. Hind mar- p unc t U red; postero-lateral angles pro- 
gins of terga markedly crenulate. Male duced, acutely pointed; hind margins 
genitalia as in Fig. 6. markedly crenulate. 

Description. — Female: Black. The follow- Male: Coloration identical to that of 

ing are dark reddish brown: distal half of f ema i e 

mandible; upper surface of pronotum; Length 7.6-8.8 mm (average of 4: 
tegula; narrow posterior margin of scutel- g ^ mm ). i en gth of front wing 5.4-6.6 mm 
lum; median part of metanotum; terga I ( ave rage of 4: 5.8 mm); hamuli 6 to 8. 
and II (except for triangular black antero- More gracile than the female but struc- 
medial areas); terga III and IV laterally and tura iiy similar, apart from the more mark- 
posteriorly; tergum V postero-medially; ec ^ y swo llen antennal club and the more 
"knees" of all legs; underside of mid and produced, very acutely pointed postero- 
hind tibiae and tarsi of these legs. Wings lateral angles of the terga. Hind margins of 
browned. terga even more markedly crenulate. 

Length 8.8-10.0 mm (average of 6: Genitalia in ventral view as in Fig. 6; 

9.4mm); length of front wing 6.6-6.8 mm p ar amere in dorsal view laterally subpar- 

(average of 4: 6.7 mm); length of extended a \\ e \ r terminally markedly concave be- 

tongue 5.4-5.8 mm (average of 2: 5.6 mm); tween produced, acutely pointed outer 

hamuli 8 or 9. angle and subrightangular inner angle. 

Head 1.3X as wide as long (measured Etymology. — The name arenarius, a Latin 

across eyes and from vertex to bottom of a djective relating to sand, serves to char- 

emargination of clypeus respectively), to- acterize the substrate on which the species 

tally devoid of clypeal and frontal carinae. was collected. 

Clypeal disc convex, steeply raised lateral- Material examined. — Holotype 9, SOUTH 

ly, with surface closely reticulate punctate AFRICA: NORTHERN CAPE: Richters- 

and with microsculptured interstices; V eld, Pachtvlei [circa 7 km NE of Alexan- 

frons, vertex and occiput slightly more der Bay] (28.33S 16.34E), 15.ix.1996 (F. W. 

coarsely reticulate punctate. Frons slightly and S. K. Gess) (nesting in burrow in sand) 

raised above and between antennal sockets [AMG]. Paratypes: SOUTH AFRICA: 

and very slightly depressed medially NORTHERN CAPE: Richtersveld, Pachtv- 

above swollen area. lei [circa 7 km NE of Alexander Bay] 

Pronotum, mesonotum, mesopleuron, (28.33S 16.34E), 15.ix.1996 (F. W. and S. K. 

scutellum, posterior two thirds of tegula Gess), 4 99, 4 $3 (all basking on sand); 

and dorso-lateral part of propodeum some- same locality and collectors, 18. ix. 1996, 1 9 

what more coarsely punctured than frons. (on sand) [AMG] 

Scutellum steeply raised above level of Geographic distribution. — (Fig. 15): Known 

postero-medially depressed mesoscutum, only from the type locality, Pachtvlei, 

medially markedly longitudinally de- situated on fine wind-blown alluvial sand 

pressed. Tegula long (circa 2.2 X as long on the southern bank of the Orange about 

as maximum width), posteriorly narrowed 7 km from its mouth. 

with outer margin curving inwards to- Floral associations. — Unknown, no flower 

wards rounded but acute posterior angle, visiting having been observed. Pollen from 



Volume 16, Number 1, 2007 



21 



the excavated cell was identified as possi- 
bly that of Lebeckia multiflorum E. Mey. 
(Fabaceae: Papilionoideae) growing in the 
vicinity of the nest. 

Celonites michaelseni von Schulthess 
Celonites Michaelseni von Schulthess, 1923: 137, 
male. Holotype: male, Namibia: Windhoek 
(Hamburg, destroyed in 1943 bombing). - 
Richards, 1962: 243 (speculation as to de- 
struction of type). 
Celonites michaelseni von Schulthess. Neotype: 
male, Namibia: Orjitundu River, 42 km W of 
Okahandja (21.54S 16.31E) (AMG). 
Celonites gariepensis Gess, 1997: 41, female, male. 
Holotype: female, South Africa: Richtersveld 
National Park (AMG). New synonym - Gess 
(S. K.) et al., 1997: 75 (flower visiting, 
nesting); Gess and Gess, 2003: 38 (flower 
visiting). 

Celonites michaelseni von Schulthess 1923 
was described from a single specimen 
collected by Dr W. Michaelsen of the 
Hamburger deutsch-sudwestafrikanischen 
Studienreise 1911 at Windhuk (now Wind- 
hoek in present day Namibia) during the 
period 29. iv. - 8. v. 1911. The holotype, 
a male, was deposited in the Zoologisches 
Museum, Hamburg). 

Richards (1962) omitted C. michaelseni 
from his revisional study, speculating that 
"the type was probably at Hamburg and 
may well have been destroyed". It is 
evident that he did not have access to any 
Celonites material from Namibia and there- 
fore did not see any specimens answering 
to Schulthess' description. 

Recent collecting in Namibia, in particu- 
lar that of F. W. and S. K. Gess during the 
period 1997-2004, has yielded a wealth of 
material of a common and widespread 
species which without any doubt is con- 
specific with C. michaelseni. In common 
with some other species of wide distribu- 
tion, a considerable variation is shown in 
the colour pattern; however, structural 
features, most importantly including the 
male genitalia, are constant across the 
range. The colour pattern exhibited by 
specimens from the north-central and 



central part of Namibia, is that described 
by von Schulthess for the type from 
Windhoek. 

Confirmation was received in 2005 from 
Dr Rudolf Abraham of the Zoologisches 
Museum, Hamburg that "we cannot find 
the type of Celonites michaelseni in our 
collection, so it is indeed destroyed in 
1943 during WW2". 

In view of the desirability of clarifying 
the taxonomic status of what appears to be 
the most common Celonites species occur- 
ring in Namibia, it is appropriate to 
designate a neotype for C. michaelseni. The 
specimen chosen for this purpose is a male 
from a series of 5 99 and 3 6*6" from 
Orjitundu River, 42 km W of Okahandja 
(21.54S 16.31E), circa 90 km NW of Wind- 
hoek. The colour pattern is consistent with 
that of the destroyed holotype. 

Celonites gariepensis Gess 1997 is a syno- 
nym. The name was applied to specimens 
from the southern part of the species' 
range, their true identity not being recog- 
nized at the time. 

As already stated, C. michaelseni shows 
a remarkable degree of colour variation 
across its range. 

The most strikingly colored specimens, 
characterized by very clearly defined white 
lateral markings and orange-red postero- 
medial markings on the otherwise black 
terga, a white-marked black pronotum and 
an unmarked black mesopleuron occur in 
the north central part of Namibia (near 
Tsumeb; the Etosha National Park; be- 
tween Outjo and Okaukuejo; between 
Omaruru and Kalkveld; between Omaruru 
and Karibib; and between Karibib and 
Okahandja. 

Moving westwards there is a tendency at 
least for some specimens to have the terga 
and pronotum red rather than black but 
having the same markings and to have 
a white-marked black mesopleuron (26 km 
W of Kamanjab; 24 km N of Palmwag; 
120 km from Khorixas on road to Palm; 
40 km E of Springbokwater). 



22 



Journal of Hymenoptera Research 



Further west still and in the Central 
Namib most specimens have the terga with 
red posterior bands which are unmarked 
or at most have white postero-medial 
markings; the pronotum and mesopleuron 
being generally black and unmarked 
(40 km E of Springbokwater; between Uis 
and Henties Bay; Solitaire; between Usakos 
and Swakopmund; Swakopmund District). 

In south central and south-eastern Na- 
mibia specimens are similar to the last 
form but the red posterior bands have 
diffuse white markings both laterally and 
postero-medially and in some the prono- 
tum is white marked (S. of Windhoek, 
Gaub bridge, E. of Hardap Dam; S. of 
Mariental, near Karasburg; between Karas- 
burg and Ariamsvlei). 

At the southern extremity of the species" 
range, that is in the Richtersveld (Northern 
Cape, South Africa) specimens are gener- 
ally melanistic and are also somewhat 
smaller than those from more northern 
localities. 

Specimens from two isolated localities in 
Limpopo (= Northern Province), South 
Africa, are similar to those from the 
savanna in northern Namibia. 

Male genitalia in ventral view as in 
Fig. 3. 

Material examined. — Neotype: $, NAMIBIA: 
Orjitundu River, 42 km W of Okahandja (21.54S 
16.31E), 1 & 2.iv.2004 (F. W. and S. K. Gess) 
(visiting purple/violet flowers of Aptosimum 
arenarium Engl., Scrophulariaceae) [AMG]. Oth- 
er specimens: NAMIBIA: 26 km W of Kaman- 
jab (19.36S 14.28E), 7.iv.l998, 7 99, 2 33 (visiting 
purple/violet flowers of Aptosimum angustifo- 
liuiu Weber & Schinz, Scrophulariaceae); 24 km 
N of Palmwag (19.43S 13.51E), 18.iii.1999, 3 99 (2 
visiting purple/violet flowers of Anticharis 
inflata Marloth & Engl., Scrophulariaceae; 1 
visiting blue/violet flowers of Aptosimum angu- 
stifolium); 27 km NW of Outjo on road to 
Okaukuejo (19.44S 15.53E), 26.iii.1997, 1 <J 
(visiting purple flowers of Aptosimum decumbent 
Schinz); Two Palms, near Palmwag (19.53S 
13.54E), 28.iii.2004, 1 <J (visiting white flowers 
of Heliotropium tubulosum E. Mey. ex DC, 
Boraginaceae); 120 km from Khorixas on road 



to Palm (20.17S 14.05E), 8.iv.l998, 1 9, 3 <$<$ (1 9 
and 1 o* visiting purple/violet flowers of Antic- 
haris inflata; 2 33 visiting white flowers of 
Boerhavia deserticola Codd, Nyctaginaceae); 
40 km E of Springbokwater (20.17S 13.57E), 
ll.iv.2002, 2 99 (visiting violet flowers of Antic- 
haris inflata); Uis to Khorixas (20.54S 15.05E), 
15.iii.2004, 1 9 (visiting purple-violet flowers of 
Anticharis); 24 km N of Omaruru on road to 
Kalkveld (21.15S 16.01E), 23.iii.1997, 29 99 (28 99 
visiting purple flowers of Aptosimum arenarium 
Engl.); 156 km from Khorixas, betw. Uis and 
Henties Bay (21.24S 14.46E), 9.iv.l998, 2 99, 1 3 
(visiting purple/violet flowers of Anticharis 
ebracteata Schinz); 20 km S of Omaruru on road 
to Karibib (21.35S 15.59E), 23 and 24.iii.1997, 5 
99, 1 3 (3 99 visiting purple flowers of 
Aptosimum arenarium; 2 99/ 1 3 on ground next 
to this plant); 30 km S of Omaruru on road to 
Karibib (21.41S 15.59E), 26.iv.2002, 5 9?, 3 33 
(visiting bluish violet flowers of Aptosimum 
arenarium); Karibib to Omaruru (21.51S 
15.55E), 12.iii.2004, 1 3; Orjitundu River, 42 km 
W of Okahandja (21.54S 16.31E), 1 & 2.iv.2004, 5 
99, 2 33 (visiting purple/violet flowers of 
Aptosimum arenarium); 72 km E of Karibib on 
road to Okahandja (21.54S 16.31E), l.iv.1998, 1 9, 

2 33 (visiting purple flowers of Aptosimum 
arenarium); 94 km E of Karibib on road to 
Okahandja (21.57S 16.43E), l.iv.1998, 1 9 (visit- 
ing purple/violet flowers of Aptosimum arenar- 
ium); 77 km E of Henties Bay on road to Klein 
Spitzkuppe (21.54S 14.58E), 19.iv.2002, 1 9 
(visiting white flowers of Heliotropium tubulo- 
sum); 58 km SW of Usakos on road to Swakop- 
mund (22.12S 15.10E), 23.iv.2002, 2 99 (1 9 
visiting violet flowers of Aptosimum spinescens 
(Thunb.) Weber; 1 9 on ground next to this 
plant); 7 km from Gaub bridge towards Kuiseb 
River (23.27S 15.48E), 14.iv.1998, 7 99 (visiting 
purple/violet flowers of Aptosimum lineare 
Marloth & Engl.); Solitaire (23.52S 16.00E), 
30.iv.2002), 3 99, 2 33 (visiting purple/violet 
flowers of Aptosimum spinescens); E of Hardap 
Dam (24.29S 17.53E), 4.iv.l997, 1 9, 1 3 (visiting 
purple flowers of Aptosimum glandulosum Weber 
& Schinz); 25 km S of Mariental (24.50S 17.56E), 
16. iv. 1998, 2 99 (visiting purple/violet flowers 
of Aptosimum spinescens); S of Maltahohe on 
D811 (25.16S 17.03E), 23.iii.1999, 1 3 (visiting 
purple/violet flowers of Aptosimum spinescens); 
Klein-Aus-Vista (26.41S 16.13E), 23.ix.2003, 4 99, 

3 ; ; (visiting purplish violet flowers of Aptosi- 



Volume 16, Number 1, 2007 



23 



mum tragacanthoid.es E. Mey. ex Benth.); circa 
3 km NNW of Karasburg (27.58S 18.43E), 
6.iii.l999, 1 9 (visiting purple flowers of Aptosi- 
mum spinescens); between Karasburg and Ar- 
iamsvlei (28.05S 19.25E), 18.iv.1998, 1 9 (ex nest) 
- (all F. W. and S. K. Gess) [all AMG]; 10 km SE 
of Tsumeb [19.17S 17.48E], 8.iii.l990, 1 J; 62 km 
E of Karibib [21.53S 16.25E], 20.ii.1990, 1 9, 4 $& 
17 km W of Okahandja [21.57S 16.44E], 
19.ii.1990, 1 J; 28 km S of Windhoek [22.49S 
17.08E], 17.ii.1990, 2 33 - (all W. J. Pulawski) [all 
CAS]; Etosha Nat.Park, Nau-Obes (19.19S 
16.37E), 17.L1987 (J. Irish, E. Marais) 2 99; 
Damaraland, 6 km N Arandis (22.22S 14.59E), 
l.iii.- 9.iv.l985, 1 9, 9.iv.- 6.V.1985, 1 9 (both J. 
Irish, H. Rust); Swakopmund Dist, Rossing 
Mine (22.28S 15.02E), 8.v.- 5.vi.l984 (J. Irish, H. 
Liessner), 1 9; Swakopmund Dist., Lower 
Ostrich Gorge (22.30S 14.58E), 5.vi.- 3.vii.l984 
(J. Irish, H. Liessner), 1 9; Hoogland 132, 
Maltahohe (SE2416Cd) [24.45S 16.15E], 1- 
3.ii.l974, 1 9; Vogelstrausskluft 87, Bethanien 
(SE2717Ba) [27.00S 17.31E], 24-29. ix. 1974 (? 
collector), 1 3 - [all NNIC]. SOUTH AFRICA: 
LIMPOPO (formerly NORTHERN PROVINCE): 
Louis Trichardt (2239BB sic!!) [23.03S 29.55E], 
27.vi.1991 (M. Lorenz), 1 9; Rust de Winter Dam 
[25.15S 28.29E], 12.X.1980 (F. J. Herbst), 1 9, 1 3 
[all AMG]; NORTHERN CAPE: Namaqualand, 
Richtersveld National Park, Koeroegabvlakte 
(28.11S 17.03E), 17-21 and 24.ix.1995, holotype 
9, 33 paratype 99, 4 paratype SS of C. gariepensis 
Gess (33 99 incl.holotype, 3 jj 1 in deep violet 
flowers of Peliostomum sp., Scrophulariaceae; 1 
9, 1 3 in purple-violet flowers of Aptosimum 
spinescens); same locality, 6.ix.l996, 1 paratype 9, 
7 paratype 3 3 of C. gariepensis (9, 6 jj on 
ground near flowering Peliostomum sp.); Nama- 
qualand, Richtersveld National Park, Paradise 
Kloof (28.19S 17.01E), 22.ix.1995, 1 paratype 9 of 
C. gariepensis (associated with nest); Namaqua- 
land, Richtersveld National Park, 1.5 km from 
Helskloof gate (28.18S 16.57E), 8 and 9.ix.l996, 3 
paratype 99, 1 paratype o* of C. gariepensis (all on 
ground near Aptosimum spinescens) - (all F. W., 
S. K. and R. W. Gess) [all AMG]. 

Geographic distribution. — (Fig. 16): Known 
in Namibia from collection localities span- 
ning ten degrees of latitude in the Mopane 
Savanna, Saline Desert with Dwarf Scrub 
Savanna Fringe, Mountain Savanna and 
Karstveld, Thornbush Savanna, Central 
Namib, Semi-desert and Savanna Transi- 



tion, Highland Savanna, and Dwarf Scrub 
Savanna of Giess (1971). In South Africa it 
is known from the Richterveld National 
Park (Northern Cape) in the Namaqualand 
Broken Veld of Acocks (1953) and from 
two localities in Limpopo in the Mixed 
Bushveld of Acocks. 

Floral associations. — Scrophulariaceae: 
Aptosimeae (Anticharis spp., Aptosimum 
spp., Peliostomum sp.). 

Known species of Celonites not 
previously recorded from Namibia. 

Celonites andrei Brauns 

Celonites andrei Brauns, 1905: 228, female. 
Holotype: female, South Africa : Willowmore 
(TMP). - Brauns, 1913: 206, male, female, nest; 
Richards, 1962: 236 (revision); Gess, 1996: 222 
(flower visiting); Gess and Gess, 2003: 36 
(flower visiting). 

Male genitalia (Fig. 7) 

Material examined. — NAMIBIA: Sperrgebiet 
(Diamond Area 1): Tsaukhaib (26.43S 15.40E), 
13 and 15.ix.2005 (F. W. and S. K. Gess), 1 9, 2 
33 (both sexes associated with Aptosimum 
spinescens (Thunb.) Weber, Scrophulariaceae, 
flying around plants, alighting on ground next 
to them, 9 observed entering a flower). 

Geographic distribution. — (Fig. 17): Known 
in Namibia from a single locality on the 
old wagon track from Luderitz to Aus, in 
the north of Diamond Area 1 in the Desert 
and Succulent Steppe (Winter rainfall area) 
of Giess (1971). It is widely distributed in 
the Karoo Biome of South Africa, speci- 
mens in the Albany Museum collection 
being from the Northern Cape (Twee 
Rivieren in the Kalahari Gemsbok National 
Park, the Richtersveld National Park, Ane- 
nous, Springbok, and near Norvalspont), 
from the Western Cape (near Prince Al- 
bert) and from the Eastern Cape (Steytler- 
ville). 

In South Africa, as in Namibia, the 
species has been recorded visiting flowers 
solely of Scrophulariaceae: Aptosimeae 
(Aptosimum procumbens (Lehm.) Steud., A. 



24 



Journal of Hymenoptera Research 









Uy* 


-> \< 


V 




• 










.25' 




^ 


_y V 




'A - 


■31 






or 




.11 - 


12 


15' ^ 




jT 


250 


500 




Kilometres 

IS 






f 

ly-V V. S 

\ • 

\ *** 
\ • 

\ 


^ 


V 

>^ 250 500 


14 V 

15 *" 




>■——•**' Kilometres 

15 







or 

X 250 500 


15 V 




^»— —^ Kilo me Ires 

35 








}\. < 


1 • 

\ • 

\ • 




% / 250 500 


17 V 




C — ** Kilometres 



Figs 12-17. Celonites species. Distributions. 12, C. heliotropii. 13, C. pulcher. 14, C. kalahariensis. 15, C. arenarius. 
16, C. michaelseni. 17, C. andrei. 

spinescens (Thunb.) Weber, and Peliostomum Brauns, 1913: 205, male; Richards: 1962: 235 
vitgatum E. Mey. ex Benth.) (revision); Gess, 1996: 223 (flower visiting); Gess 

and Gess, 2003: 37 (flower visiting). 

Celonites capensis Brauns Male genitalia (Fig. 8) 

Celonites capensis Brauns, 1905: 231, female. Material examined.— NAMIBIA: 16 km S of 

Holotype: South Africa: Willowmore (TMP). - Rosh Pinah (28.04S 16.51E), 13-15.X.2000, 24 99 



Volume 16, Number 1, 2007 



25 








■ / \ r i 




iU-^tV 


N • 
\ • 

v. • . 

\ • 

\ •. 




• O/ 


19 V 




s—*^ 



Figs 18-20. Celonites species. Distributions. 18, C. capensis. 19, C. clypeatus. 20, C. tumidiscutellatus. 



(14 99 visiting yellow flowers of Tripteris 
microcarpa Harv., Asteraceae; 3 99 visiting 
yellow flowers of Gazania lichtensteinii Less. , 
Asteraceae; 3 99 visiting yellow flowers of 
Didelta carnosa (LA.) Ait., Asteraceae; 4 99 
visiting pinkish white flowers of Aizoaceae: 
Mesembryanthema) - (all F. W. and S. K. Gess) 
[all AMG]; Diamond Area 1, Daberas (28.12S 
16.49E), 14-29.ix.1994 (E. Marais) 1 9 (Pres. pitf. 
traps) [NNIC]. 

Geographic distribution. — (Fig. 18): Known 
in Namibia from two closely adjacent 
localities in the extreme south of the Desert 
and Succulent Steppe (Winter rainfall area) 
of Giess (1971). It is widely distributed in the 
Karoo Biome of South Africa, specimens, 
mostly in the Albany Museum collection, 
being from the Northern Cape (Richterveld, 
between Vioolsdrif and Springbok, Voelklip 
near Springbok, and Sors Sors near Kamie- 



skroon), from the Western Cape (Nuwerus, 
near Ceres, Malmesbury, Ladismith, and 
near Oudtshoorn) and from the Eastern 
Cape (several localities near Grahamstown). 
Richards (1962) recorded the species from 
Worcester, Montagu and Matjesfontein in 
the Western Cape and from Willowmore 
and Somerset East in the Eastern Cape. 

In South Africa the species has been 
recorded visiting the flowers of a wide 
range of plants: Asteraceae (Berkheya spp., 
inch Berkheya heterophylla (Thunb.) O. 
Hoffm., and Senecio pterophorus DC), Ai- 
zoaceae: Mesembryanthema (Drosanthe- 
iiiinii sp. and Prenia pollens (Ait.) N. E. 
Br.), Geraniaceae (Pelargonium myrrhifolium 
(L.) L'Herit.), Scrophulariaceae (Phyllopo- 
dium cuneifolium (L. f.) Benth.), Campanu- 
laceae (Wahleribergia ecklonii Buek), Irida- 



26 



Journal of Hymenoptera Research 




21 






Figs 21-24. Celonites species. Diagrammatic representations of postero-lateral part of propodeum. 21, C. 
capensis. 22, C. pulcher. 23, C. michaelseni. 24, C. kalahariensis. 



ceae (Fermria sp.), and Boraginaceae (Ehre- 
tia rigida (Thunb.) Druce). 

Celonites clypeatus Brauns 

Celonites clypeatus Brauns, 1913: 206, female. 
Holotype: South Africa: Willowmore (TMP). - 
Richards, 1962: 236 (revision); Gess, 1996: 223 
(flower visiting); Gess and Gess, 2003: 38 
(flower visiting). 

Male genitalia (Fig. 9) 

Material examined. — NAMIBIA: 62 km E of 
Karibib [21.53S 16.25E], 20.ii.1990 (W. J. Pu- 
lawski), 3 99 [CAS]; 10 km W of Usakos on road 
to Swakopmund (21.59S 15.30E), 12 & 



13.iii.2004, (F. W. and S. K. Gess), 2 99, 3 ^ 
(visiting purple violet flowers of Aptosimum 
arenarium Engl., Scrophulariaceae) [AMG]; Ex- 
celsior 206, Windhoek (22.27S 17.38E), 1- 
31.viii.1979 (S. Louw, M.-L. Penrith), 1 <? 
[NNIC]; 25 km S of Mariental (24.50S 17.56E), 
16.iv. 1998, 3 99, 1 o (visiting purple/violet 
flowers of Aptosimum spinescens (Thunb.) We- 
ber); 2 km from CI 7 on R511 road to Mata Mata 
(25.37S 19.25E), 8.iii.2000, 1 9; circa 3 km NNW 
of Karasburg (27.58S 18.43E), 6.iii.l999, 2 99 
(visiting purple flowers of Aptosimum spines- 
cens); between Karasburg and Ariamsvlei 
(28.05S 19.25E), 18.iv.1998, 1 9 (visiting pur- 
ple/violet flowers of Aptosimum spinescens) - (all 
F. W. and S. K. Gess) [all AMG]; 



Volume 16, Number 1, 2007 



27 



Geographic distribution. — (Fig. 19): Known 
in Namibia from collection localities span- 
ning seven degrees of latitude in the 
southern half of the country in the Thorn- 
bush Savanna, the Highland Savanna, and 
the Dwarf Scrub Savanna of Giess (1971). It 
is widely distributed in the Karoo Biome of 
South Africa, specimens in the Albany 
Museum collection being from the Northern 
Cape (Twee Rivieren in the Kalahari Gems- 
bok National Park, the Richtersveld Nation- 
al Park, Springbok, Leliefontein and near 
Norvalspont), from the Western Cape (near 
Prince Albert) and from the Eastern Cape 
(between Cradock and Hofmeyr, Grahams- 
town, Steytlerville and Willowmore). 

In South Africa, as in Namibia, the 
species has been recorded visiting the 
flowers solely of Scrophulariaceae: Aptosi- 
meae (Aptosimum indivisum Burch., A. 
procumbens (Lehm.) Steud., A. spinescens 
(Thunb.) Weber, and Peliostomum virgatum 
E. Mey. ex Benth.) 

Celonites tumidiscutellatus Gess 

Celoiiites tumidiscutellatus Gess, 1997: 44, fe- 
male, male. Holotype: female, South Africa, 
[Northern Cape Province,] Namaqualand, 
Springbok, Hester Malan [now Goegap] Nature 
Reserve (AMG). - S. K. Gess et al, 1997: 76 



(flower visiting); Gess and Gess, 2003: 40 
(flower visiting). 

Male genitalia (Fig. 5) 

Material examined. — NAMIBIA: E of Oran- 
jemund, 37 km from checkpoint on road to 
Sendelingsdrif (28.23S 16.44E), 24.ix.1997, 4 99 
(visiting purple/violet flowers of Aptosimum 
spinescens (Thunb.) Weber, Scrophulariaceae); E 
of Oranjemund, 28 km from checkpoint on road 
to Sendelingsdrif (28.26S 16.42E), 25.ix.1997, 2 
99, 1 j (visiting violet flowers of Peliostomum 
leucorrhizum E. Mey. ex Benth., Scrophularia- 
ceae) - (all F. W. and S. K. Gess) [all AMG]. 

Geographic distribution. — (Fig. 20): Known 
in Namibia from the northern bank of the 
Orange in the extreme south of the Desert 
and Succulent Steppe (Winter rainfall area) 
of Giess (1971). In South Africa it is known 
from the Northern Cape (Richtersveld Na- 
tional Park , near Springbok and Leliefon- 
tein) and from the Eastern Cape (Willow- 
more). 

In South Africa, as in Namibia, the 
species has been recorded visiting the 
flowers solely of Scrophulariaceae: Aptosi- 
meae (Aptosimum indivisum Burch., A. 
procumbens (Lehm.) Steud., A. spinescens 
(Thunb.) Weber, and Peliostomum virgatum 
E. Mey. ex Benth.) 



KEY TO SPECIES OCCURRING IN NAMIBIA 



Propodeum postero-laterally with long, anteriorly directed slit separating off lateral 
lamella from medial part of propodeum; slit straight or sinuous, not or only 
minimally incurved anteriorly (Figs 21, 22) 2 

Propodeum postero-laterally with anteriorly directed slit which after very short 
distance is incurved and ends in circular emargination or extends medially in 
transverse direction, in both conditions cutting off lateral lamella leaving part of 
hind margin of median part of propodeum as narrow finger-like process pointing 
towards end of lateral lamella (Figs 23, 24) 3 

Propodeal slit narrow and straight (Fig. 21); lateral lamella distally broadly truncate, 
more or less in same plane as adjacent median part of propodeum. Neither clypeus 
nor frons with carina (though frons may have low swelling). Male genitalia: 
Fig. 8 capensis Brauns 

Propodeal slit narrow and sinuous (Fig. 22); lateral lamella distally narrowly and 
obliquely truncate, at angle to adjacent median part of propodeum; median part of 
propodeum postero-laterally markedly produced, lamellate (Fig. 22). Both clypeus 
and frons with carina (though clypeal carina may be weak or absent in male). 



2 g Journal of Hymenoptera Research 

Clypeus and frons shiny, with small, well separated punctures and smooth 
interstices; mesoscutum and scutellum markedly longitudinally reticulate-punc- 
tate. Male genitalia: Fig. 2 pulcher Gess n. sp. 

3. Frons with V-shaped carina (sometimes weak medially); clypeus with an M-shaped 

carina (if medially weak and diffuse, at least well developed laterally) 4 

Frons without a V-shaped carina 8 

4. Meso- and metapleura with pronounced, postero-ventrally directed, apically rounded, 

processes (situated below base of lateral lamella of propodeum). Pronotum, 
mesopleuron, tegula, axilla, scutellum, metanotum, propodeum and gaster 
largely red. In some specimens (particularly females) clypeus baso-medially 
and frons supra-carinally with transverse red markings. Male genitalia: 

Fig. 4 kalahariensis Gess n. sp. 

Meso- and metapleura without such processes 5 

5. Mesopleuron with variously sized (to minute) red marking 6 

Mesopleuron either totally black or with white marking 7 

6. Frons with small red (or yellow) spot on each side [in female situated next to upper 

margin of ocular sinus, that is, above carina; in male situated within lower half of 
ocular sinus, that is, below carina]. Clypeus of female occasionally with red spot, 
that of male usually with red or yellow spot. Carinae on frons and clypeus of 
female poorly developed medially, those of male even more poorly developed and 
almost effaced respectively. Scutellum falling abruptly onto mesoscutum (espe- 
cially in female). Male genitalia: Fig. 7 andrei Brauns 

Frons without small red spot on each side in female but in male occasionally with red 
spot within lower half of ocular sinus. Clypeus of both sexes immaculate. Carinae 
on frons and clypeus of female well developed throughout, both carinae less 
developed but indicated in male. Scutellum falling gradually onto mesoscutum. 
Male genitalia: Fig. 9 clypeatus Brauns 

7. Frons, clypeus and mesopleuron totally black, dorsal part of pronotum red, abdominal 

terga black with reddish posterior bands; scutellum medially raised (subcorneal) 
and anteriorly falling gradually onto mesoscutum (see Gess 1997: Fig. 9). Male 

genitalia: Fig. 5 tumidiscutellatus Gess 

Frons, clypeus and mesopleuron often with white spots, dorsal part of pronotum black 
with white humeral spots and white or yellow (to orange) hind margin, abdominal 
terga with reddish posterior bands and white lateral and medial spots; scutellum 
antero-medially almost overhanging mesoscutum and falling very abruptly onto it 
(see Gess 1997: Fig. 3). Male genitalia: Fig. 3 (See also Gess 1997: Figs 5 and 
6) michaelseni von Schulthess 

8. Fore femur (more particularly that of female) produced postero-ventrally in proximal 

half (Fig. 10). Pronotum, tegula, scutellum and posterior bands on terga red. Male 
with most of mandibles, entire labrum and clypeus and variably developed facial 

markings lemon yellow. Male genitalia: Fig. 1 heliotropii Gess n. sp. 

Fore femur unmodified. Hind margins of terga very markedly crenulate. Scutellum 
steeply raised anteriorly, markedly longitudinally depressed medially. Pronotum, 
tegula and terga (largely) red. Pale facial markings absent. Male genitalia: 
Pig- 6 arenarius Gess n. sp 



ACKNOWLEDGMENTS of happy, productive and synergistic fieldwork, for 

valuable discussion and encouragement; the Nami- 

Thartks are expressed to the following for much bian Ministry of Environment and Tourism for 

appreciated assistance as specified: Sarah Gess of the granting permits to conduct research and collect 

Albany Museum, Grahamstown, co-collector of most biological specimens in that country; NAMDEB (Pty) 

of the Albany Museum's material, for over thirty vears Ltd for granting permits to enter the Sperrgebiet 



Volume 16, Number 1, 2007 



29 



(Diamond Area No 1); Coleen Mannheimer of the 
National Herbarium of Namibia, Windhoek for her 
determination of voucher specimens of Namibian 
plants; Eugene Marais of the Namibian National 
Insect Collection, Windhoek, Connal Eardley of the 
National Collection of Insects, Pretoria and Wojciech 
Pulawski of the California Academy of Sciences, San 
Francisco for the loan of specimens from their 
respective collections; Rudolf Abraham of the Zoolo- 
gisches Museum Hamburg for confirming that the 
type of Celonites michaelseni was destroyed in 1943 
during WW2; Fred Herbst of Durban (formerly of 
Gauteng) for the gift of two specimens of C. 
michaelseni from the Limpopo Province; Shirley 
Pinchuck of the Electron Microscopy Unit and 
Bronwyn McLean of the Graphics Services Unit, both 
of Rhodes University, Grahamstrown, for help with 
the SEMs and the distribution maps repectively; 
Ashley Kirk-Spriggs of the Albany Museum, Gra- 
hamstown for cleaning and enhancing the images 
used for figures 1-11; the South African Foundation 
for Research Development (FRD) and the South 
African National Research Foundation (NRF) for 
running expenses grants awarded either to the author 
and to Sarah Gess or to Sarah Gess and the author for 
field work during the course of which much of the 
present material was collected; the reviewers Wojciech 
Pulawski and James Carpenter for their useful 
comments. 

LITERATURE CITED 

Acocks, J. P. H. 1952. Veld types of South Africa. 
Memoirs of the Botanical Survey of South Africa 29: 
i-iv, 1-92. 

Bartholomew, J., P. J. M. Geelen, H. A. G. Lewis, P. 
Middleton, and B. Winkleman (Editorial direc- 
tion), 1981. The Times Atlas of the World (compre- 
hensive Edition). London: Times Books Limited. 

Brauns, H. 1905. Masaridae von Sudafrika. Annates 
Musei Nationalis Hungarici 3: 219-234. 

. 1913. Drifter Beitrag zur Kenntnis der Masar- 

iden (Hym.) von Sudafrika. Entomologische Mittei- 
lungen 2: 193-209. 

Gess, F. W. 1989. New species of the genus Celonites 
Latreille (Hymenoptera: Masaridae) from South 
Africa. Annals of the Cape Provincial Museums 
(Natural History) 18 (4): 83-94. 



. 1992. A new species of the genus Celonites 

Latreille associated with the flowers of Wahlen- 
bergia in southern Africa. Journal of Hymenoptera 
Research 1: 141-144. 

. 1997. Contributions to the knowledge of 

Ceramius Latreille, Celonites Latreille, fugurtia 
Saussure and Masarina Richards (Hymenoptera: 
Vespidae: Masarinae) in South Africa, journal of 
Hymenoptera Research 6: 36-74. 

. 2004. A revision of the Afrotropical species of 

the wasp genus fugurtia de Saussure, 1854 
(Hymenoptera: Vespidae: Masarinae). Journal of 
the Kansas Entomological Society 77 (4): 669-720. 

Gess, S. K. 1996. The Pollen Wasps: Ecology and Natural 
History of the Masarinae. Harvard University 
Press, Cambridge, Massachusetts. 

, F. W. Gess, and R. W. Gess. 1997. Update on 

the flower associations of southern African 
Masarinae with notes on the nesting of Masarina 
strucki Gess and Celonites gariepensis Gess (Hyme- 
noptera: Vespidae: Masarinae). journal of Hyme- 
noptera Research 6: 75-91. 

and F. W. Gess. 2003. A catalogue of flower 

visiting records for aculeate wasps and bees in the 
semi-arid to arid areas of southern Africa. Grahams- 
town: Albany Museum. 

and F. W. Gess. 2004a. Distributions of flower 

associations of pollen wasps (Vespidae: Masar- 
inae) in southern Africa, journal of Arid Environ- 
ments 57: 17-44. 

and F. W. Gess. 2004b. The distributions of the 



genera of pollen wasps (Hymenoptera: Vespidae: 
Masarinae) in the semi-arid to arid areas of 
southern Africa in relation to their requirements 
for successful nesting. Transactions of the Royal 
Society of South Africa 59 (2): 59-64. 

Giess, W. 1971. A preliminary vegetation map of 
South West Africa. Dinteria 4: 1-114. 

Leistner, O. A. and J. W. Morris. 1976. Southern 
African Place Names. Annals of the Cape Provincial 
Museums (Natural History) 12: i-iv, 1-565. 

Richards, O. W. 1962. A Revisional Study of the Masarid 
Wasps (Hymenoptera, Masaridae). London: British 
Museum (Natural History). 

Schulthess, A. von. 1923. Hymenoptera VI: Vespidae. 
Pp. 135-140 in: Michaelsen, W., ed. Beitrdge zur 
Kenntnis der Land- und Susswasserfauna Deutsch- 
Siidwestafrikas 2 Lief 2. Hamburg, L. Friederichsen 
and Co. 



J. HYM. RES. 

Vol. 16(1), 2007, pp. 30-50 

Notes on Nesting and Flower Visiting of some Anthidiine Bees 
(Hymenoptera: Megachilidae: Megachilinae: Anthidiini) in 

Southern Africa 

Sarah K. Gess* and Friedrich W. Gess 

Department of Entomology, Makana Biodiversity Centre, Albany Museum, Grahamstown, 6139 

South Africa; email: s.gess@ru.ac.za; email: f.gess@ru.ac.za 
* Address for correspondence: Sarah Gess, Albany Museum, Grahamstown, 6139 South Africa; 

email: s.gess@ru.ac.za; tel 0466222312 



Abstract.— For Anthidiini occurring in southern Africa, descriptive notes on nests of seven 
species belonging to three genera, have been published. All are constructed from plant fibres and, 
depending upon the species, are situated aerially on plants or in pre-existing cavities. To these are 
added first descriptions of nests of three further species representing three genera. Serapista ruftpes 
(Friese), like the only other species of Serapista, S. denticulata (Smith), for which nesting is known, 
was found to construct nests from plant fibres, however, although similarly found aerially on plant 
stems, a nest with its builder was discovered in a burrow in the ground. Afranthidium 
(Nigranthidium) concolor (Friese) was found nesting in a burrow in the ground, like other species 
of Afranthidium, using plant fibres. Plesianthidium, represented by P. (Spinanthidiellum) volkmanni 
(Friese), was found constructing groups of separate resinous, spouted, pot-like cells, similar to those 
constructed by some extraterritorial species of Anthidiellum. Additional nest records are given for 
Afranthidium (Immanthidium) repetitum and Afranthidium (Afranthidium) ablusum. A first record (that 
of Alan Weaving) of a host, Megachile (Gronocerus) feline Gerstaecker, of Euaspis abdominalis 
(Fabricius) is reported. Anthidiini in southern Africa are relatively polyphagous. In the present 
analysis inter-generic and intra-generic similarities, differences, and preferences in flower families 
visited are indicated. 



Anthidiini are worldwide in distribu- junodi (Friese) (as Anthidium junodi Friese - 
tion. Michener (2000) recognizes 37 genera Skaife 1950, Taylor 1962, Michener 1968 
of which 18 are represented in Africa south and as Immanthidium junodi (Friese) - Gess 
of the Sahara - 15 in southern Africa. 1981); Afranthidium (Immanthidium) repeti- 
Anthidiine bees are generally divisible into turn (Schulz) (as Anthidium repetitum Schulz 
two groups on the basis of the materials - Michener 1968); Afranthidium (Branthi- 
used for nest construction. One group uses dium) micrurum (Cockerell) (as Anthidium 
plant hairs or plant fibres and the other micrurum Cockerell - Michener 1968); 
resin, often together with pebbles (Mich- Afranthidium (Branthidium) braunsi (Friese) 
ener 1968, Pasteels 1977, Michener 2000). (as Branthidium braunsi - Gess 1981); 
However, V achy anthidium bicolor (Lepele- Afranthidium (Afranthidium) ablusum (Cock- 
tier) is exceptional as it uses a mixture of erell) (as Afranthidium (Oranthidium) prob- 
plant down and resin (Michener 1968, ably odonturum (Cockerell) (Gess and Gess 
2000). 1999); Pseudoanthidium (Micranthidium) 

Surprisingly little has been published on truncatum (Smith) (as Micranthidium trun- 

the nesting of anthidiine bees in southern catum Smith - Friese 1902, Michener 1968); 

Africa. Nesting records are available for Serapista denticulata (Smith) (Stadelmannn 

seven species from three genera, all using 1898, quoted by Friese 1905, 1909, 1916, 

plant fibres: Afranthidium (Immanthidium) Michener 1968). 



Volume 16, Number 1, 2007 



31 



To these published records are added 
first descriptions of the nests of Plesianthi- 
dium (Spinanthidiellum) volkmanni (Friese), 
a resin user, and of Serapista rufvpes (Friese) 
and Afranthidium (Nigranthidium) concolor 
(Friese), plant fibre users, additional rec- 
ords for Afranthidium (Immanthidium) re- 
petition and Afranthidium (Afranthidium) 
ablusum, and a first record (that of Alan 
Weaving) of a host of Euaspis abdominalis 
(Fabricius). 

Although provision from the nests was 
not investigated, flowers visited are dis- 
cussed based on the records assembled for 
13 of the 15 genera known from southern 
Africa. The two genera not represented are 
parasitic in the nests of other bees and 
therefore visit flowers only for adult 
nourishment and egg production. In a com- 
parative overview of flower visiting by 
non-Apis bees in the semi-arid to arid areas 
of southern Africa (Gess and Gess 2004) the 
authors compared diversity of choice be- 
tween and within bee families, in the case 
of Megachilidae to the level of sub-tribe, 
but no comparisons were made within sub- 
tribes. The present contribution, restricted 
to Anthidiini, examines flower visiting at 
inter- and intra-generic levels. 

METHODS 

Nests of anthidiines have been rarely 
encountered by the authors and no full 
nesting studies were conducted by them. 
The notes on nesting are limited to a small 
number of opportunistic observations and 
collections. Voucher specimens of the bees 
and of their nests have been deposited in 
the insect collection of the Albany Muse- 
um. 

Field research by the authors, spanning 
35 years, has been concentrated principally 
within the Karoo Biome and associated dry 
savanna, dry fynbos and desert. Apart 
from in the area around Grahamstown in 
the Eastern Cape where the authors reside 
it has not been possible to work in the field 
throughout the year. In the southeast and 
south peaks of rainfall are in September 



and March and the summer is rarely 
excessively hot and dry. Thus, in these 
areas, sampling has been principally from 
late October through to March. In the 
winter rainfall region of the southwest, 
north to Liideritz and Aus in southwestern 
Namibia, sampling has been from August, 
when temperatures start to rise, through to 
December, when most flowering is over 
and the land as a general rule becomes 
parched. North of the Orange River in the 
southern Kalahari, southeastern Namibia 
and in western Namibia to the north of the 
winter rainfall area, sampling has been in 
March and April, which is when rains are 
expected and excessive heat is past. 

Insects visiting flowers were collected 
using a hand net. All plants in flower at the 
study site were observed for visitors and, 
when possible, were sampled throughout 
the day. In effect anthidiine bees in an area 
all had the choice of visiting all those 
plants that were in flower. Records encom- 
pass both types of visit, for collecting 
nectar and for collecting pollen, indiscrimi- 
nately. 

The records used in the present analysis 
are listed in the Appendix. Except where 
indicated, they are condensed from the 
authors' collection labels. Many of the 
specimens were collected after the cata- 
logue (Gess and Gess 2003) and the 
electronic bee database were closed in 
March 2002. Full locality data, including 
co-ordinates, and exact dates of all speci- 
mens are available from label data and of 
those for specimens processed before 
March 2002 in addition from the electronic, 
relational database. 

The great majority of records relate to 
specimens in the terrestrial insect collection 
of the Albany Museum, Grahamstown. 
Voucher specimens for many but not all 
plants have been deposited in the Schon- 
land Herbarium, Albany Museum, Gra- 
hamstown. Duplicates of specimens from 
Namibia are in the National Herbarium of 
Namibia, Windhoek. A few additional 
records have been added from literature 



32 Journal of Hymenoptera Research 

and from specimens collected by V. White- Plesianthidium (Spinanthidiellum) volk- 
head, determined by C.D.Michener and manni (Friese) was found nesting abun- 
deposited in the collection of the South dantly in electricity boxes placed 1.25 m 
African Museum, Cape Town (see Appen- above the ground, provided for campers in 
dix). The records listed in Struck (1994) the Clanwilliam Dam Resort, Olifants 
have not been included because it is not River Valley (Figs 1, 2). Sixty-nine electric- 
clear whether "the flower visiting records .. ity boxes were inspected. Of these 24 had 
compiled from direct field observations" been used for nesting by P. (S.) volkmanni. 
are supported by voucher specimens. Within these pre-existing cavities clus- 

The value of including single records of ters of up to seven separate spouted, pot- 
visits to a particular plant family by like cells with resinous walls had been 
a particular species of bee have been constructed. The cells were attached hor- 
questioned. Taken singly such records are izontally to a wall of the cavity or 
of little value but taken together with horizontally on the floor of the cavity 
records for other species they are of value (Fig. 3). All of the cells within a cavity 
in indicating plant families visited by bees were orientated in the same direction but 
at higher taxonomic levels. there was no constancy between cavities. 

Specific flower visitors show varying In no cases was there more than one cell 

degrees of diversity of choice, i.e. of under construction, suggesting that each 

oligophagy and polyphagy. In order to cluster had been constructed by a single 

make comparisons between groups of female. 

flower visitors constituted of unequal Construction of a cell was in all cases 

numbers of species Gess (1992, unpub- preceded by the construction of a small 

lished) developed an Index of Diversity of " pa d or saucer", noticeably different in 

Choice at the specific level, using the texture from the cell walls, and attached to 

formula: t he substrate. The cell was then con- 



D = a-b/b x 100 



structed on this base. When a cell had 

been only partially constructed, at the close 

where (7 = the sum of the number of of day, the female slept head down within 

species recorded visiting each of the flower the cell with her gaster curved over within 

families and b = the number of species of the cell so that only the arched terga were 

flower visitors (published in Gess 1996 exposed (Fig. 4). 

page 47). This is an index by which to Measurements were taken from a sample 

compare the degree of oligophagy or of 13 cells (deposited in the collection of the 

polyphagy exhibited by taxa of differing Albany Museum). The average total length 

numbers of species. 'D' would equal if is 14.9 mm including the spout. The cell 

each species only visited one species of without the spout is 11 mm in length and 

plant; the higher the value of 'D' the 7.7 mm in diameter. The walls of the cells 

greater the degree of polyphagy. are circa 0.5 mm in thickness and the 

diameter of the opening of the spout at 

NEW NEST RECORDS the tip circa 0.5 - 1 mm. The channel 

nt . tf ... „ within the spout is filled with very short 

Plesianthidium Cameron , ., , £ . r , L . , , ,.< 

lengths of fine plant material, not fibres or 

Plesianthidium consists of four sub-gen- "fluff". The resinous walls are yellow 

era, Carinanthidium Pasteels, Plesianthidium ochre in colour, initially smooth and pli- 

Cameron, Spinanthidiellum Pasteels, and able. When under construction the walls of 

Spinanthidium Mavromoustakis known the wide portion of the cell are made 

from South Africa onl> , principally from higher than the final height of the cell- 

the west (Michener 2000). proper . The ed are then crimped to 



Volume 16, Number 1, 2007 



33 




Fig. 1. Vegetation on the eastern side of the Clanwilliam Dam, Olifants River Valley, Western Cape. 

Fig. 2. Electricity box in the caravan park on the eastern side of the Clanwilliam Dam. 

Fig. 3. Cells of Plesianthidium (Spinanthidiellum) volkmanni (Friese) constructed inside an electricity box. Note: 

initial "pad or saucer" on which a cell is constructed, partially constructed cell and five completed cells, throe 

attached to the vertical wall and one to the floor. Average actual length of cells 14.9 mm including spout. 

Fig. 4. Two cells of Plesianthidium (Spinanthidiellum) volkmanni (Friese) constructed inside an electricity box. 

Note nest builder sleeping in the incomplete cell. 



34 



Journal of Hymenoptera Research 



form the top of the cell-proper but not 
reaching the centre, construction then 
continues with a gentle narrowing of the 
tube of the spout. 

Provisioning takes place before crimping 
and spout construction. Provision taken 
from a cell (Fig. 5) appeared to be of mixed 
provenance with an admixture of bright 
yellow oil. Pollen from the flowers of 
Aspalathus spinescens, the only flowers from 
which these bees were collected in the 
vicinity of the nests, is represented mixed 
with bright yellow oil. The completed 
provision filled circa 2/3 of the cell-proper. 

The resinous walls are probably imper- 
vious to air. The function of the spout is 
most probably for ventilation. 

The cocoon is attached to the inner 
surface of the wall of the cell-proper with 
the papilla within the base of the spout. 
Silken threads are visible within the brittle, 
highly varnished, brown walls of the 
cocoon. 

Emergence from the cocoon is through 
the side of the cell-proper, a large exit hole 
being cut by the emerging adult (Fig. 6). 

There was a high level of success, P. (S.) 
volkmanni having emerged from most of 
the circa 20 cells collected. However, 
a meloid larva emerged from one of the 
cells and pupated attached to the outside 
of the cell (Fig. 7). 

Serapista Cockerell 

The Afrotropical genus Serapista Cock- 
erell consists of four species, two of which 
are represented in southern Africa - the 
widespread S. denticulata (Smith) recorded 
from central, eastern and southern Africa 
and the more restricted S. rufipes (Friese) 
recorded only from southern Africa (Pas- 
teels, 1984). 

Published data concerning the nests of 
Serapista all appear to apply to S. denticulata 
Smith (Stadelmann, 1898, quoted by Friese, 
1905 and 1909, Friese 1916, Michener 1968, 
with republication of his Figs 27-28 by 
Roubik 1989, Michener, 2000). Michener 
(1968) illustrated and gave a full descrip- 





Fig. 5. Cell of Plesianthidium (Spinanthidiellum) volk- 
manni (Friese) cut through longitudinally to show 
provision and young larva. 

Fig. 6. Cell of Plesianthidium (Spinanthidiellum) volk- 
manni (Friese) showing emergence hole and imagine. 
Fig. 7. Cell of Plesianthidium (Spinanthidiellum) volk- 
manni (Friese) from which a meloid beetle emerged, 
showing pupal skin of and adult of the beetle. 

tion of a nest from Malawi, commented on 
further nests in the British Museum from 
Natal, and summarized the accounts of the 
earlier authors. However, the assumption 
has been made that the form of the nest of 
S. denticulata and its aerial situation holds 



Volume 16, Number 1, 2007 35 

good also for the other species of the of southern Africa (Gess and Gess 2004). 

genus. Thus it is stated by Michener Although Serapista rufipes was commonly 

(1968) that "bees of this genus make encountered and observed throughout the 

exposed nests of down" and by Michener south-western areas of South Africa and 

(2000), with regard to the genus, that widely in Namibia, no further nests were 

"nests, masses of plant down often inter- found, nor were nests found in the collec- 

mixed with animal hairs or even feathers tions of the South African Museum (Mar- 

and placed on plant stems have been gie Cochran pers. com.) nor in the National 

described by several authors..." Collection, PPRI, Pretoria (Connal Eardley 

In the Albany Museum, collected by the pers. com.), both collections with bees as 

present authors in the vicinity of Grahams- one of their specializations. The authors 

town, Eastern Cape (Fig. 8), are two nests had frequently observed these bees flying 

of S. denticulata (Figs 9, 10), identified as low over the ground but had been unable 

such from, in one instance, the capture of to observe what they were doing. It was 

the builder and, in both from the bees only in October 2005, when the authors 

reared from them (a female and a male were sampling flower visitors on the banks 

from one and eight females and three of the lagoon at Lamberts Bay (Fig. 13) that 

males from the other). Additional identi- the second author noticed a Serapista rufipes 

fied typical nests of this species are two in bee disappearing into the ground. On 

a public display in the Albany Museum closer examination it was found that it 

and others held in the collection of the had entered a plant down tube projecting 

South African Museum (Margie Cochran from the ground. Excavation showed that 

pers. com.). Typical nests are roughly oval it had constructed a five-celled, down nest 

in shape with a short entrance tube at the within a, presumably pre-existing, burrow 

higher end. in the sandy soil (Fig. 14). The total length 

The four Grahamstown nests like those of this nest was 90 mm of which 42 mm 
described in literature are aerial nests built was of an entrance tube, 10 mm in di- 
on shrubs, the recorded height above ameter. The lower part of the nest contain- 
ground of one being circa 50 cm. Two ing the cells was 20 mm across at its widest 
consist entirely of plant down, the other point, at which two cells had been con- 
two in patches incorporate fine gray structed side by side. One cell, from which 
mammalian fur. The nests illustrated are no emergence had taken place, was opened 
76 mm and 70 mm in height and 29 mm in June 2006. In it was uneaten provision, 
and 50 mm in diameter respectively. They Some of this, examined microscopically, 
were constructed on stems of Elytropappus was found to contain pollen of mixed 
(Asteraceae) and Rhus (Anacardiaceae). provenance, two pollen types being pres- 
From a third nest, similar in construction ent. Most of the pollen in the sample was 
and placement, collected by W.A. Clarke spherical and thin-smooth-walled. It was 
from Lebeckia (Fabaceae: Papilionoideae) possible that it was from Conicosia sp. 
near Twee Rivieren, Gemsbok National (Aizoaceae: Mesembryanthema), several 
Park in 1966 (Figs 11, 12), were reared four large plants of which were growing in 
female and three male Serapista rufipes close proximity to the nest. A small 
Friese (nest and bees in the collection of percentage of the pollen grains, also rela- 
the Albany Museum), supporting the belief tively thin, smooth-walled, were about 
that Serapista constructs only aerial nests, twice the size, elongate-oval and consider- 
This nest was similar in size, 72 mm in ably longer than broad, 
height and 43 mm in diameter. In late summer two adults emerged, 

The authors sampled bees visiting flow- each making its way out through the side 

ers throughout the semi-arid to arid areas of its cell (Fig. 15). 



36 



Journal of Hymenoptera Research 




Fig. 8. Strowan Farm, northwest of Grahamstown, Eastern Cape, the area in which nests of Serapista denticulata 

Smith were collected. 

Fig. 9. Nest of Serapista denticulata Smith constructed on a shrub, Elytropappus rhinocerotis (Asteraceae), at 

Strowan Farm. 

Fig. 10. Nest of Serapista denticulata Smith constructed on a shrub, Rhus sp. (Anacardiaceae) at Goodwin's 

Kloof, neighbouring Strowan Farm. 



Volume 16, Number 1, 2007 



37 






Fig. 11. Southern Kalahari, near Twee Rivieren, Northern Cape, the area in which a nest of Serapista rufipes 
Friese constructed on a shrub, Lebeckia linearifolia (Fabaceae: Papilionoideae), as seen in the foreground, 
was collected. 

Fig. 12. Nest of Serapista rufipes Friese constructed on a stem of a shrub, Lebeckia linearifolia (Fabaceae: 
Papilionoideae), in the southern Kalahari near Twee Rivieren. 



38 



Journal of Hymenoptera Research 




Fig. 13. Southern bank of the lagoon at Lamberts Bay, Western Cape, the area in which a nest of Serapista 

rufipes Friese constructed in a cavity in the ground was found. The clip board on the ground is midway between 

the site of the nest and a large plant of Conicosia (Aizoaceae) beyond the nest. 

Fig. 14. Sand cleared away to show nest of Serapista rufipes at Lamberts Bay. 

Fig. 15. Nest of Serapista rufipes Friese from Lamberts Bay showing emergence holes and an imagine. 



Volume 16, Number 1, 2007 39 

Clearly, S. rufipes is remarkable in that it the vicinity of the Obib Mountains and 

may either construct an aerial nest or may Boegoeberg northwards to the Klinghardt 

construct its nest within a pre-existing Mountains and the north/south road west 

cavity in the ground. The fact that the of Grillental. 

authors have on several occasions and at All the recorded nests of A. (Immanthi- 

several sites observed bees of this species dimn) junodi were constructed in pre-exist- 

flying low over the ground suggests that ing tubular cavities, which necessitated the 

the latter strategy may not be unusual. construction of the cells in a single linear 

series, one female nest builder per cavity. 

Afranthidium Michener Similar nests have been obtained by the 

The genus Afranthidium, divided into present authors from the vicinity of Gra- 

eleven sub-genera, is principally sub-Sa- hamstown, Eastern Cape. By contrast the 

haran with one sub-genus in the Palaearc- nests of A. (I.) repetitum are constructed in 

tic and at least two other species occurring relatively large cavities. The record pub- 

in this region (Michener 2000). lished by Michener (1968) is based on part 

A nest of A. (Nigranthidium) concolor (estimated at one fifth) of a nest removed 

(Friese) was found in a bare sandy area from an electricity meter box in Estcourt 

on a slope at SorsSors in the Kamiesberg, Natal and housed in the Natal Museum. 

Namaqualand. It was in an early stage of This part of the nest, constructed from 

construction. Projecting from the mouth of plant down, contained an estimated 350 

a vertical burrow, 6 mm in diameter and cells or cocoons resulting in the estimation 

41 mm deep, apparently pre-existing, was of the total number of cells in the nest 

a short entrance tube constructed from having been 1,750 and the conclusion that 

plant fluff and at the base of the burrow the nest had been constructed by a consid- 

was an open, as yet un-provisioned cell, erable group of females, 

similarly constructed from plant fluff. The remains of a much smaller nest and 

Examination of the fluff showed it to have an associated adult bee of A. (I.) repetitum 

been obtained from the seeds of Eriocepha- were submitted to the second author for 

his (Asteraceae) growing nearby. identification. They came from a Cape 

Afranthidium (Afranthidium) ablusum Town householder who had noticed a bee 
(Cockerell) was found nesting and shelter- entering a heavy-duty vice on a work- 
ing (females and males) in empty shells of bench in his garage. He had later ex- 
desert snails, Trigonephrus (Mollusca: Gas- tracted this nest from a cavity in the vice, 
teropoda: Dorcasiidae), in sparsely vege- in which it had been constructed, and 
tated, desertic areas north and south of the a single cell from an electrical double 
Orange River, east of Oranjemund and adapter in the same garage. Regrettably 
Alexander Bay (Gess and Gess 1999). The the remains of the nests received were so 
cells were embedded in a mass of white, mangled that no further information could 
closely packed, cotton-wool like plant be derived from them. Clearly, the size of 
fibres. Further nests in Trigonephrus shells the cavity used for nesting will dictate the 
were found at eight additional sites by the number of bees which can nest in the 
present authors during the course of more cavity and the number of cells which can 
extensive sampling and investigation of be constructed. 

desert snail shells in 2002, 2003 and 2005. Subsequently, in June 2000, J. Cardale of 

One site, the most southerly, was east of CSIRO, Canberra, Australia wrote that this 

Port Nolloth, Namaqualand, and the other bee had become established in southern 

seven were in the Sperrgebiet, Diamond Queensland, Australia. Being a nester in 

Area no. 1, in the winter rainfall area of pre-existing cavities, the species is an ideal 

Namibia, in the south from the plains in stowaway candidate, making its accidental 



40 



Journal of Hymenoptera Research 



transfer with household goods an easy 
matter. 

The origin of the "wool" used by A. (I.) 
repetitum has not been established, howev- 
er, Taylor (1962) recorded that Jacot Guil- 
larmod had seen A. (I.) junodi removing 
fibres from the stems of Helichrysum 
(Asteraceae) in Lesotho (as Basutoland). 

Euaspis Gerstaecker 

Euaspis is widespread in Africa, from 
Nigeria to Kenya and south to South 
Africa, and in southern and eastern Asia 
(Michener 2000). Of the 12 described 
species two only are African. Euaspis has 
been recorded as parasitizing other Mega- 
chilidae (Lithurge and chalicodomiform 
Megachile) (Michener 2000). Iwata (1976: 
page 420) states of "Parcvaspis = Euaspis" 
that, "while it is reported to parasitize 
Lithurge (Lieftinck, 1939), observations 
show that the Japanese P. basalis is a clep- 
toparasite of Chalicodoma sculpturalis only 
(Iwata, 1933). This species does not para- 
sitize any species other than those which 
make nests with resin. In Southeast Asia, it 
is reported to live in the nest of Ch. 
disjuncta or other allied species and even 
in Japan, it probably lives on Ch. disjuncti- 
formis ". He goes on to give a detailed 
account of the activities of this bee in 
converting its host's nest to its own. 

Three Euaspis abdominalis (Fabricius) 
were reared from nests of Megachile (Gro- 
nocerus) felina Gerstaecker from northern 
Natal by A.J.S. Weaving in 1992 (specimens 
in the Albany Museum collection). Of these 
one, a male, is from a nest constructed in 
an old Synagris (Eumeninae) mud nest 
from Umlalazi Nature Reserve and two, 
a female and a male, are from a nest 
constructed within a length of reed put out 
as a trap-nest in the Lake St Lucia Game 
Reserve. 

Some further support for a possible pre- 
dilection for parastising chalicodomiform 
Megachile is given by two observations in 
the vicinity of Grahamstown. A female E. 
abdominalis was found by R.W. Gess shel- 



tering in an aerial mud-nest, probably of 
Megachile (Gronocerus) cincta (Fabricius), 
build under the windowsill of a brick 
building (specimen in collection of the 
Albany Museum). The same was observed 
investigating holes in a vertical sandstone 
bank but it was not established which 
nests were being parasitized although 
circumstantial evidence suggested that it 
may have been associated with the nests 
of Megachile (Pseudomegachile) schulthessi 
Friese (as Chalicodoma (Pseudomegachile) 
schulthessi (Friese)) (Gess 1981). 

DISCUSSION 

The discovery that Serapista rufipes con- 
structs nests in two very different situa- 
tions, one exposed above ground and the 
other within a burrow in the ground, was 
a surprise. However, nesting by a single 
species both aerially and in cavities is not 
unknown in the aculeate hymenoptera, 
Celouites michaelseni von Schulthess (Vespi- 
dae: Masarinae), the cells of which are 
known to be constructed attached to rocks, 
albeit in a somewhat sheltered position (as 
Celouites gariepensis Gess - Gess et al 1997), 
has been found to nest also in pre-existing 
cavities in the ground (Gess and Gess 
fieldnotes 1998). 

As noted above Serapista rufipes has 
frequently been seen flying low over the 
ground, suggesting that nesting in the 
ground may be common. It therefore seems 
possible that other species of anthidiines 
which have been found nesting in exposed 
aerial situations may be found also to nest 
in cavities, perhaps explaining why so few 
anthiidine nests have been found. 

The construction of spouted pot-shaped 
resin cells here recorded for P. (S.) volk- 
manni is not unique to Plesianthidium. A 
commonly known example is the Europe- 
an species Authidiettum (Anthidiellum) stri- 
gatum Panzer for which there are several 
published accounts. An account of cell 
construction, provisioning and closure to- 
gether with excellent photographic illus- 
trations is to be found in Bellmann (1995, 



Volume 16, Number 1, 2007 



41 



pages 275-277). Two other examples of 
spouted pot-shaped resin cells of species of 
Anthidielium (Anthidiellum), A. notation La- 
treille and A. perplexum (Smith), are illus- 
trated by Baker et al. (1985, page 293). The 
recorded positioning of the cells of A. (A.) 
strigatum and A. (A.) notatum differs from 
that here recorded for P. (S.) volkmanni. 
Instead of being concealed within a cavity 
they are build in the open. Those of A. (A.) 
strigatum are attached to the sunnyside of 
tree trunks, plant stems, stones or rocks. 
The cells of A. perplexum illustrated in 
Baker et al. had been constructed in 
cavities but, unlike those of P. (S.) volk- 
manni, had been seated in pebbles and 
separated from each other by cell parti- 
tions. Clearly these cells will always be 
found to be constructed within cavities, 
however, it should not be assumed that the 
free-standing cells of P. (S.) volkmanni are 
always constructed within cavities. The 
fact that S. rufipes constructs its nests either 
on plant stems or in cavities indicates that 
such an assumption would not be justified. 
A second but probably constant difference 
is the orientation of the cells. Whereas the 
cells of P. (S.) volkmanni were all positioned 
with their long axis sub-horizontal those of 
A. (A.) strigatum and those of A. notatum 
hang down from the substrate so that the 
ventilation tube points downwards. Clo- 
sure of the cell is undertaken in a similar 
manner. Both P. (S.) volkmanni and A. (A.) 
strigatum, using their mandibles, crimp the 
cell wall at the opening in such a way that 
the edges are drawn inwards leaving only 
a narrow collar-like opening, much in the 
way in which a potter draws in the walls of 
a jug to form the neck. The crimping is not, 
however, smoothed away but is visible in 
the completed cells. The spout or ventila- 
tion tube is then constructed. Those of A. 
(A.) strigatum and A. notatum are slightly 
flared at the opening whereas that of P. (S.) 
volkmanni is not. Whereas the resin walls of 
the concealed cells of P. (S.) volkmanni are 
smooth and in colour contrast with the 
walls of the cavity those of A. (A.) strigatum 



are camouflaged by the addition of bark 
fragments to the resin giving the outer 
surface a flaky finish which furthermore in 
colour blends with that of the substrate. 

FLOWER VISITING 

In a study of host plant specialization in 
western Palaearctic anthidiine bees (Muller 
1996), pollen sources of 72 anthidiine 
species of Europe, North Africa and Asia 
Minor were investigated by analysis of 
pollen loads from females. Muller found 
that 43% of the species studied by him 
were oligolectic (relatively specialized as to 
pollen source), 35% were polylectic (visit- 
ing up to 17 plant families for pollen), and 
the remaining 4% insufficiently known. 
The principal pollen sources throughout 
were Fabaceae, Lamiaceae, Asteraceae, 
Dipsacaceae, Campanulaceae, and Zygo- 
phyllaceae. Of these all except Dipsaceaeae 
are families included in the list of families 
visited by southern African anthidiines. 

In the comparative overview of flower 
visiting by non-Apis bees in the semi-arid 
to arid areas of southern Africa (Gess and 
Gess 2004) the Gesses reported that of the 
34 plant families found to be attracting 
visits to flowers by bees 30 were visited by 
Megachilidae (Gess and Gess 2004: Ta- 
ble 2) and of these 20 were visited by 
anthidiines (Gess and Gess 2004: Table 5). 
Of the plant families Fabaceae (almost 
exclusively Papilionoideae) received visits 
from 61% of the 44 species of anthidiines 
for which records were obtained. The three 
other families receiving visits from a signif- 
icant, though appreciably lower, percent- 
age of the species of anthidiines are 
Asteraceae (32%), Malvaceae (all species 
of Hermannia, formerly Sterculiaceae) (27%) 
and Lamiaceae (23%). 

In the overview (Gess and Gess 2004) 
flower visiting within the Megachilidae 
was compared between sub-families and 
tribes. Values for Diversity of Choice (see 
Methods, present paper) were calculated at 
the family level. The values obtained were 
- Andrenidae 177.77, Apidae 195.93, Col- 



42 



Journal of Hymenoptera Research 



Table 1 Numbers of species by genus of Anthidiini (Megachilinae) recorded visiting flowers of the listed 
plant families Numbers of species below generic names denote the number of species for which flower visiting 
records are available. Numbers before asterisks denote the numbers of species for which five or more such 
records were obtained. 



Afranthidium 
21 species 
Anthidkllum 
1 species 
Anthidioma 

1 species 
Anthidium 
3 species 
Aspidosmia 

2 species 
Cyphanthidium 

2 species 
Eoanthidium 

1 species 
Pachyanthidiwn 

3 species 
Plesianthidium 
6 species 
Pseudoanthidium 

2 species 
Serapista 
2 species 
Trachusa 
1 species 

Total number of 
species (45) 



I 1 



5 


2 


1 2 


13 


1* 




1* 


6* 

1 

1 




1 




1 
1* 

1 

1 
1* 




1 


1 


3 






1* 


3* 


4 




1 


5 






1* 


3* 

1 
1* 

1 

1* 


9 


4 


1 4 


28 


1* 




3* 


16* 



M 



O 



1 1 







1 






1 






r 






1* 




1 


l 








5 




2 






1 5 


1* 




1* 






1* 
1 1 

r 


1 








1 




1* 












12 


1 


9 


1 


2 


5 10 


4* 




2* 


1* 


3* 


3* 1 



2 
1* 



15 

8* 



A = Asphodelaceae; B = Iridaceae; C = Aizoaceae; D = Amaranthaceae; E = Molluginaceae; F 
Zygophyllaceae; G = Fabaceae; H = Polygalaceae; I = Brassicaceae; ] = Malvaceae; K = Loasaceae; L 
Boraginaceae; M = Vahliaceae; N = Apocynaceae; O = Acanthaceae; P = Lamiaceae; Q = Scrophulariaceae; R 
Apiaceae;S = Asteraceae; T = Campanulaceae. 



letidae 50.00, Halictidae 202.90, Megachili- 
dae 130.61, Melittidae 70.00. When the 
formula is applied to Table 1 a value of 
146.67 is obtained for Anthidiini. This 
suggests a relatively high degree of po- 
lyphagy for the Anthidiini overall. How- 
ever, whereas some of the solitary apids 
were recorded from flowers of over twenty 
families no anthidiine was recorded from 
more than seven families. 

In the present contribution flower visit- 
ing at the levels of genera and for some 
genera sub-genera and species are exam- 
ined. In most cases the number of females 
is too low for analysis of their pollen loads 
to give an accurate evaluation of oligolecty 



and polylecty and so this has not been 
attempted. However, summation of flower 
visiting records for females and males does 
give some indication of preferences for and 
differences in preferences for flowers of 
particular taxa and possibly degree of 
oligophagy and polyphagy (diversity of 
plants visited to obtain pollen and nectar 
combined) within the southern African 
anthidiines. 

At the generic level nine of the 13 
genera of anthidiines included were re- 
corded from Fabaceae, however, none was 
restricted to Fabaceae, flowers of 1-13 
plant families being visited. However, if 
numbers of records of visits to Fabaceae 



Volume 16, Number 1, 2007 



43 



Table 2. Numbers of species by subgenus of Afranthidium (Megachilinae: Anthidiini) recorded visiting 
flowers of the listed plant families. Numbers of species following generic names denote the number of species 
for which flower visiting records are available. Asterisks denote the numbers of species for which five or more 
such records were obtained. 



Afranth-5 spp. Branth-6 spp. Capanth-1 sp. Domanth-1 sp. Immanth-3 spp. Nigranth-1 sp. Oranth-3 spp. 



Aizoaceae 


2 


2* 


Molluginaceae 






Amaranthaceae 




2 


Zygophyllaceae 
Fabaceae 


1 
5** 


1 


Malvaceae 


1 


l 


Boraginaceae 






Vahliaceae 




l 


Apocynaceae 
Lamiaceae 




l 


Scrophulariaceae 
Asteraceae 




2* 


Campanulaceae 




2 



1* 

2 

3 
1 
1 
1 



3 



compared with visits to other plant 
families are considered (Table 1), Faba- 
ceae is the most frequently visited family. 

When visiting is considered at sub- 
generic and specific levels, some possible 
differences in preferences are apparent. Of 
the seven sub-genera of Afranthidium (Ta- 
ble 2), only four were recorded from 
Fabaceae. Of these, two of the five species 
of A. (Afranthidium), four of the six species 
of A.(Branthidium) and one of the three 
species of A. (Immanthidium) were recorded 
five or more times from Fabaceae but none 
of the three species of A. (Oranthidium) was 
recorded five or more times. Two species 
of Oranthidium were, however, the only 
species of the genus Afranthidium recorded 
five or more times from Hermannia (Mal- 
vaceae). One species of A. (Branthidium) 
and the single species of A. (Capanthidium) 
(not recorded from flowers of any other 
plant family) were recorded at least five 
times from Asteraceae. 

When flower visiting records for the six 
species of A. (Branthidium) are separated 
(Table 3), it is seen that braunsi, haplogas- 
trum, matjiesfonteinense and minutulum 
were recorded 29 (14 female and 15 male), 
11 (seven female and four male), 7 (one 
female and six male) and 17 (12 female and 
five male) times respectively from Faba- 



ceae suggesting a preference for Fabaceae 
by at least three of these species. However, 
minutulum was recorded 30 (15 female and 
15 male) times from Asteraceae suggesting 
an equal, if not greater, preference for 
Asteraceae. This species was furthermore 
recorded seven times (five female and two 
male) from Aizoaceae and was also found 
visiting flowers of four other families 
suggesting that this species at least is 
polyphagous. 

Of the three sub-genera of Plcsianthidium 
(Table 4) all were recorded from Fabaceae 
and furthermore the single species of P. 
(Carinanthidium), the single species of P. 
(Smnanthidiellum) and two of the four 
species of P. (Spinanthidium) were recorded 
five or more times from Fabaceae. In 
addition, one of the four species of P. 
(Spinanthidium) was recorded five or more 
times from Boraginaceae and one five or 
more times from Malvaceae (Hermannia 
only). 

When the flower visiting records for the 
four species of P. (Spinanthidium) are 
separated (Table 5), it is seen that neli and 
trachusiforme were recorded from Fabaceae 
18 (5 female and 13) and 19 (6 female and 
13 male) times respectively suggesting 
a possible preference by these species for 
Fabaceae. The records of three female and 



44 



Journal of Hymenoptera Research 



Table 3. Numbers of records by species of the subgenus Branthidium of the genus Afranthidium 
(Megachilinae: Anthidiini) of visits to flowers of the listed plant families. 



Aizoaceae 

Amaranthaceae 

Zygophyllaceae 

Fabaceae 

Malvaceae 

Vahliaceae 

Apocynaceae 

Asteraceae 

Campanulaceae 



braunsi guillarmodi 



2 
29 



liavhwtiftrum 



2 
2 

11 



jocosum matjiesfonteinense 



minutulwn 



3 

17 
2 
4 
1 

30 



two male of taclmsifornie from Boraginaceae 
probably indicate a possible secondary 
preference only. The species for which 
more than five records were obtained from 
Malvaceae is cnllescens (5 female and 12 
male). Although this may possibly indicate 
a preference it should be noted that only 
five of the 17 records are for females and 
that two females were recorded from 
Lamiaceae and one female from Astera- 
ceae. 

Further sampling of flowers for anthi- 
diines would clearly be rewarding and 
would make more substantiated conclu- 
sions with respect to flower preferences 
possible. 

Table 4. Numbers of species by subgenus of 
Plesianthidium (Megachilinae: Anthidiini) recorded 
visiting flowers of the listed plant families. Asterisks 
denote the numbers of species for which five or more 
such records were obtained. 

Carinanthidium Spinanthidiellum Spinanthidium 
1 species 1 species 4 species 



Asphodelaceae 


1 


Iridaceae 




Aizoaceae 




Zygophyllaceae 


1 


Fabaceae 


1* 


Polygalaceae 


1 


Brassicaceae 




Malvaceae 




Boraginaceae 




Acanthaceae 




Lamiaceae 


1 


Asteraceae 


1 



1 

3 



1 

4* 

1* 

3* 
1 



ACKNOWLEDGEMENTS 

Grateful thanks are expressed to: all those who gave 
access to their land or land in their care; all those 
bodies who issued permits for the collection of insects 
and plant samples; the authors' sons, David, Harold 
and Robert Gess for field assistance at various times; 
Margie Cochrane, South African Museum, Iziko 
Museums, Cape Town for providing data from labels 
of specimens in the collection in her care; Bronwyn 
McLean of the Graphic Services Unit, Rhodes Univer- 
sity for assistance with the preparation of the figures; 
the Council for Scientific and Industrial Research 
(CSIR), the Foundation for Research Development 
(FRD), and currently the National Research Founda- 
tion of South Africa (NRF) for running expenses 
grants; the Board of Trustees of the Albany Museum 
for Research Contracts granted to the authors since 
2003, which have given the authors continued use of 
the museum's facilities since their retirements; and the 
editors of the Journal of Hymenoptera Research and 
a referee for constructive criticism. 



Table 5. Numbers of records by species of the 
subgenus Spinanthidium of the genus Plesianthidium 
(Megachilinae: Anthidiini) of visits to flowers of the 
listed plant families. 



bruneipet 



callescens neli trachusijbrme 



Iridaceae 




Aizoaceae 


1 


Fabaceae 


3 


Brassicaceae 




Malvaceae 


1 


Boraginaceae 




Lamiaceae 


2 


Asteraceae 







2 


2 




18 


19 


7 


3 


1 




1 


5 


6 




2 


-> 







Volume 16, Number 1, 2007 



45 



LITERATURE CITED 

Armbruster, W. S. and K. E. Steiner. 1992. Pollination 
ecology of four Dalechampia species (Euphorbia- 
ceae) in northern Natal, South Africa. American 
Journal of Botany 79: 306-313. 

Baker, J. R., E. D. Kuhn, and S. B. Bambara. 1985. Nests 
and immature stages of leafcutter bees (Hyme- 
noptera: Megachilidae). Journal of the Kansas 
Entomological Society 58: 290-313. 

Bellmann, H. 1995. Bienen, Wespen, Ameisen - Haut- 
flilgler Mitteleuropas. Stuttgart: Franckh-Kosmos. 
336 pp. 

Brauns, H. 1926. V. Nachtrag zu "Friese, Bienen 
Afrikas". Zoologische Jahrbiicher, Abteilung fiir 
Systematic, Geographie und Biologic der Tiere 52: 
187-230. 

Friese, H. 1902. Zwei neue Bienengatrungen. Zeitschrift 
fiir Systematische Hymenopterologie und Dipterologie 
2: 186-187. 

. 1905. Die Wollbienen Afrikas, Genus Antlu- 

diiini. Zeitschrift fiir Systematische Hymenopterologie 
und Dipterologie 5: 65-75. 

. 1909. Die Bienen Afrikas nach dem Stande 

unserer heutigen Kenntnisse. pp. 83-476, pis ix - 
\ in: Schultz, L., ed. Zoologische und Anthropolo- 
gische Ergebnisse einer Forschungsreise im westlichen 
und zentralen Sudafrika ausgefilhrt in den jahren 
1903 - 1905. Band 2, Lieferung 2, X Insecta (ser.3) 
[Jenaische Denkschriften Vol. 14]. Jena: Fischer. 

. 1916. Hymenoptera V: Apidae. pp. 415^152, 

pi. viii in: Michaelsen, W. ed. Beitrage zur Kenntnis 
der Land- und Siisswasserfauna Deutsch-Sudwesta- 
frikas. Hamburg: L. Friederichsen and Co. 

Gess, F. W. 1981. Some aspects of an etiological study 
of the aculeate wasps and the bees of a karroid 
area in the vicinity of Grahamstown, South 
Africa. Annals of the Cape Provincial Museums 
(nat. Hist.) 14: 1-80. 

and S. K. Gess. 1999. The use by wasps, bees 

and spiders of shells of Trigonephrus Pilsb: 
(Mollusca: Gasteropoda: Dorcasiidae) in the des- 
ertic winter rainfall areas of southern Africa. 
Journal of Arid Environments 43: 143-153. 

Gess, S. K. 1992. Ecology and natural history of the 
masarid wasps of the world with an assessment of their 
role as pollinators in southern Africa (Hymenoptera: 
Vespoidea: Masaridae). Ph.D. dissertation, Rhodes 
University. 412 pp. (unpublished) 

. 1996. The Pollen Wasps: ecology and natural 

history of the Masarinae. Cambridge, Mass: Har- 
vard University Press. 340 pp. 

and F. W. Gess. 2003. A catalogue of flower 

visiting records for aculeate wasps and bees in the 
semi-arid to arid areas of southern Africa. Grahams- 
town: Albany Museum. 529 pp. 

and F. VV. Gess. 2004. A comparative overview 

of flower visiting by non- Apis bees in the semi- 
arid to arid areas of southern Africa. Howard 



Evans Memorial Volume, journal of the Kansas 
Entomological Society 77: 602-618. 
— and F. W. Gess. 2006. Survey of flower visiting 
by solitary aculeate wasps and bees in semi-arid 
to arid areas of southern Africa. Annals of the 
Eastern Cape Museums 5: 1-52. [Previously listed 
as in press Cimbebasia 20 but withdrawn due to 
the publication of that journal having been 
discontinued indefinitely] 
-, F. W. Gess, and R. W. Gess. 1997. Update on 



the flower associations of southern African 
Masarinae with notes on the nesting of Masarina 
strucki Gess and Celonites gariepensis Gess (Hyme- 
noptera: Vespidae: Masarinae). Journal of Hyme- 
noptera Research 6: 75-91. 

Iwata, K. 1933. Studies on nesting habits and parasites 
of Megachile sculpturalis Smith (Hym.: Megachili- 
dae). Mushi 6: 4-24. 

. 1976. Evolution of Instinct: Comparative Ethology 

of Hymenoptera. New Delhi: Amerind Publishing 
Co. Pvt. Ltd. 535 pp. 

Lieftinck, M. A. 1939. Uit het leven van Lithurgus 
atratus, een Indisch houtbijrje. De Tropische Natuur 
28: 193-201. 

Michener, C. D. 1968. Nests of some African mega- 
chilid bees, with description of a new Hoplitis 
(Hymenoptera, Apoidea). journal of the entomolog- 
ical Society of soutliern Africa 31: 337-359. 

. 2000. The bees of the world. Baltimore: John 

Hopkins University Press. 913 pp. 

Muller, A. 1996. Host-plant specialization in western 
Palearctic anthidiine bees (Hymenoptera: Apoi- 
dea: Megachilidae). Ecological Monographs 66: 
235-257. 

Pasteels, J. J. 1977. Une revue comparative de 
l'ethologie des Anthidiinae nidificateurs de l'an- 
cien monde. Annates de la Societe Entomologique de 
France (n.s.) 13: 651-667. 

. 1984. Revision des Anthidiinae (Hymenop- 
tera, Apoidea, Megachilidae) de I' Afrique sub- 
saharienne. Memoires de la Classe des Sciences, 
Academic Royale de Belgique Collection in-4 - 2 erie 
19: 1-165. 

Roubik, D. W. 1989. Ecology and natural history of 
tropical bees. Cambridge: Cambridge University 
Press. 514 pp. [Fig. 3.16 is a republication of Figs 7 
and 28 of Michener, 1968.] 

Skaife, S. H. 1950. Some South African solitary bees. 
South African journal of Science 46: 183-187. 

Stadelmann, H. 1898. Die Hymenopteren Ost-Afrikas, 
pp. 1-55, 1 pi. in: Mobius, K., ed. Die Tierwelt Ost- 
Afrikas und der Nachbargebiete. Band IV. Berlin. 

Struck, M. 1994. A check-list of flower visiting insects 
and their host plants of the Goegab Nature 
Reserve, Northwestern Cape, South Africa. Bon- 
tebok9: 11-21. 

Taylor, J. S. 1962. A note on the carder bee Antltidium 
junodi melanosomum Cameron (Hymenoptera: 



46 



Journal of Hymenoptera Research 



Megachilidae). Pan-Pacific Entomologist 38: 
244-248. 

APPENDIX: FLOWER VISITING 

RECORDS FOR ANTHIDIINE BEES 

(HYMENOPTERA: MEGACHILIDAE: 

MEGACHILINAE: ANTHIDIINI) IN 

SOUTHERN AFRICA 

The data given below, except where 
indicated, are condensed from the authors' 
collection labels. Many of the specimens 
were collected after the catalogue (Gess 
and Gess, 2003) and the electronic bee 
database were closed in March 2002. Full 
locality data, including co-ordinates, and 
exact dates of all specimens are available 
from label data and of those for specimens 
processed before March 2002 are available 
from the electronic, relational database. 
Afranthidium Michener - eleven sub-genera, 10 
found in the Afrotropical Region (Michener 
2000). 
Afranthidium (Afranthidium) Michener - 
known from the "Cape Province", South 
Africa, and from Namibia (Michener 2000). 
Pasteels (1984) includes nine species but 
Michener (2000) considers that some may 
belong elsewhere. 
Afranthidium (A.) ablusum (Cockerell): Fabaceae 
(Papilionoideae), Aspalathus, 36 j, Clanwil- 
liam in the Olifants River Valley in the 
Western Cape, early October. 
Afranthidium (A.) biangulatum Pasteels: Fabaceae 
(Papilionoideae), Indigofera , 299, between 
Rosh Pinah and Sendelingsdrif, Namibia, 
mid-October; Aizoaceae (Mesem- 

bryanthema), 19, between Steinkopf and 
Vioolsdrif in northern Namaqualand, mid- 
October. 
Afranthidium (A.) hamaticauda Pasteels: Fabaceae 
(Papilionoideae), Indigofera and Lcbeckia, 11 99 
and 11 S$, six sites from Wallekraal, North- 
ern Cape to Luderitz, Namibia, late Septem- 
ber to early October; Malvaceae (formerly 
Sterculiaceae), Hermannia , 1 j\ Wallekraal 
and 14 99, northeast of Aus and at Swakop- 
mund in Namibia, early to mid-September; 
Zygophyllaceae, Zygophyllum, 699 and 11 $$, 
Swakopmund, Namibia, mid-March. 
Afranthidium (A.) karooense (Brauns): Fabaceae 
(Papilionoideae), indigofera, 499 and 56*6*/ in 
the vicinity of Oranjemund and Aus, Nami- 



bia, late September and early March; Aizoa- 
ceae (Mesembryanthema), 299 and lj 1 , Rich- 
tersveld and southeastern Namaqualand, 
Northern Cape, late September; Zygophylla- 
ceae, 299 and 1 j, Richtersveld and south- 
eastern Namaqualand, Northern Cape, late 
September. 

Afranthidium (A.) reicherti Brauns: Fabaceae 
(Papilionoideae), Aspalathus and Psoralea, I9 
and 2jo, vicinity of Ceres and of Oudt- 
shoorn, Western Cape, late October and early 
December. 

Afranthidium (Brantliidium) Pasteels - found 
from Lesotho to the Western Cape, north to 
Zaire and Kenya (Michener 2000). Pasteels 
(1984) recognized 10 species. 

Afranthidium (B.) braunsi (Friese): Fabaceae 
(Papilionoideae), Aspalathus and Indigofera, 
1599 and 196*6*/ southwestern Cape, mid- 
October and late-November, and south and 
north of the Orange River, western Northern 
Cape and southwestern Namibia, late-Sep- 
tember to mid-October; Amaranthaceae, 
Hermbstaedtia, 26*6*/ Richtersveld, Northern 
Cape, late-September; Asteraceae, Tripteris, 2 
99, 16*/ north of the Orange River, western 
Northern Cape, mid-October. 

Afranthidium (B.) guillarmodi (Mavromoustakis): 
Campanulaceae, Wahlenbergia, I9, northwest 
of Grahamstown, Eastern Cape, early-March. 

Afranthidium (B.) haplogastrum (Mavromousta- 
kis): Fabaceae (Papilionoideae), Indigofera and 
Lessertia, 1299 and 66*6\ three localities in 
Namaqualand and from Aus, Namibia, early- 
September to mid-October; Aizoaceae (Me- 
sembryanthema), Prenia, 299, Springbok, Na- 
maqualand, early October; Amaranthaceae, 
Hermbstaedtia, 19 and 1$, Richtersveld, late- 
September. 

Afranthidium (B.) ? joeosum Pasteels: Campanu- 
laceae, Wahlenbergia, 1 J, between Okahandja 
and Karibib, northwestern Namibia, late- 
March. 

Afranthidium (B.) matjesfonteinense (Mavromous- 
takis): Fabaceae (Papilionoideae), Indigofera, 
19 and 66*6*/ Richtersveld, Northern Cape, 
late-September. 

Afranthidium (B.) minutulum (Brauns): Fabaceae 
(Papilionoideae), mainly Indigofera, 699 and 
116*6*/ six sites from the Orange River north to 
Omaruru in northwestern Namibia, late- 
February to early-April; Aizoaceae (Mesem- 
bryanthema), 599 and 26*6*, Augrabies on the 



Volume 16, Number 1, 2007 



47 



Orange River north to Karibib in northwest- 
ern Namibia, late-February, mid-April, mid- 
October; Asteraceae, Geigeria and Osteosper- 
Diuiu (including Tripteris), 1799 and 166*6% 
Orange River north to Usakos in northwest- 
ern Namibia, mid-March and mid-October; 
Aizoaceae (non-Mesembryanthema), 599, 
26*6*/ Augrabies, Orange River, Northern 
Cape, northwestern Namibia, late-February 
to early-April; Apocynaceae, Asclepias, 1J, 
southeastern Namibia, mid-March; Malva- 
ceae, Hermannia, 19, 16\ Gibeon, southern 
Namibia, Bullsport, western central Namibia, 
late-March, mid-April; Vahliaceae, Vahlia, 39, 
lo\ southeastern Namibia, early-March; Zy- 
gophyllaceae, Zygophyllum, 29, 16\ Richters- 
veld and northwestern Namibia, early-Octo- 
ber, late March to early April. 

Afranthidium (Capanthidium) Pasteels - has 
a disjunct distribution, southern Africa (Cape 
Province, South Africa and Namibia) and 
western Mediterranean (Morocco and Spain) 
(Michener 2000). Ten species were recognized 
from southern Africa by Pasteels (1984). 
There appear to be no records of nesting for 
this sub-genus. 

Afranthidium (C.) capicola (Brauns): Asteraceae, 
H99, 86*6\ Oudtshoorn in the Little Karoo and 
Karoo Poort near Ceres, Clanwilliam and be- 
tween Clanwilliam and Klawer in the Olifants 
River Valley, early December, late October. 

Afranthidium (Domanthidium) Pasteels, mono- 
specific, found in Namibia and South Africa, 
in the "Cape Province", probably also Natal 
(Michener 2000). 

Afranthidium (D.) abdominale: Malvaceae (for- 
merly Sterculiaceae), Hermannia, 16 between 
Murraysberg and Hutchinson, central Nama 
Karoo, South Africa, late-October. 

Afranthidium (Immanthidium) Pasteels is wide- 
spread in eastern Africa from Sudan to South 
Africa, Natal west to "Cape Province", and 
Namibia (Michener 2000). 

Afranthidium (1.) immaculatum (Smith): Astera- 
ceae, Senecio, 1 specimen, Bastervoetpad [near 
Barkly East], Eastern Cape, South Africa ; 
Scrophulariaceae, Diascia capsularis and Dia- 
scia fetacaniensis, 2 specimens, Mt Kemp and 
Zuurberg Pass, Eastern Cape (col. V.B. White- 
head, det. C.D.Michener, specimens in South 
African Museum, Cape Town) 

Afranthidium (I.) junodi (Friese): Fabaceae (Papi- 
lionoideae), Melolobium, Aspalathus and "lu- 



cerne", 499, 16*6\ northeast Grahamstown, 
Eastern Cape, Clanwilliam, Olifants River 
Valley, Western Cape, Augrabies, Orange 
River, Northern Cape, late September, mid- 
October, late February; Lamiaceae, Ballota, 
299, Kamiesberg, Namaqualand, Northern 
Cape, early-October; Asteraceae, Berkhei/a, 
19, Eastern Cape, late-September; Boragina- 
ceae, Trichodesma, I9, Richtersveld, Northern 
Cape, late-September. 

Afranthidium (I.) repetitum (Schulz): Lamiaceae, 
1 J, Richtersveld, Northern Cape, late-Sep- 
tember. 

Afranthidium (I.) sjoestedti (Friese): Lamiaceae, 
Ballota, 3oo, southeastern Namaqualand and 
Richtersveld, Northern Cape, late-September; 
Boraginaceae, Anchusa, 1 j, Grahamstown, 
Eastern Cape, late September. 

Afranthidium (Nigranthidium) Pasteels - found 
in Namibia and in "Cape Province", South 
Africa (Michener 2000). There are two named 
species and a third undescribed species 
(Michener 2000). 

Afranthidium (Nigranthidium) concolor (Friese): 
Aizoaceae (Mesembryanthema), Herrea, 299, 
Nieuwoudtville, southeastern Namaqualand, 
late-September; Asteraceae, Pentzia, Senecio, 
399, southeastern Namaqualand, northern 
Namaqualand, Northern Cape, September. 

Afranthidium (Oranthidium) Pasteels - repre- 
sented by five or six species, has been found 
in Namibia and in South Africa, "Cape 
Province" east to the "Transvaal" (Michener 
2000). 

Afranthidium (O.) folliculosum (Buysson): Faba- 
ceae (Papilionoideae), Indigofera, I9, Kalahari 
fringe, southeastern Namibia, late-March. 

Afranthidium (O.) sp. Gess 1: Fabaceae (Papilio- 
noideae), Indigofera, 399, Kalahari fringe, 
southeastern Namibia, late March; Malva- 
ceae, Hermannia, 499 and 11 >j, Kalahari 
fringe, southeastern Namibia, Northern 
Cape, early-March, early-April; Apiaceae, 
Deverra, I9, 16\ southern Kalahari, early- 
March; Molluginaceae, Liineuiu, 299, southern 
Kalahari, Northern Cape, early March. 

Afranthidium (O.) sp. Gess 2: Fabaceae (Papilio- 
noideae), Indigofera, 1 ;, east of Oranjemund, 
northern bank of the Orange River, south- 
western Namibia, late-September; Malvaceae, 
Hermannia, 899, Springbok and Kamiesberg, 
Namaqualand, Northern Cape, late Septem- 
ber, early-October. 



48 



Journal of Hymenoptera Research 



Anthidiellum Cockerell - five sub-genera are 
recognized (Michener 2000). Of these only 
two, Chloranthidiellum Mavromoustakis and 
Pycnanthidiellum Krombein, have been re- 
corded from Africa south of the Sahara. 

Anthidiellum near melanocephalum (syn. of apica- 
tum Smith): Euphorbiaceae, Dalechampia vo- 
lubilis E. Mey. ex Baill., False Bay, northern 
KwaZulu-Natal (Armbruster and Steiner 
1992). 

Anthidiellum (Pycnanthidiellum) spilotum (Cock- 
erell): Malvaceae (formerly Tiliaceae), Grewia 
occidentalis L., l.j 1 , northeast of Grahamstown, 
Eastern Cape, early December. 

Anthidioma Pasteels - known from Namibia 
and the Western Cape, South Africa from two 
species, one undescribed (Michener 2000). 

Anthidioma sp. : Fabaceae (Papilionoideae), 
"yellow legume", one specimen, from Po- 
mona, Sperrgebiet, southwestern Namibia 
(col. V. Whitehead, det. C.D.Michener, spec- 
imen in South African Museum, Cape Town.) 

Anthidium Michener - found on all continents 
except Australia; rather poorly represented in 
sub-Saharan Africa (Michener 2000). 

Anthidium (Anthidium) Fabricius - found 
throughout the range of the genus; represent- 
ed in Africa by only a few species (Michener 
2000). 

Anthidium (A.) pontis (Cockerell): Acanthaceae, 
Blepharis capensis (L. f.) Pers., lo, northeast 
Grahamstown, Eastern Cape, early January. 

Anthidium (Niv anthidium) Pasteels - known 
from only one species; apparently previously 
recorded only from eastern Africa (Mozam- 
bique and Malawi) (Michener 2000). 

Anthidium (N.) niveocinctum Gerstaecker: Aster- 
aceae, Geigeria, 2 99, Oshivelo, Namibia, 
early-April (D.W. & G.T. Gess). 

Anthidium (Severanthidium) Pasteels - known 
from the Eastern Cape north to Senegal and 
the Arabian Peninsula (Michener 2000). 

Anthidium (Severanthidium) soni Mavromousta- 
kis: Fabaceae (Papilionoideae), Crotalaria, 1 J, 
Kamanjab, Namibia, late-March. 

Aspidosmia Brauns, represented by two species 
only - known solely from Namibia and the 
"Cape Province" (Michener 2000). 

Aspidosmia arnoldi (Brauns): Fabaceae (Papilio- 
noideae), Lebeckia and Wiborgia, 399, 3;;, 
four sites from Clanwilliam to Springbok, 
early-late-September; Lamiaceae, Stachys, 19, 
Richtersveld, Northern Cape, late September. 



Aspidosmia volkmanni (Friese): Asteraceae, Ber- 
kheya, Gorteria and Osteospermum, I899, 336*6*/ 
four sites from Kamiesberg to north of the 
Orange River, late-September, early-October; 
Amaranthaceae, Hermbstaedtia, 299, Richters- 
veld, Northern Cape, late-September. 

Cyphanthidium Pasteels - known from Zimbabwe, 
Namibia and the "Cape Province" from three 
species, two described (Michener 2000). 

Cyphanthidium intermedium Pasteels: Fabaceae 
(Papilionoideae), Crotalaria and Indigofera, 19, 
2.$S, three sites in western Namibia, Rosh 
Pinah/Sendelingsdrif (south), Uis/Henties 
Bay and Khorixas (north), mid-October 
(south), early April (north). 

Cyphanthidium sp.: Acanthaceae, Blepharis, 399, 
3J j, between Springbok and Kamieskroon in 
Namaqualand, October. 

Eoanthidium Popov - known from Africa, the 
Middle East, the Indian Peninsula and Asia 
Michener 2000); divided into four sub-genera, 
two of which, Clistanthidium Michener and 
Griswold and Eoanthidium Popov are repre- 
sented in Africa south of the Sahara but only 
the former from southern Africa (Michener 
2000). 

Eoanthidium. (Clistanthidium) turnericum (Mavro- 
moustakis): Fabaceae (Papilionoideae), Indigo- 
fera, "a yellow flowered papilionate shrub", 
19, 2cTcT/ Richtersveld, Northern Cape, be- 
tween Rosh Pinah and Sendelingsdrif, south- 
western Namibia, late-September, mid-Octo- 
ber; Acanthaceae, Moneehma and Petal idi 11 in, 
66*6*/ Richtersveld, Northern Cape, between 
Uis and Plenties Bay, northwesternNamibia, 
late-September (south), early-April (north); 
Boraginaceae, Heliotropium, 19, between Uis 
and Omaruru, northwestern Namibia, March; 
Boraginaceae (formerly Hydrophyllaceae), Co- 
don, 399, 36*6*/ Richtersveld, Northern Cape, 
between Uis and Plenties Bay, northwestern 
Namibia, late September (south), early April 
(north); Brassicaceae (formerly Capparaceae), 
Cleome, I9, 1 J, between Bullsport and Sesriem, 
westcentral Namibia, mid-April. 

P achy anthidium Friese - known from Africa 
east to China (Michener 2000); four sub- 
genera recognized; all present in Africa south 
of the Sahara; only three recorded from 
southern Africa (Michener 2000). 

P 'achy anthidium (Aus anthidium) Pasteels, 
known from a single species found in 
Namibia. 



Volume 16, Number 1, 2007 



49 



Pachyanthidium (A.) ausense (Mavromoustakis): 
Fabaceae (Papilionoideae), Indigofera, lo, 86*6*/ 
east of Oranjemund, southwestern Namibia, 
east of Alexander Bay, Richtersveld, North- 
ern Cape, late September; Amaranthaceae, 
Hermbstaedtia, 26*6*/ Richtersveld, late Sep- 
tember; Boraginaceae, Trichodesmia, 2 ] ], 
Richtersveld, late September; Zygophylla- 
ceae, Zygophyllum, 299, 56*6*/ Richtersveld, 
between Palm and Khorixas in northwestern 
Namibia, late-September (south), early-April 
(north); Loasaceae, Kissenia, 36*6*/ between 
Keetmanshoop and Aus in southwestern 
Namibia, between Uis and Henties Bay, 
northwestern Namibia, early March (south), 
early April (north); Malvaceae (formerly 
Sterculiaceae), Hermannia, 2 J 1 J 1 , southwest of 
Aus, southwestern Namibia, September. 

Pachyanthidium (Pachyanthidium) Friese - 
widespread in Africa, from Senegal to Ethio- 
pia and south to KwaZulu-Natal and the 
Cape Province, South Africa; 11 species 
(Michener 2000). 

Pachyanthidium (P.) cordatum (Smith): Fabaceae 
(Papilionoideae), Psoralen pinnata L., 19, near 
Grahamstown, Eastern Cape (Jacot Guillar- 
mod label data); Euphorbiaceae, Dalechampia 
sp., northern KwaZulu-Natal (Armbruster 
and Steiner 1992). 

Pachyanthidium (Triclianthidium) Cockerell - 
known in Africa from the Ivory Coast to 
southern Egypt, south to Angola and Kwa- 
Zulu Natal, South Africa, and in Asia from 
India to Yunnan Province, China; " at least 
three species" (Michener 2000). 

Pachyanthidium (T.) benguelense (Vachal): Fabaceae 
(Papilionoideae), Aspalathus, 19, Graafwater, 
west of the Olifants River, Western Cape, late 
September; Asteraceae, Senecio, lo, near Gra- 
hamstown, Eastern Cape, late December. 

Plcsianthidium Cameron, known from South 
Africa only, principally from the west; con- 
sisting of four sub-genera, Carinanthidium 
Pasteels, Plesianthidium Cameron, Spinanthi- 
diellum Pasteels, and Spinanthidium Mavro- 
moustakis (Michener 2000). 

Plesianthidium (Carinanthidium) Pasteels, rep- 
resented by a single species found in the 
Western Cape, but the type specimen was 
reported to be from the "northern Transvaal" 
(Michener 2000). 

Plesianthidium (C.) cariniventre (Friese): Fabaceae 
(Papilionoideae), Aspalathus and Lebeckia, 49, 



196*6*/ Olifants River Valley, Western Cape, 
Namaqualand, Northern Cape, late-Septem- 
ber, early-October; Polygalaceae, Polygala, I9, 
Namaqualand, late-October; Asteraceae, Pter- 
onia, I9, Namaqualand, early-October; La- 
miaceae, Balotta, I9, Namaqualand, early- 
October; Zygophyllaceae, Zygophyllum, 36*6*/ 
Namaqualand, early-October; Asphodela- 
ceae, Albuca, 1 j, Namaqualand, late Septem- 
ber. 

Plesianthidium (Spinanthidiellum) Pasteels - 
known from two species from the Western 
Cape Province of South Africa (Michener 2000). 

Plesianthidium (S.) volkmanni (Friese): Fabaceae 
(Papilionoideae), mostly Aspalathus and Le- 
beckia, 3499, 686*6\ Western Cape, Namaqua- 
land, Northern Cape, early-September to late- 
October; Malvaceae (formerly Sterculiaceae), 
Hermamiia, I9, southeastern Namaqualand, 
late-September; Zygophyllaceae, Zygophyl- 
lum, I9, 46*6*/ southeastern Namaqualand, 
late-September; Lamiaceae, Stachys, 36*6*/ 
southeastern Namaqualand, late-September; 
Aizoaceae (Mesembryanthema), Herrea, 2 3 j , 
southeastern Namaqualand, late-September; 
Polygalaceae, Polygala, 26*6*/ Namaqualand, 
late-September, late-October. 

Plesianthidium (Spinanthidium) Mavromousta- 
kis - known from five species all from the 
"Cape Province", South Africa (Michener 
2000). 

Plesianthidium (S.) bruneipes (Friese): Fabaceae 
(Papilionoideae), Lebeckia, 299, \3, Namaqua- 
land, Northern Cape, early-September; Ai- 
zoaceae (Mesembryanthema), Herrea, I9, near 
Springbok, Namaqualand, early-October; La- 
miaceae, Ballota and Stachys, 36*6*/ near 
Springbok, early-October; Malvaceae (for- 
merly Sterculiaceae), Hermannia, 1 $, Kamies- 
berg, Namaqualand, early-October. 

Plesianthidium (S.) callescens (Cockerell): Aster- 
aceae, Arctotheca and Pteronia, 19, lj, Nama- 
qualand, Northern Cape, late-September; 
Lamiaceae, Ballota, 29, 4j, southeastern Na- 
maqualand, late-September; Malvaceae (for- 
merly Sterculiaceae), Hermamiia, five species, 
299, 156*6*/ Namaqualand, Ratelfontein, West- 
ern Cape, late-September, early-October. 

Plesianthidium (S.) neli (Brauns): Fabaceae (Papi- 
lionoideae), Aspalatlms, Lebeckia and Melolo- 
biitiu, 5q, 136*6*/ western Western Cape, 
Namaqualand, early-September, early-Octo- 
ber; Aizoaceae (Mesembryanthema), Prenia, 



50 



Journal of Hymenoptera Research 



266, near Ratelfontein, western Western 
Cape, early-October; Boraginaceae, Anchusa, 
1<$, Kamiesberg, Namaqualand, early Octo- 
ber; Malvaceae (formerly Sterculiaceae), Her- 
mannia 19, 233, Namaqualand, late-Septem- 
ber, early-October; Iridaceae, Homeria, 19, 
vicinity of Springbok, Namaqualand, early- 
October. 

Plesianthidium (S.) trachusiforme (Friese): Faba- 
ceae (Papilionoideae), Aspalathus, Lcbeckia, 
Wiborgia, Melolobium and Indigofera, 699, 
136*6*, Olifants River Valley and westwards, 
Western Cape, Namaqualand, Northern 
Cape, early-September to mid-October; Ai- 
zoaceae (Mesembryanthema), Herrea, 299, 
Springbok, Namaqualand, early-September, 
early-October; Boraginaceae, Anchusa, 399, 
26*6*/ Kamiesberg, Namaqualand, early-Octo- 
ber; Brassicaceae, Heliophila, I9, Kamiesberg, 
early-September; Lamiaceae, Ballota, 233, 
Kamiesberg, early-October; Malvaceae (for- 
merly Sterculiaceae), Hermannia, 1 j, Nieu- 
woudtville, southeastern Namaqualand, late- 
September. 

Pseudoanthidium (Tuberanthidium) Pasteels, 
four species are recognized - known from 
Tanzania, Botswana, Namibia, Lesotho, and 
South Africa, "Cape Province to Natal" 
(Michener 2000). 

Pseudoanthidium (T.) tuberculiferum (Brauns): As- 
teraceae, Berkhei/a and Pteronia, 599, 46*6*/ 
Nieuwoudtville, southeastern Namaqualand, 
Springbok, Namaqualand, Northern Cape, 
northwest of Aus, southwestern Namibia, 
late-September; Acanthaceae, Monechma, 16, 
Richtersveld, Northern Cape; Lamiaceae, Bal- 
lota and "labiate", 56*6*/ Nieuwoudtville, Ka- 
miesberg, Richtersveld, Namaqualand, early- 
late-September, Acanthaceae, Monechma, 1 3, 
Richtersveld, Northern Cape, late September. 

Pseudanthidium (T.) damarensis (Mavromousta- 
kis): Asteraceae, Giegeria, 19, 26*6/ between 
Oshivelo and Ondongwa, northern Namibia, 
early-April (D.W. and G.T. Gess). 

Serapista Cockerell - known from the Afrotro- 
pial Region, two of the four species repre- 
sented in southern Africa - widespread S. 
denticulata (Smith) recorded from central, 
eastern and southern Africa and more re- 
stricted S. rufipes (Friese) recorded only from 
southern Africa (Pasteels, 1984). 



Serapista rufipes: Fabaceae (Papilionoideae), Cro- 
talaria, Indigofera, Lebeckia and Aspalathus, 
2199, 166*6"/ 14 sites in Namibia, Nababeep, 
Namaqualand, Northern Cape, four sites in 
Olifants River Valley and vicinity, Western 
Cape, early-March to early-April (north), 
mid-October (south); Malvaceae (formerly 
Sterculiaceae), Hermannia, 999, 66*6*/ six sites 
in Namibia, two sites in Namaqualand, early- 
March to late-April (north), early-September 
to early-October (south); Acanthaceae, Mon- 
echma, 19, northwest Namibia, late-April; 
Brassicaceae (formerly Capparaceae), Cleome, 
16, northwest Namibia, late- April; Scrophu- 
lariaceae, Jamesbrittenia, I9, westcentral Na- 
mibia, mid-April; Apocynaceae, Asclepias, I9, 
1 ;, southern Karoo, Western Cape, late- 
November. 

Tracliusa Panzer - known from Africa, Asia, 
the Mediterranean and North America. The 
sub-genus Massauthidium Pasteels is known 
from four species - three described by 
Pasteels, from Kenya and Eritrea and one 
recorded from a single specimen from 
Namibia by Michener (2000). In the course 
of their survey of flowers visited by aculeate 
wasps and bees in the semi-arid to arid 
areas of southern Africa the authors encoun- 
tered an undescribed species in northwest- 
ern Namibia, most probably the species 
known to Michener from a single specimen. 
Eighteen females and five males were 
collected. 

Tracliusa (Massauthidium) sp. undescribed: Fa- 
baceae (Papilionoideae), Crotalnria and Indi- 
gofera, 29, 26*/ west of Palm in northwestern 
Namibia, west-northwest of Omatjete in 
northwestern Namibia, late-March; Fabaceae 
(Caesalpinoideae), Adeuolobus, 499, 13, be- 
tween Palm and Khorixas and a drainage 
channel between Gaub and Kuiseb passes, 
northwestern Namibia, mid-late-March; Ped- 
aliaceae, Sesamum, 699, south of Swartboois- 
drif, Kunene River, northwestern Namibia, 
late-March; Brassicaceae (formerly Cappara- 
ceae), Cleome, south of Swartbooisdrif, late- 
March; Acanthaceae, Monechma, 599, 1 3, Two 
Palms, Palmwag, northwestern Namibia, 
late-March; Lamiaceae, Hemizygia, 16, south 
of Palmwag, late-March. 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 51-146 

Diversity, Classification and Higher Relationships of 
Mymarommatoidea (Hymenoptera) 

Gary A. P. Gibson 1 , Jennifer Read and John T. Huber 

(GAP) Agriculture and Agri-Food Canada, Biodiversity and Integrated Pest Management, K. W. 

Neatby Bldg., 960 Carling Avenue, Ottawa, Ontario, Canada, K1A 0C6; email: gibsong@agr.gc.ca 

(JR) Agriculture and Agri-Food Canada, Biodiversity and Integrated Pest Management, K. W. 

Neatby Bldg., 960 Carling Avenue, Ottawa, Ontario, Canada, K1A 0C6; email: readj@agr.gc.ca 

(JTH) Canadian Forestry Service, Natural Resources Canada, c/o ECORC, K. W. Neatby Bldg., 960 

Carling Avenue, Ottawa, Ontario, Canada, K1A 0C6; email: huberjh@agr.gc.ca 



Abstract. — The supraspecific, extinct and extant fauna of Mymarommatoidea (Hymenoptera) is 
revised. Ten extinct and ten extant described species are classified in six genera and two families, 
Mymarommatidae and Gallorommatidae n. fam. A key to the families and genera is provided. 
Classified in Gallorommatidae is the extinct Cretaceous genus, Galloromma Schliiter, including the 
type species, G. bezonnaisensis Schliiter, and G. agapa (Kozlov and Rasnitsyn) n. comb, (from 
Palaeomymar). Classified in Mymarommatidae is one extinct Cretaceous genus, Archaeromma 
Yoshimoto, one extinct Tertiary genus, Palaeomymar Meunier, and three extant genera, Mymaromma 
Girault, Mymaromclla Girault, and Zcalaromma n. gen. Mymaromma and Mymaromella rev. stat. are 
resurrected from prior synonymy under Palaeomymar and a neotype male is designated for 
Palaeomymar succini Meunier. Protooctonus Yoshimoto is transferred from Mymaridae to 
Mymarommatidae and is newly synonymised under Archaeromma. Newly classified in Archaeromma 
are the Cretaceous species A. masneri (Yoshimoto) n. comb, (from Protooctonus), and A. 
mandibulatum (Kozlov and Rasnitsyn) n. comb., A. senonicum (Kozlov and Rasnitsyn) n. comb. 
and A. japonicum (Fursov, Shirota, Nomiya and Yamagishi) n. comb, (all from Palaeomymar). 
Classified in Mymaromma are M. anomalum (Blood and Kryger) rev. comb., M. buyckxi Mathot rev. 
comb., M. goethei Girault rev. comb., M. mirissimum (Girault) n. comb, and M. ypt (Triapitsyn and 
Berezovskiy) n. comb, (all from Palaeomymar). Classified in Mymaromella is the extinct Tertiary 
species, M. duerrenfeldi (Schliiter and Kohring) n. comb., and the extant species M. chaoi (Lin) n. 
comb., M. cyclopterus (Fidalgo and De Santis) n. comb, and M. mira Girault rev. comb, (all from 
Palaeomymar). Classified in Zealawmma are the newly designated type species of the genus, Z. 
insulare (Valentine) n. comb, (from Palaeomymar) and Z. valentinei n. sp. Description of the structural 
diversity of extant and extinct mymarommatids is based on the 20 described species plus 15 
undescribed extant morphospecies and several fossils from Tertiary Baltic amber and Burmese, 
Canadian and New Jersey Cretaceous amber. Evidence for monophyly of the genera is presented 
and the following phylogenetic relationships are hypothesized: Galloromma + {Archaeromma + 
(Zealawmma + (Mymaromella + (Palaeomymar + Mymaromma)))). Absence of mesotibial and metatibial 
spurs are newly proposed synapomorphies for Mymarommatoidea. New information is given on 
the structure of Serphitidae (Serphitoidea) based on study of several Taimyr and Canadian 
Cretaceous fossils, and structural features that are shared among Serphitoidea, Mymarommatoidea, 
Chalcidoidea and Platygastroidea are discussed relative to establishing their relationships. Several 
features, including gastral laterotergites, a mesopectal region similar to a netrion, and a forewing 
venation that could be ancestral to that of Platygastroidea suggest Serphitoidea is closely related to 
Platygastroidea. No new evidence was found to support a Serphitoidea + Mymarommatoidea 
sister-group relationship. Independent parameres in the groundplan of the male genitalia of 
Mymarommatoidea and Serphitoidea is a likely symplesiomorphy that differentiates them from 
Chalcidoidea and Platygastroidea. Two different types of specialized claval sensilla could support 



Corresponding author (e-mail: gibsong@agr.gc.ca) 



„ Journal of Hymenoptera Research 

monophyly of Mymarommatoidea + Chalcidoidea, but further study throughout parasitic 
Hymenoptera is necessary to substantiate character-state distribution and homology. 



The superfamily Mymarommatoidea sphere had been collected in shady and 

(Hymenoptera) has been referred to as relative moist areas such as deciduous 

"arguably the most enigmatic wasp taxon" forests. Clouatre et al. (1989) did extract 

(Vilhelmsen and Krogmann 2006, p. 290). specimens from forest litter in eastern 

Individuals are among the smallest of Canada, but they were also extracted from 

microhymenoptera, only about 0.3-0.8 mm vegetation litter samples on three subant- 

in body length, but Mymarommatidae is arctic islands (Valentine 1971). Based on 

one of the easiest families of Hymenoptera sweep samples, Kryger (cited in Bakkendorf 

to recognize because of several highly 1948, p. 216) suggested that mymaromma- 

distinctive features. Most conspicuously, tids are associated with low vegetation and 

the head has a hyperoccipital band of remarked on their "very slow-moving gait 

pleated membrane that enables the occipital — as an old man tired to death". 

region to expand and contract in a bellows- Partly because of the lack of comprehen- 

like manner, the forewing membrane has sive comparative studies, Mymarommati- 

a mesh-like pattern, the hind wing is dae has a complex nomenclatural history 

reduced to an apically bifurcate haltere-like and uncertain phylogenetic relationships 

structure, and the petiole is composed of within Apocrita. Gibson (1986) and Vil- 

two tubular segments (Gibson 1986, Gibson helmsen and Krogmann (2006) postulated 

et al. 1999, Vilhelmsen and Krogmann several autapomorphies to support mono- 

2006). Partly because of their minute size, phyly of the group, but these hypotheses 

mymarommatids are rarely collected and and other character-state knowledge are 

are poorly represented in most collections, based primarily on a single European 

but they have been captured on several species whose morphology has been stud- 

subantarctic and Pacific islands (Valentine ied in detail (Debauche 1948, Vilhelmsen 

1971, Beardsley et al. 2000) and on all and Krogmann 2006). Knowledge of other 

continents north into Canada (Clouatre et mymarommatid species and genera is 

al. 1989), Scandinavia (Hansen 1997) and far limited largely to original descriptions. 

eastern Russia (Triapitsyn and Berezovskiy All 9 previously described extant species 

2006). Specimens are also known in Domin- and 6 of the 10 extinct species are currently 

ican, Sicilian, Baltic, Canadian, Japanese, classified in Palaeomymar Meunier, 1901, 

Taimyr, New Jersey, Burmese, French, which was established for a species in 

Spanish and Lebanese amber, indicating Baltic amber. Kozlov and Rasnitsyn (1979) 

the group has been present for at least 120 stated that the diversity of the fossils they 

million years (Grimaldi and Engel 2005) knew from the Cretaceous extended be- 

and has long had a world distribution, yond the limits of a single genus, but that it 

Despite their long and apparently ubiqui- did not seem possible to introduce any 

tous presence, almost nothing is known of clear generic classification without an 

their biology. Yoshimoto (1984) suggested analysis of all accumulated material. Gib- 

that they are egg parasitoids, but the life son et al. (1999) also suggested that the 

stage they attack and their hosts remain to extant species could be classified in two 

be discovered. A single individual was genera based on differences in foretibial 

reared from a bracket fungus (Gibson spur and forewing structure. 

1993) and Huber (1987) noted that most The primary purpose of our study is to 

specimens captured in the Northern Hemi- describe and illustrate the range of mor- 



Volume 16, Number 1, 2007 53 

phological variability encompassed by the idae) and established the new genus and 

extant and extinct mymarommatoid fauna, species, Palaeomymar succini Meunier, 

This is necessary to classify the described based on five males of Duisburg's material. 

species in higher taxa within a phylogenetic Unaware of either Stein (1877) or Meunier 

perspective and to determine what features (1901), Bakkendorf (1948) rediscovered 

are of specific or generic value so that these Duisburg (1868) and concluded that the 

are included in future descriptions. Our illustration seemed to be identical with 

study is also intended to provide the what Blood and Kryger (1936) had illus- 

accurate morphological data for Mymar- trated as the female of Petiolaria anomala. 

ommatoidea necessary for reliable phylo- Doutt (1973) later synonymised Mymar- 

genetic analyses of Hymenoptera. We do omma, Mymaromella and Petiolaria under 

not attempt to resolve the higher relation- Palaeomymar, though he incorrectly cited 

ships of Mymarommatoidea within Apoc- M. duisburgi as the type species of Palaeo- 

rita because our study of other groups is mymar. Petiolaria had previously been 

insufficient for reliable hypotheses of char- synonymised under Mymaromma by Gir- 

acter state homology and distribution, ault (1930) and Mymaromella under Mymar- 

However, we did examine other parasitic omma by Annecke and Doutt (1961). Ever 

Hymenoptera, particularly those groups since Doutt (1973), Mymaromma, Mymaro- 

that have been postulated as closely related mella and Petiolaria have all been consid- 

to mymarommatids, to investigate shared ered as synonyms of Palaeomymar. Based 

features. We discuss our observations so on our study, we recognize Palaeomymar 

that the features can be examined more only for P. succini and classify all the extant 

comprehensively for these and other Hy- species in Mymaromma, Mymaromella, and 

menoptera prior to comprehensive phylo- Zealaromma n. gen. This classification and 

genetic analyses. our newly proposed nomenclatural acts are 

summarized in Table 1. 

HISTORICAL REVIEW In addition to p sucani and the extant 

Taxonomy. — Three genera have been es- species, the following extinct species have 

tablished in Mymarommatidae for extant been classified in Palaeomymar: P. duerren- 

species: Mymaromma Girault (1920), Petio- feldi Schluter and Kohring (1990), P. japo- 

laria Blood and Kryger (1922), and Mymar- nicum Fursov et al. (2002), and P. agapa, P. 

omella Girault (1931). These authors were mandibulatus and P. senonicus Kozlov and 

unaware that Meunier (1901) had estab- Rasnitsyn (1979). These species were de- 

lished Palaeomymar for a mymarommatid scribed from Tertiary and Cretaceous 

in Baltic amber that Duisburg (1868) amber deposits spanning about 5-100 

illustrated and discussed but did not name, mya (Table 1). The other four described 

Although Duisburg did not formally name extinct species are in three different genera, 

the species, he gave reasons why he Yoshimoto (1975) established Archaeromma 

believed it probably belonged to Mi/mar for Ooctonus minutissimus Brues, 1937, and 

Curtis (Mymaridae). Prior to Meunier, his new species, A. nearcticum, from Cana- 

Stein (1877) had discovered a female in dian Cretaceous amber. He also described 

Baltic amber that he thought was the same Protooctonus masneri from the same materi- 

species as the female illustrated by Duis- al and assigned this taxon to Mymaridae 

burg and named it Mymar duisburgi. (Chalcidoidea), but we consider that it is 

Meunier (1901) subsequently examined a mymarommatid (see below). Finally, 

Duisburg's original amber material and Galloromma bezonnaisensis Schluter (1978) 

realized that Stein's interpretation of Duis- was established for a specimen from 

burg's species was incorrect. He reassigned French Cretaceous amber. Based on our 

M. duisburgi to Eustochus Haliday (Mymar- study, we classify all the Cretaceous 



54 



Journal of Hymenoptera Research 



Table 1 Proposed classification of Mymarommatoidea including temporal and spatial distribution of type 
material. Abbreviations: E = extant, K = Cretaceous, T = Tertiary; ages based on Grimaldi and Engel (2005). 



Classification proposed 



Original genus 



Temporal and spatial distribution 



* Gallorommatidae n. fam. 
Galloromma Schliiter, 1978 

agapa (Kozlov & Rasnitsyn, 1979) n. comb. 

bezonnaisensis Schliiter, 1978 

Mymarommatidae 

1 Archaeromma Yoshimoto, 1975 

japonicum (Fursov, Shirota, Nomiya & 

Yamagishi, 2002) n. comb. 
mandibulatum (Kozlov & Rasnitsyn, 1979) n. comb. 

masneri (Yoshimoto, 1975) n. comb. 

minutissimum (Brues, 1937) 

nearcticum Yoshimoto, 1975 

senonicum (Kozlov & Rasnitsyn, 1979) n. comb. 

f Palaeomymar Meunier, 1901 
succiiii Meunier, 1901 

Mymaromma Girault, 1920 rev. stat. 

anomalum (Blood & Kryger, 1922) rev. comb. 

buyckxi Mathot, 1966 rev. comb. 

goethei Girault, 1920 rev. comb. 

mirissimum (Girault, 1935) n. comb. 

ypt (Triapitsyn & Berezovskiy, 2006) n. comb. 
Mymaromella Girault, 1931 rev. stat. 

chaoi (Lin, 1994) n. comb. 

cycloptents (Fidalgo & De Santis, 1882) n. comb. 
; duerrenfeldi (Schliiter & Kohring, 1990) n. comb. 

mira Girault, 1931 rev. comb. 
Zealaromma Gibson, Read & Huber n. gen. 

insulare (Valentine, 1971) n. comb. 

valentinei Gibson, Read & Huber n. sp. 



Palaeomymar 
Galloromma 

Palaeomymar 

Palaeomymar 

Protooctonus 

Yoshimoto, 1975 
Ooctonus Haliday, 

1833 

Archaeromma 

Palaeomymar 
Palaeomymar 



Petiolnria Blood & 

Kryger, 1922 
Mymaromma 
Mymaromma 
Mymaromella 
Palaeomymar 

Palaeomymar 
Palaeomymar 
Palaeomymar 

Mymaromella 

Mymaromma 
Zealaromma 



K: Taimyr 

[Cenomanian (95 mya)] 
K: French 

[Cenomanian (100 mya)] 



K: Japanese 

[Santonian (85 mya)] 
K: Taimyr 

[Santonian (85 mya)] 
K: Canadian 

[Campanian (75 mya)] 
K: Canadian 

[Campanian (75 mya)] 
K: Canadian 

[Campanian (75 mya)] 
K: Taimyr 

[Santonian (85 mya)] 

T: Baltic 

[Eocene (44 mya)] 

E: Britain 

E: Congo 
E: Australia 
E: Australia 
E: Far East Russia 

E: China 

E: Argentina 

T: Sicilian [Pliocene (5 mya)] 

E: Australia 

E: New Zealand 
E: New Zealand 



species in Archaeromma and Galloromma, 
and the Tertiary species either in Palaeo- 
mymar or extant genera (Table 1). 

Classification ami relationships. — Mymar- 
ommatids were included in Mymaridae 
prior to Debauche (1948), who established 
the family Mymarommidae (sic) after 
a comprehensive morphological compari- 
son of M. anomalum with several mymarid 
genera. Annecke and Doutt (1961) rejected 
Debauche's classification and regarded 



Mymaromma as merely an extremely aber- 
rant mymarid, placing it in the subfamily 
Mymarinae, tribe Ooctonini. Subsequently, 
some authors included mymarommatids in 
Mymaridae (e.g. Valentine 1971, Doutt 
1973), sometimes as their own subfamily 
(Yoshimoto 1975), whereas others treated 
them as a separate family in Chalcidoidea 
(e.g. Mathot 1966, Konigsmann 1978, 
Yoshimoto 1984). Without explanation, 
Nikol'skaya (1978) classified Palaeomymar 



Volume 16, Number 1, 2007 



55 



in the otherwise extinct Cretaceous family 
Serphitidae (Serphitoidea). Kozlov and 
Rasnitsyn (1979) subsequently provided 
rationale for this placement based on 
a single feature shared by the two taxa — 
a two-segmented petiole. They noted that 
Yoshimoto (1975) had also described the 
extinct Canadian Cretaceous genus Disty- 
lopus (Chalcidoidea: Tetracampidae) as 
having a two-segmented petiole, but they 
considered the original description was not 
sufficiently detailed to establish its system- 
atic position. Gibson (1986) later found that 
only a single segment formed the petiole of 
the unique specimen of the type species, D. 
bisegmentus Yoshimoto. He studied petiolar 
structure as one of 23 adult and larval 
characters throughout parasitic Hymenop- 
tera and proposed that mymarommatids 
constituted a monophyletic taxon based on 
four autapomorphies — head consisting of 
two plates connected by pleated mem- 
brane, hind wing stalk-like with hamuli 
forming a distal bifurcation, forewing with 
reticulate pattern, and mesotergal-mesotro- 
chanteral muscle with axillar portion ab- 
sent. He also proposed that Chalcidoidea, 
including Mymaridae, was monophyletic 
based on three autapomorphies and that 
Chalcidoidea and Mymarommatidae were 
sister-groups based on three synapomor- 
phies — unique presence of axillar phrag- 
mata as sites of origin for all or part of the 
mesotrochanteral depressor muscles, me- 
sotrochanteral depressor muscle without 
a mesofurcal or mesoscutal portion, and 
the male genitalia without an independent 
basal ring. The last similarity likely is 
homoplastic because Chiappini and Maz- 
zoni (2000) showed that the male genitalia 
of some Mymaridae have a distinct basal 
ring. Regardless, the proposed synapomor- 
phies for Chalcidoidea + Mymarommatoi- 
dea are all internal and likely never will be 
informative for testing relationships with 
Serphitidae because these are known only 
as amber fossils. 

Kozlov and Rasnitsyn (1979) suggested 
that their new genus Microserphitcs (Ser- 



phitidae) was intermediate between My- 
marommatidae and Serphitidae because 
the pronotum did not appear to extend to 
the base of the forewing, unlike in other 
serphitids. However, the single specimen 
constituting Microserphitcs is damaged (top 
of head and mesonotum not preserved), 
and the pronotum was described as ex- 
tending to the tegula in one of their three 
Cretaceous mymarommatid species. Gib- 
son (1986) suggested that the single unique 
feature shared between mymarommatids 
and serphitids, a 2-segmented petiole, 
likely was derived independently because 
otherwise members of the two taxa are so 
dissimilar. Rasnitsyn (1988) subsequently 
considered the ancestry of mymaromma- 
tids as uncertain, being questionably most 
closely related to either Chalcidoidea or 
Serphitidae. When Ronquist et al. (1999) 
reanalyzed Rasnitsyn's data using cladistic 
methods they recovered Mymarommati- 
dae as the sister group of Chalcidoidea and 
Serphitidae as either the sister group of 
Platygastroidea or within a clade that 
contained Platygastroidea. Rasnitsyn 
(2002) later included Mymarommatidae 
and Serphitidae as separate families in 
Platygastroidea, and these two families as 
the sister group of Chalcidoidea (Rasnitsyn 
2002, fig. 331). This classification more 
closely reflects the views of Ronquist et al. 
(1999), but renders Platygastroidea 
sensu Rasnitsyn (2002) paraphyletic. 
More recently, Rasnitsyn et al. (2004) 
included Mymarommatidae as one of two 
families of Serphitoidea, whereas 
Grimaldi and Engel (2005) treated the 
families as separate superfamilies. To date, 
mymarommatids have not been included 
in published molecular analyses of hyme- 
nopteran (Dowton and Austin 2001, 
Castro and Dowton 2006) or chalcid 
(Campbell et al. 2000) relationships. Re- 
gardless, molecular techniques cannot re- 
solve the question of whether mymarom- 
matids are more closely related to serphi- 
tids than to chalcids because serphitids are 
extinct. 



56 



Journal of Hymenoptera Research 



When Gibson (1986) proposed Mymar- 
ommatidae as the sister group of Chalci- 
doidea he left them unplaced to superfam- 
ily. He did this because he considered that 
mymarommatid relationships were ambig- 
uous and because if mymarommatids and 
chalcids were sister groups then mymar- 
ommatids could be included or excluded 
from "Chalcidoidea" depending on wheth- 
er internal or external features were used 
to define that taxon. Noyes and Valentine 
(1989) first treated mymarommatids as the 
superfamily Mymarommatoidea, as re- 
viewed by Gibson et al. (1999). The latter 
authors proposed two additional autapo- 
morphies for the group — mesopleuron, 
metapleuron and propodeum fused ven- 
tral to propodeal spiracle, and propleura 
and presternum fused into carapace below 
pronotum. Most recently, Vilhelmsen and 
Krogmann (2006) proposed four additional 
autapomorphies for Mymarommatoidea — 
absence of a mesothoracic spiracle, fusion 
of the propleural arm with the profurcal 
arm, presence of a pair of rods on the 
anterior surface of the prophragma, and 
absence of a metafurca. 

MATERIALS AND METHODS 

Sources of material. — This study was 
based on specimens obtained from the 
collections listed below. An asterisk indi- 
cates the collection included amber mate- 
rial. The names of individuals who facili- 
tated loans of specimens are given in 
parentheses. 

ANIC Australian National Insect Collec- 

tion, Canberra, ACT, Australia 
(John LaSalle, Nicole Fisher). 

AMNH* Division of Invertebrate Zoology, 
American Museum of Natural His- 
tory, New York, NY, USA (David 
Grimaldi). 

BMNH The Natural History Museum, 

London, England (John Noyes). 

BPBM Bernice P. Bishop Museum, De- 

partment of Entomology, Hono- 
lulu, HI, USA (John Beardsley). 



CIRAD Centre de cooperation internatio- 

nale en recherche agronomique 
pour le developpement (CIRAD), 
Montpellier, France (Gerard Del- 
vare). 
CNC* Canadian National Collection of 

Insects, Ottawa, ON, Canada. 

FAUF Biological Control Research Insti- 

tute, Fujian Agricultural Universi- 
ty, Fuzhou, Fujian Province, China 
(Naiquan Lin). 

GPPC* George Poinar Personal Collection, 

maintained at Oregon State Uni- 
versity, Corvallis, OR, USA 
(George Poinar). 

GZG* Geowissenschaftliches Zentrum 

der Universitat Gottingen, Muse- 
um, Gottingen, Germany (Mike 
Reich). 

ISNB Institut Royal des Sciences Natur- 

elles de Belgique, Brussels, Bel- 
gium (Paul Dessart). 

MCZ* Museum of Comparative Zoology, 

Cambridge, MA, USA (Philip Per- 
kins). 

MLPA Museo de la Plata, Facultad de 

Ciencias Naturales y Museo, Uni- 
versidad Nacional de La Plata, La 
Plata, Argentina (Marta Loiacono). 

NHRS* Naturhistoriska riksmuseet, Stock- 

holm, Sweden (Dave Karlsson). 

NZAC New Zealand Arthropod Collec- 

tion, Entomology Division, DSIR, 
Auckland, New Zealand (Jo Berry). 

PIN* Palaeontological Institute, Russian 

Academy of Sciences, Laboratory 
of Arthropods, Moscow, Russia 
(Alex Rasnitsyn). 

QMBA Queensland Museum, Queensland 

Cultural Centre, Brisbane, QLD, 
Australia (Chris Burwell). 

ROMT* Department of Natural History, 

Royal Ontario Museum, Toronto, 
ON, Canada (Janet Waddington). 

UCRC UCR Entomological Teaching and 

Research Collection, University of 
California, Riverside, CA, USA 
(Johan Liljeblad, Jeremiah George, 
Serguei Triapitsyn, John Heraty). 

USNM United States National Museum of 

Natural History, Smithsonian In- 
stitution, Washington, DC, USA 
(Michael Gates). 



Volume 16, Number 1, 2007 57 

ZMB* Institute of Systematic Zoology, We did not locate specimens of the 

Museum fur Naturkunde der apterous species that Valentine (1971) said 

Humboldt-Universitat zu Berlin, he had from ma i n land New Zealand, 

Berlin, Germany (Michael Ohl, whJch WQuld be important for assessing 

Thomas Schliiter). ,, cc , , 

^» ,i t^*. ^ . • . » , r r ■ ■ r the effect of wine reduction on mesosomal 

ZMUC* Zoological Museum, University of , , . ,, . . 

^ , „ , j\ morphology. Among the material exam- 
Copenhagen, Copenhagen, Den- r °/ ° 

mark (Lars Vilhelmsen). ined ' we distinguished 15 morphospecies 

in addition to the described species (Ap- 

Techniques. — The minute size of mymar- pendix II). Our study was intended to 
ommatids, the comparatively poor state of evaluate morphological diversity rather 
preservation of many amber fossils, and than formally name species. We therefore 
the inability to examine body parts of describe only one species that is important 
a specimen from all angles in any single for phylogenetic inference and for which 
amber inclusion prevented us from de- sufficient specimens are available to in- 
termining the exact structure and morpho- terpret structure confidently. Appendix II 
logical variation encompassed by Tertiary is given as an aid to locate the specimens 
and Cretaceous representatives to the same we examined and the morphospecies we 
extent as for extant taxa. For this reason, differentiated for future species descrip- 
less complete and comparable descriptions tions. Many of the specimens are mounted 
are provided for the extinct genera and such that features are not visible or directly 
a comprehensive description and compar- comparable because of the state of preser- 
ative and functional analysis of extant vation (air vs. critical-point dried), method 
mymarommatid structure is given prior of mounting (card vs. point) and/or one 
to discussing the extinct fauna. Character body part concealing another. Consequent- 
states observed in the fossil taxa are ly, an accurate appraisal of the distribution 
sometimes discussed in the comparative of some character states was not possible 
and functional analysis for the purpose of and we did not attempt to fully resolve the 
justifying character polarity hypotheses. species limits of Mymaromma. Based on 

The terminology of Basibuyuk and variation in intensity of the reticulate 

Quicke (1995) is used for the components pattern of the first petiolar segment, what 

of the foreleg antenna cleaner, whereas we interpret as Mymaromma sp. 7 may 

other terms for structure follow Gibson constitute a species complex. Specimens 

(1997) and Vilhelmsen and Krogmann from Taiwan have quite a strongly re- 

(2006). The abbreviations used to designate ticulate first petiolar segment and therefore 

structures in the illustrations are listed in are very similar to M. anomalum and M. ypt. 

Appendix I. In the text, figure numbers The specimens from Taiwan assigned to M. 

that precede an abbreviation designate sp. 7 were one of three samples of 

figures that have the abbreviation for the Mymaromma for which we had numerous 

relevant structure illustrated, whereas fig- ethanol-preserved individuals; the other 

ure numbers following the abbreviation two were M. anomalum from Sweden and 

illustrate the structure but do not have M. sp. 6 from Taiwan. Specimens of these 

these specifically indicated. Our descrip- three taxa were critical-point dried, dis- 

tion of the mesosoma is intended to sected, and gold coated for more detailed 

supplement the comprehensive study of anatomical study using a Philips XL30 

the internal and external anatomy of M. environmental scanning electron micro- 

anomalum by Vilhelmsen and Krogmann scope (SEM). A single female of Z. valenti- 

(2006) and does not repeat many, normally nei was also gold coated for observation, 

concealed, anatomical features they de- but specimens of the other species were left 

scribed. uncoated as originally mounted on card 



58 



Journal of Hymenoptera Research 



points or rectangles. In some instances, the 
electron beam caused the setae of uncoated 
specimens to bend (e.g., Figs 55, 71, 144c) 
so that some images do not accurately 
reflect setation. 

Amber inclusions were examined after 
the upper surface was covered with a thin 
film of glycerine and a cover slip. This was 
done in order to improve visibility through 
scratches and other minor surface irregu- 
larities. In order to obtain an optimal 
viewing angle, amber blocks were often 
positioned in a V-like glass well or inserted 
into a piece of plasticine at the desired 
angle before glycerine and a cover slip 
were added. The amber piece was washed 
subsequently in water to remove the 
glycerine. After one re-examination of the 
male paratype of P. agapa, the tiny amber 
shard containing the specimen was lost 
while it was being transferred to its storage 
vial by the senior author. The inclusions 
were examined with a Nikon SMZ1500 
binocular microscope using 15x oculars 
and a 1.6x HR Plan APO objective for 
a maximum magnification of 270x. When 
possible, they were examined also with 
a Nikon Optiphot compound microscope, 
usually at a magnification of 200x. The 
binocular microscope had a light base with 
a mirror for transmitted light, and a halo- 
gen spot light and a hand-held fibre optic 
ring light were used to obtain optimal 
lighting. Inclusions were photographed 
with a Leica DC500 digital camera attached 
to a Leica Z16 APO macroscope or a Nikon 
E800 compound microscope. The serial 
images obtained were combined with 
AutoMontage® and these and the scan- 
ning electron microphotographs were dig- 
itally retouched using Adobe Photoshop® 
to enhance clarity. 

In the original publications, the holo- 
types of Gnlloromma bezonnaiseusis Schliiter 
(1978) and Palaeoiin/nmr duerrenfeldi Schlii- 
ter and Kohring (1990) were stated as 
deposited in the "Institute fiir Palaontolo- 
gie, Freie Universitatat, Berlin". These 
types are now in ZMB. 



EXTANT FAUNA 

Description. — Body less than 1 mm in 
length. Body yellow to partly brown 
without metallic luster, often with a dark 
brown triangular region or band on meso- 
pleuron below base of forewing (Fig. 82: 
sa), and tarsal segments often with extreme 
apices brown; forewing disc sometimes 
more or less infuscate within basal half 
(Figs 166, 168). 

Head capsule: Head capsule composed 
of three parts, a strongly convex anterior or 
"frontal" plate (Figs 14: frp; 13, 30, 53), 
a flat, semicircular, posterior or "occipital 
plate" (Figs 14, 23: ocp; 41, 50), and 
a ventral "postgenal" plate (Figs 23: pgp; 
41, 50). Frontal plate separated dorsally 
and laterally from occipital plate by 
pleated membrane originating front above 
base of each mandible (Figs 13, 14, 53). 
Occipital plate articulating with postgenal 
plate along transverse margin above occip- 
ital foramen (Figs 23, 41, 50) and capable of 
rotating anteriorly into head capsule 
(Fig. 15) or posteriorly beyond vertex in 
a bellows-like manner (Figs 13, 42). Fron- 
tovertex transversely reticulate-scabrous to 
strigose and sparsely setose (Figs 25, 30, 

32, 33, 36, 46, 53). Face usually quite 
smooth, only very finely striate to strigose 
(Figs 25, 31, 33, 47, 48), though sometimes 
with conspicuous mesh-like sculpture 
(Figs 49, 52, 54); with two "subtorular" 
setae on midline immediately below toruli 
(insert, Figs 33, 52) plus 6-12 "interorbital" 
setae in region between eyes and oral 
margin (Figs 31, 33, 47-49, 52, 54); oral 
margin slightly reflexed medially (Figs 25, 

33, 56), but clypeus undifferentiated by 
sutures or evident anterior tentorial pits. 
Ocelli present (Figs 32, 36, 45, 53) or absent 
(Figs 25, 30, 33, 46). Toruli subcontiguous 
and slightly protuberant, at or above level 
of dorsal margin of eyes (Figs 25, 30, 33, 36). 
Gena bare except for long seta overlapping 
base of mandible (Figs 19, 29, 48); malar 
space variable in length, sometimes long 
(Figs 47, 52), particularly if eye with few 



Volume 16, Number 1, 2007 



59 



ommatidia (Fig. 47), but usually short to 
sublinear (Figs 25, 29, 31, 36, 49) and only 
rarely with distinct malar sulcus (Fig. 30). 
Eye variable in size, composed of about 5- 
55 comparatively large ommatidia (cf. 
Figs 33, 47). Occipital plate with sculpture 
mesh-like (Fig. 41) to more or less wrin- 
kled-rugulose (Fig. 14); bare except for 
long seta at extreme ventrolateral corner 
(Figs 16, 45, 46, 51) and sometimes with 
a pair of setae paramedially near center 
(Figs 45, 46: ms); rarely with a ventromedial 
pit above occipital foramen (Figs 45, 46: 
opp). Postgenal plate bare except for two 
long setae laterally in line with ventrolat- 
eral seta of occipital plate (Fig. 41), with 
region between occipital foramen and 
labiomaxillary complex sclerotized and 
smooth, the postgenae comparatively 
widely separated medially (Figs 23, 41, 
50). Occipital foramen near dorsal margin 
of postgenal plate, the orifice surrounded 
laterally and ventrally by U-shaped region 
divided by oblique lateral suture (Fig. 24), 
with single setiform sensillum dorsolater- 
ally, three setiform sensilla ventrolaterally, 
and with posterior tentorial pits (Fig. 19: 
ptp) usually visible immediately below 
ventrolateral sensilla. 

Mouthparts: Mandible bidentate or tri- 
dentate, exodont, their apices not meeting 
medially when closed (Figs 23, 25, 31, 33, 
49, 52), and with 3-5 setae on outer surface, 
including dorsal seta usually associated 
with a campaniform sensillum (Figs 27, 34: 
cs); bidentate mandible with more or less 
straight dorsal margin ending as acute 
dorsoapical angulation and with shorter, 
acute tooth near middle of ventral margin 
(Figs 34: vt; 47-49, 52, 54); tridentate 
mandible both with small ventral tooth 
(Figs 26, 27: vt) and variably distinct dorsal 
tooth (Fig. 27) or angulation (Fig. 26) ex- 
tending at least slightly above oral margin 
(Figs 29, 30: dt). Labrum (Figs 20-22: lbr) 
thin, flaplike, often convoluted apically, the 
exposed surfaces smooth (Figs 19-22) but 
with papillae directed ventrally from inner, 
ventral surface. Labiomaxillary complex in 



ventral view a more or less oval to tri- 
angular, flat plate having medial labium 
(Fig. 20: lab) separated from lateral maxil- 
lae (Fig. 20: max) over apical third to two- 
thirds (Figs 20, 41, 50, 55), and with medial, 
apparently socketed, papilliform process 
projecting (Figs 21, 22: pap) externally 
between labrum and labium. Externally 
visible part of labium undifferentiated, 
without palpi, but with pair of setae 
paramedially near its base (Figs 19-21, 
41, 41, 55). Externally visible part of maxilla 
with long seta laterally near presumptive 
base; distally differentiated into small sub- 
apical lobe (Fig. 41: mxs) bearing terminal 
spine (Fig. 41: mxp) and longer lateral lobe 
(Fig. 41: mxg), or with apically narrowed 
ventral lobe (Figs 21, 57: mxs) bearing 
spine (Figs 21, 22, 57: mxp) and with dorsal 
flap (Fig. 21: mxg) or lobe (Fig. 57: mxg) 
having short, distally projecting papillae 
(Figs 21, 57) and pustulate surface apically 
(Figs 20, 21, 22, 57: mxg). 

Antenna: Antenna geniculate; flagellum 
without multiporous plate sensilla; first 
flagellar segment often shorter than second 
segment but not anelliform (Figs 171-178). 
Female antenna 9-11 -segmented, distinctly 
clavate; funicle with 7 (Figs 173: fu; 178) or 
6 (Figs 174, 175) segments, the segments 
progressively more setose toward clava 
and with unmodified trichoid (hairlike) 
setae; clava composed of 1 (Figs 173: cl; 
174, 175, 178) or 2 (Figs 71, 172: cl) 
segments, and with at least four different 
types of sensilla, including trichoid setae 
on inner and outer surfaces similar to setae 
of funicle (Figs 58, 72: si), a few much 
longer and thicker, usually basally bent 
sensilla on dorsal or sometimes outer 
surface (Figs 63, 64, 70: s2), one or more 
rows of comparatively short and more or 
less sinuate sensilla on ventral surface 
(Figs 64, 72: s3), two or three basally bent 
and slightly lanceolate sensilla, each aris- 
ing from distinct depression, on outer 
surface near midline or more dorsally 
(Figs 62, 63, 70, 71: s4), and often a straight, 
spine-like sensillum projecting from apex 



60 



Journal of Hymenoptera Research 



of clava (Figs 64: as; 65, 67, 69). Male 
antenna rarely 12-segmented (Fig. 176), 
usually more (Fig. 171) or less (Fig. 177) 
distinctly 13-segmented; filiform, but api- 
cal 2-4 segments usually somewhat more 
closelv associated or partly fused to form 
inconspicuously differentiated clava 
(Figs 60, 61, 66, 68, 73); flagellum relatively 
sparsely setose, most segments usually 
with a whorl of long trichoid sensilla 
medially to subapically (Fig. 171), fl 9 and 
flio or flg-flio each with a basally bent, 
slightly lanceolate sensillum arising from 
distinct depression distally (Figs 60, 66, 68: 
s4; 61, 73), and apical segment with spine- 
like sensillum projecting from apex. 

Mesosoma: Pronotum in dorsal view not 
visible (Figs 78, 79), very short and vertical 
mediallv (Fig. 111). Pronotum in lateral 
view triangular with acute posterodorsal 
angle extending almost to base of forewing 
(Figs 80, 103), with single seta posteriorly 
near dorsal margin (Fig. 80); dorsal and 
posterior margins often appressed to later- 
al margin of mesoscutum and anterior 
margin of mesopleuron, respectively 
(Figs 80, 95, 96, 101, 108), but not rigidly 
connected by tongue and groove interlock- 
ing system so that posterior margin often 
displaced from anterior margin of meso- 
pleuron over dorsal third to half (Figs 13: 
am; 53, 82, 97, 99); posterior margin 
sometimes with distinct notch (Figs 16, 
82: pn; 99, 101) near dorsal quarter, but 
without evident spiracle. Pronotum in 
ventral view not continuous between sides 
(i.e., not annular) (Fig. 86). Propleura form- 
ing entire lateral and ventral surfaces of 
propectus (Fig. 86), divided mediolongi- 
tudinally (Figs 43, 112) or more or less 
extensively and indistinguishably fused 
medially (Figs 18, 44, 86). Prosternum re- 
flexed internally at about 90 degrees to 
posterior margin of propleura, comprising 
a transverse, vertical surface largely con- 
cealed between propleura and base of 
procoxae (Vilhelmsen and Krogmann 
2006, fig. 8). Prepectus not visible external- 
ly. Tegula absent, but base of forewing 



with bare, oval, humeral plate (Figs 82, 
101: hp) resembling a tegula. Mesoscutum 
(Figs 45, 78, 79, 91, 93) scabrous, without 
evident notauli or parapsidal lines, and 
bare except for four long setae posteriorly 
in a transverse line; transscutellar articula- 
tion straight-transverse. Scutellum 
(Figs 45, 78, 79, 91, 93, 109) in dorsal view 
without differentiated axillae, consisting of 
convex, more or less hourglass-shaped (i.e., 
lateral margins incurved) anterior scutel- 
lum and transverse, concave posterior 
scutellum (frenum) (Figs 78, 79, 91, 93). 
Anterior scutellum in dorsal view scabrous 
to longitudinally scabrous-strigose, but 
bare except for seta on either side within 
anterior half (Figs 79, 108, 109); in lateral 
view, the almost vertical side finely sculp- 
tured (Figs 91, 93, 97, 109). Posterior 
scutellum more or less distinctly, longitu- 
dinally strigose (Figs 78, 79, 9l', 93, 109). 
Mesopleuron high-rectangular with anteri- 
or margin reflexed as slender rim (Figs 13: 
am; 82, 97) normally concealed by posterior 
margin of pronotum; dorsal margin with 
three setiform "subalar" sensilla below 
base of forewing (Fig. 98: sas); without 
differentiated mesepisternum and mesepi- 
meron, and distinct subalar area (acro- 
pleuron) not differentiated except often 
by colour difference described above. 
Metanotum dorsomediallv slender and 
more or less concealed under posterior 
margin of scutellum (Figs 78, 84: no 3 ; 79, 
91, 93, 97, 99, 101); laterally separate from 
metapleuron and propodeum (Fig. 105: 
no 3 , pl 3/ pro) or fused with metapleuron 
(Fig. 84: no 3 , pl 3 ), propodeum (Figs 102: 
"no 3 ", pro; 97, 98, 100), or both meta- 
pleuron and propodeum (Figs 109, 110: 
"no 3 "). Meso- and metapleuron completely 
fused (Figs 80, 83, 103, 108) except for short 
suture below base of wings (Fig. 80: mms) 
or partly (Fig. 87: mms) to completely 
separated by oblique suture (Figs 82: 
mms; 95-97, 99, 101); bare, variable in 
sculpture. Meso/metapleural complex, 
when separated by suture, often with 
suture slightly widened or posterior mar- 



Volume 16, Number 1, 2007 



61 



gin of mesopleuron with tiny notch 
(Figs 97, 98: mn) dorsally at same level as 
notch on posterior margin of pronotum. 
Meso/metapleural complex, when fused, 
usually with variably distinct, curved 
groove (paracoxal sulcus sensu Vilhelmsen 
and Krogmann 2006, fig. 1) extending from 
intersegmental pit (Fig. 80: isp) to meta- 
pleural pit (Figs 80-82, 87: pl 3 p) or propo- 
deal spiracle (Figs 103, 108), and rarely 
with very tiny pit (Fig. 87: plip) at height 
similar to metapleural pit (Fig. 87: p^p). 
Metapleuron indistinguishablv fused with 
propodeum except usually above propo- 
deal spiracle (Figs 80, 82-84, 97-102); with- 
out (Figs 95, 96, 99, 103, 108) or with 
variably distinct metapleural pit (Figs 80- 
82, 87, 97, 101: pl 3 p); dorsal margin with 
a single setiform sensillum near base of 
hind wing (Figs 85, 98, 105, 110: dms); 
anteroventral corner often projecting 
slightlv anteriorlv to abut reflexed rim of 
slightly projecting mesocoxal foramen 
(Figs 86, 87) (in lateral view the two 
protrusions usually form a pit (Figs 80, 
86: isp) between them). Metathoracic-pro- 
podeal complex with single setiform or 
digitiform "postalar sensillum" (Figs 84, 
85, 98, 100, 102, 105, 110, 127: pas), almost 
always with two setiform "prespiracular 
sensilla" (Figs 84, 85, 98, 110, 127: pss; 100, 
105), and usually with one, or rarelv two, 
"suprapleural" setiform sensilla (Figs 85, 
98: sps) anterior to postalar sensillum 
between dorsal margin of metapleuron 
and ventral margin of scutellar-axillar 
complex. Propodeum with sculpture vari- 
able, more or less reticulate to transverse- 
strigose at least dorsally (Figs 78, 79, 91, 93, 
107); with single seta near posterior margin 
of spiracle (Figs 84: prs; 80-83, 95-102, 105, 
109); posteriorly reflexed into variably 
distinct and high flange (Figs 92, 94, 104: 
pf) on either side of short foraminal tube 
(Figs 90-97, 99, 101, 103, 104, 107) or into 
continuous fl-like flange encircling petio- 
lar insertion dorsally and Laterally (Figs 81, 
87: pf) (in either instance the flange and 
dorsal rim of metacoxal foramen abut to 



form pincer-like structure surrounding 
deep pit (Figs 80, 86: isp; 104, 108)). 
Propodeal spiracle (Figs 84, 100, 105, 109: 
sp) below level of propodeal surface, 
surrounded by spiracular aperture 
(Figs 84, 100, 105, 109: spa); spiracular 
aperture usually circular to oval (Figs 80, 
84, 100, 105, 109: spa), rarely slit-like 
(Fig. 82: spa), and usually continuous to 
anterolateral margin of propodeum as 
deep slit (spiracular peritreme) (Figs 80, 
82, 84, 100: spp), with the peritreme and 
dorsal margin of metapleuron either form- 
ing acute angle (Figs 82-84) or more 
uniformly convex margin (Figs 97-102). 

Wings: Forewing pedunculate (Figs 163- 
166, 168) with more or less lanceolate to 
broadly spatulate disc (Figs 113, 124: dsc) 
and slender basal stalk (Fig. 124: stk). 
Forewing disc sometimes almost flat 
(Fig. 166), but usually more or less dis- 
tinctly convoluted by series of longitudinal 
folds (Figs 163, 164); without venation but 
with raised lineations on both upper and 
lower surfaces of membrane forming 
a double layered mesh-like pattern interior 
to level of insertion of marginal setae 
(Figs 123, 163-165); discal setae varying 
from short and spine-like (Fig. 166) to long 
and hair-like (Figs 163, 164), sometimes 
appearing dense when very long 
(Fig. 116), but aligned in row along folds 
when disc convoluted (Figs 163, 164); 
marginal setae, when long, arising distinct- 
ly from within periphery of apical portion 
oi disc (Figs 163, 165, 168). Forewing stalk 
(Fig. 124: stk) more or less distinctly sub- 
divided into "proximal" and "distal" por- 
tions, the distal part consisting of strongly 
narrowed base of disc (Fig. 124: std). In 
dorsal view, proximal part of stalk 
(Fig. 124: stp) differentiated into convex, 
anterior and posterior longitudinal bands 
separated by furrow over most oi length 
except basally (Figs 127, 134); posterior 
band bare but anterior band with sparse, 
short spicules at least anteriorlv and with 
a campaniform sensillum distallv 
(Figs 125, 126, 128: cs) near base of long 



62 Journal of Hymenoptera Research 

seta (Figs 125-128: mc). In ventral view, tudinally, the lanceolate setae slightly 

proximal part of stalk with at least basal flattened and differentiated in length so 

half of anterior margin folded under wing apices form a concave arc. 

as basally widened region of membrane Metasoma: Metasoma 8-segmented, the 

(Figs 129, 132: cc), the membrane with basal two segments tubular, hence with 2- 

scattered spicules (Figs 129, 132) or these segmented petiole (Figs 83, 104: pti, pt 2 ; 88, 

sometimes in row (Figs 95, 130) and, at 92, 94). First petiolar segment almost 

least sometimes, with a campaniform sen- smooth (Figs 81, 83, 104, 107), transversely 

sillum basally to medially (Figs 132, 135: strigose (Figs 90, 92-94) or reticulate 

cs), and with non-folded portion differen- (Fig. 88), and often with single seta on 

tiated into anterior band having short either side in anterior half (Figs 92, 94, 

spicules and concave trough along poste- 104, 107). Second petiolar segment finely 

rior margin (Figs 130, 132: ret). Hind wing sculptured dorsally and often with tiny 

with bulbous base (Fig. 136) and slender spicules (Fig. 83, insert) or transversely 

stalk terminated by pincer-like structure; strigose (Figs 92, 94). Post-petiolar seg- 

stalk with two or rarely three subbasal ments (= gaster) of air-dried specimens 

setae and one medial to subapical seta on usually flattened-oval in cross-section, but 

anterior margin (Fig. 136), and sometimes terga broadly overlapping sterna laterally 

with a slender band of membrane poste- (i.e., without differentiated laterotergites), 

riorly (Figs 135, 136: mb), the membrane and smooth and shiny (Fig. 147); without 

sometimes also with single short seta spiracles except usually on Mt 7 (Figs 149: 

(Fig. 135); apical pincer formed by sp; 150-152, 158, 159). Gaster (Fig. 147) in 

socketed hamulus (Fig. 135: ham) and dorsal view with posterodorsal margin of 

slender projection opposite to hamulus Mt 3 broadly and deeply incurved and Mt 4 

(Figs 134, 135: op), the projection and the largest tergite; Mt 3 with two to several 

sometimes also the hamulus apically bi- setae dorsolaterally near anterior margin; 

furcate (Figs 134-136). Mt 7 with single seta near spiracle 

Legs: Meso- and metacoxae with basi- (Figs 149-152) or usually paralaterally 

coxite reduced to a small lobe (Figs 81, when spiracle absent (Fig. 154); syntergum 

104, 107: be) inserted into widely separated (Mt 8+9 ) with 2-4 setae in row near posterior 

foramina (Fig. 86). Femora with differenti- margin (Figs 150, 152, 153, 155) and usually 

ated trochantellus; posterior surface of with cerci (Figs 150, 152: cer; 149, 151, 158), 

femora sometimes with small bumps the cercus usually flat or low convex, 

(Fig. 146). Tibial spur formula 1:0:0. Proti- subcircular, differentiated from tergite by 

bia without row of modified setae (second- distinct groove, and bearing 1-4 long setae 

ary fine comb) along anterior margin (Figs 150, 151, 158), but sometimes partly 

apically; calcar curved and apically bi- fused with tergite or apparent only as 

furcate (Figs 137, 138, 139a, 140: ca) or subcircular depression (Fig. 152). Hypopy- 

straight and simple (Figs 142, 144a, 145a: gium of female (Fig. 148: hyp) with several 

ca); mesotibia ventroapically with two setae apically (Figs 148, 149, 153), the 

strong, apically divergent, socketed setae sclerite extending almost to apex of meta- 

projecting distally from tube-like eleva- soma and concealing ovipositor when 

tions of cuticle (Figs 139b, 141b, 144b: ps); appressed to syntergum (Fig. 148: syn), 

metatibia with strong setae originating but apex capable of separating widely 

from cuticular protrusions ventroapically from syntergum (Figs 149: syn; 147, 153) 

(Figs 139c, 141c, 144c: ps) and anteroapi- for ventral rotation of ovipositor (Figs 152, 

cally (Figs 139c, 144c). Tarsi 5-segmented; 153: ov; 147, 149). Hypopygium of male 

protarsus with fine comb of basitarsus bare, but with a campaniform sensillum 

(Figs 137: fc; 140, 143, 145a) aligned longi- laterally near base (Fig. 160: cs). Male 



Volume 16, Number 1, 2007 63 

genitalia without basal ring or phallobase, more likely to break the membrane rather 

consisting of large medial aedeagus than simply pushing the head away from it 

(Figs 154, 155, 158, 160: aed), ventrolateral when they are opened outwards. Such an 

volsellae with digiti (Figs 160, 162: vol, hypothesis does not explain why the 

dig), and sometimes with externally pro- occipital plate can also rotate deeply within 

truding parameres (Figs 154, 156: par; 155). the head capsule (Fig. 15), unless this is 

Aedeagus in dorsal view (Fig. 158) divided merely an artifact of drying made possible 

apically and distinctly bilobed basally, by the bellows-like structure, 

with basal lobes much smaller and more The seemingly ventral position of the 

slender than posteriorly broadened apical occipital foramen on the head (Figs 14, 15, 

lobes and each basal lobe with small pit 45, 53) is a consequence of the modified 

near its anterior margin; in ventral view head structure. The occipital foramen 

gonopore positioned apically (Fig. 160: gp); actually is near the center of the head, as 

internally with paired apodemes extending in other hymenopterans, but it appears to 

anteriorly (Figs 169, 170: aea). Volsella be more ventral because the posterior 

(Figs 160, 162: vol ) extending anteriorly surface of the head is abruptly angled 

into body as slender apodeme (Figs 169, along a transverse axis near the dorsal 

170: voa) and posteriorly broadened into margin of the occipital foramen (Figs 23, 

a digitus (Figs 161, 162: dig) having one or 41, 50). This angulation serves as the hinge 

two short spines (Figs 161, 162: dis), the that allows the occipital plate to rotate 

two volsellae together forming a posteriorly anteriorly (Fig. 15) and posteriorly (Fig. 13) 

directed, Y-shaped structure. Paramere, relative to the frontal plate. The structure 

when present, projecting externally from of the occipital foramen and the region 

between syntergum and hypopygium lat- between the foramen and labiomaxillary 

eral to aedeagal-volsellar complex as elon- complex appears to be quite consistent 

gate-digitiform process with a long termi- across the family (Figs 23, 24, 41, 50), 

nal seta (Figs 154, 156: par), and extending though we examined this in detail for very 

internally as rod-like structure (Fig. 169: few species. 

paa) articulating with genital complex Many species have the frontal plate more 

basal to volsellar apodemes. coarsely sculptured ventrolaterally near 

Comparative and functional morphology. — where the hyperoccipital band of pleated 

Head capsule: Certainly the most bizarre membrane originates than immediately 

structural modification of extant mymar- posterior to the eye (Figs 29, 30, 54). The 

ommatids is their unique "bellows-like" sculpture is developed as short vertical 

head (Figs 13, 42). The functional signifi- ridges in some species and specimens of M. 

cance of this remains uncertain, but Creta- anomalum from Sweden have a small series 

ceous fossils suggest that origin of a hyper- of ridges or denticles ventrolaterally on 

occipital band of pleated membrane and both the frontal and occipital plates 

a moveable occipital plate likely evolved (Figs 16, 17). The denticles interdigitate 

concurrently with exodont mandibles. If when the occipital plate is vertical relative 

mymarommatids are egg parasitoids it is to the frontal plate (Fig. 17) and together 

possible that the adult emerges from the they may serve as a weak locking mecha- 

host egg by opening its exodont mandibles nism to align the occipital and frontal 

and rupturing rather than chewing the plates and to inhibit rotation of the 

chorion. Expansion of the occipital region occipital plate. The exact distribution of 

to enlarge the head (cf. Figs 13, 14) may such interlocking denticles is not known, 

serve to fill the enclosing space so that the but most species apparently lack them 

mandibles are appressed firmly against the (Figs 42, 45, 51, 53). All examined species 

chorion. By doing so, the mandibles are have a single seta ventrolaterally on the 



64 



Journal of Hymenoptera Research 



occipital plate that typically projects some- 
what over the hyperoccipital band of 
pleated membrane (Figs 15, 16, 41, 51). 
This seta likely serves to sense whether the 
occipital plate is rotated within the head 
capsule. A long seta on the gena and the 
ventral-most seta of the postgenal plate 
also overlap the base of the mandible 
(Figs 19, 35). These setae likely serve to 
sense when the mandible is opened. 

The only head capsule feature of extant 
mymarommatids that appears to be of 
generic value is the presence or absence 
of paramedial setae on the occipital plate. 
Species of Mymaromella have a pair of setae 
(Figs 45, 46: ms), whereas those of Mymar- 
omma and Zealaromma do not. All exam- 
ined mymarommatids have two subtorular 
setae (Figs 33, 52, insert), but they differ in 
the number and pattern of interorbital 
facial setae (if. Figs 25, 31, 33, 47-49). Both 
species of Zealaromma have four interorbit- 
al setae above the oral margin (Fig. 56). 
There are only two interorbital seta above 
the oral margin in Mymaromma (Figs 18, 
20), but species of Mymaromella either have 
two or four such setae. Both species of 
Zealaromma share a distinctive facial sculp- 
ture (Figs 52, 54), but a few species of 
Mymaromella also have distinct facial sculp- 
ture (Fig. 49). Other features such as the 
presence or absence of ocelli and relative 
size of the eyes and number of ommatidia 
are even more variable among species. 

Mouthparts: The exodont mandibles of 
Zealaromma (Figs 52, 54) and Mymaromella 
(Figs 47-49) are bidentate and compara- 
tively gracile except in Mymaromella sp. 23. 
The single known female of this species 
uniquely has large, robust mandibles with- 
in a conspicuously large oral orifice 
(Figs 36, 37, 39). Because the left mandible 
is open its inner surface is visible (Figs 36, 
37: i). The inner surface is smooth with one 
dorsoapical and one dorsomesal seta and 
two campaniform sensilla, one below the 
dorsomesal seta and one ventromesally 
(Fig. 37: cs). The relative position of the 
two mandibles show that a mandible is 



rotated as it is opened so that the "dorsal" 
surface of a closed mandible projects 
laterally when open (Figs 36, 37: d) and 
the "outer" surface becomes "ventral" 
(Fig. 37: o). The "dorsal" surface (Fig. 40: 
d) is slightly concave, has two small, 
outcurved teeth apically, and two setae 
basally, one adjacent to a campaniform 
sensillum (Fig. 40: cs). The outer surface 
(Fig. 37: o) is flat, has three setae, and its 
apical margin is quite long, almost vertical, 
with a knife-like edge (Fig. 37: c). We infer 
that the dorsal surface of the mandible in 
its closed position in M. sp. 23 is homolo- 
gous with the external surface of the 
mandible of other mymarommatids be- 
cause it is concave with two teeth and has 
a seta associated with a campaniform 
sensillum. We did not observe a campani- 
form sensillum associated with a dorsal 
seta on the outer mandibular surface in all 
mymarommatids examined, but based on 
wide distribution of the sensillum we 
suspect that this was because the mandible 
was not clean enough for it to be observed 
rather than it being absent. In species of 
Mymaromella, the campaniform sensillum 
usually is within the depression from 
which the dorsal seta arises and therefore 
can be observed only under some angles 
(cf. Figs 34, 35: cs). Similarly, we are not 
certain whether a second, more mesal 
campaniform sensillum evident on the 
outer mandibular surface in some species 
(Fig. 35) is characteristic of all species and 
genera, or whether campaniform sensilla 
are characteristic of the internal surface of 
all mymarommatid mandibles. 

The mandibles of most species of My- 
maromma differ from those of Mymaromella 
and Zealaromma because each has an 
additional dorsal angle (Figs 25, 26) or 
tooth (Figs 19, 27, 28) that projects above 
the oral margin, at least slightly (Figs 29, 
30: dt). The only known exception is 
Mymaromma sp. 9, which has bidentate 
mandibles (Fig. 31). The mandibles of 
some species of Mymaromella, if viewed 
slightly from below, sometimes appear to 



Volume 16, Number 1, 2007 65 

have a very tiny dorsal tooth (Fig. 44), but cause the spine-like process (Fig. 41: mxp) 
this is the base of the socketed seta at the originates from a small lobe (Fig. 41: mxs) 
dorsal margin of the mandible and it does on the inner side of a longer fleshy lobe 
not project above the oral margin when (Fig. 41: mxg), which in at least some 
viewed anteriorly (Fig. 48). species has distally projecting papillae on 
The labiomaxillary complex of mymar- its upper/outer surfaces (cf. Fig. 57). We 
ommatids is strongly reduced, which consider that the small subapical lobe 
makes definitive homology of its compo- bearing the spine is a remnant of the 
nents with those of other parasitic Hyme- maxillary palpus (Fig. 41: mxp), the inner 
noptera uncertain. Extant mymarommatids surface from which it arises as likely 
have a ventral labiomaxillary plate com- a remnant of the stipes (Fig. 41: mxs), and 
posed of a medial labium (Fig. 20: lab) and the longer outer lobe as likely a remnant of 
lateral maxillae (Fig. 20: max) that are the galea (Fig. 41: mxg). We also consider 
fused together over at least their posterior the thin, apically pustulate lobe above the 
third. The externally visible part of the stipes in Mymaromma and Zealaromma as 
labium we interpret as the prementum a more internalized galea (Figs 21, 22, 57: 
(Fig. 21: 1pm). A medial papilliform pro- mxg), i.e., homologous with the outer lobe 
cess (Figs 21, 22: pap) also projects exter- of the maxilla in Mymaromella. The struc- 
nally from between the labrum and labi- tural homology of the papilliform process 
um. Except for Mymaromma sp. 9 (Fig. 31, that projects externally between the labrum 
insert), the labium is much wider in species and labium (Figs 21, 22: pap) is uncertain, 
of Mymaromma (Figs 18-21, 23, 28) than in though it might be the hypopharynx or 
Mymaromella (Figs 41, 41, 44, 48, 49) and a part of the maxilla (possibly the ligula). 
Zealaromma (Figs 50, 52, 55, 56). Because of Our interpretation of the labiomaxillary 
the condition of available specimens we complex differs from that of Debauche 
could not determine the exact structure of (1948, fig. 16a), who considered that the 
the maxilla in M. sp. 9. In other species of maxillary stipes were fused medially so 
Mymaromma and in Zealaromma the maxilla that they completely covered what we 
often appears to be composed of a single interpret as the prementum of the labium 
lobe terminated by a spine (in Zealaromma (Fig. 21: 1pm). He also considered the long 
the lobe projects slightly on either side of posterolateral setae of the labiomaxillary 
the base of the spine, Figs 56, 57), but there complex as the maxillary palpi. The two 
is also a second fleshy lobe above the very similar paramedial setae he consid- 
ventral lobe of the maxilla. This dorsal ered simply as sensory hairs. The maxillary 
maxillar lobe appears flat when the labio- regions in Mymaromma that we consider as 
maxillary complex is appressed to the head the stipes he interpreted as the galea, 
capsule (Fig. 21: mxg) and more tubular The structure of the labiomaxillary corn- 
when the complex is distended (Fig. 57: plex was not visible in most fossil mymar- 
mxg). When the dorsal maxillar lobe is ommatids we examined, but a paratype of 
expanded, oblique rows of distally project- Palaeomymar senonicus clearly has the labi- 
ing papillae are visible in lateral view um only about as wide as the maxillae and 
(Fig. 57), which appear as dorsally project- these apparently separated for most if not 
ing papillae when the lobe is flattened all of their length (Fig. 203). We could not 
(Fig. 21). Furthermore, the dorsal lobe is determine the presence of maxillary or 
differentiated apically as a pustulate lobe labial palpi. 

(Figs 20-22, 56, 57) that projects slightly Antenna: For this study we included 

beyond the terminal spine of the ventral what Beardsley et al. (2000) identified as an 

maxillar lobe. Species of Mymaromella have unnamed species from Hawaii near Mi/- 

a slightly different maxillar structure be- maromma goethei in M. goethei (Appen- 



^ Journal of Hymenoptera Research 

dix II). Females we include in M. goethei Mesosoma: The internal and external 
have six (Figs 174, 175) rather than seven mesosomal structure of Mymaromma anom- 
funicular segments. We also saw one alum was studied comprehensively by 
female of Mymaromella sp. 6 that has seven Vilhelmsen and Krogmann (2006). They 
funicular segments in one antenna and six noted that in M. anomalum the propleura 
segments in the other. Males we identify as and prosternum are fused, except for 
M. goethei from Australia and Hawaii have a short distance anteriorly, so that there is 
12-segmented antennae (Fig. 176). The a continual ventral "carapace" (Vilhelmsen 
males of other species have 13-segmented and Krogmann 2006, fig. 6; cf. Figs 18, 86). 
antennae (Figs 171, 177), though the apical Because of the fusion they were uncertain 
two segments sometimes are only indis- as to their original structure. Our survey 
tinctly separated (Fig. 177). We suggest demonstrated that both species of Zealar- 
that the 12-segmented antenna of male M. omma (Fig. 112) and at least some fossil 
goethei and what is either the same or a very mymarommatids have medially abutting 
similar species in Hawaii results from the propleura (not visible in all inclusions). We 
loss of a funicular segment, as for females, could not observe this feature in all 
rather than the loss of the apical claval examined species of Mymaromella and 
segment. Mymaromma, but species of Mymaromella 
Male mymarommatids lack the s2-type appear to have medially abutting pro- 
and s3-type sensilla present on the clava of pleura or at least a sulcus or differentiated 
females, but both sexes have s4-type line of sculpture along the ventral midline 
sensilla. The latter sensilla are more or less (Figs 43, 44), whereas species of Mymar- 
lanceolate in shape, are directed apically omma have an undifferentiated carapace 
because they are strongly bent basally, and similar to M. anomalum. 
each originates from a distinct circular Vilhelmsen and Krogmann (2006) 
depression (Figs 58, 60, 62, 64, 66, 68, 70, showed that the mesothoracic spiracle 
71: s4). Females of Mymaromma and Zealar- and an externally evident prepectus were 
omma have two s4-type sensilla and fe- missing from M. anomalum, but stated that 
males of Mymaromella two or three such it would be desirable to establish position 
sensilla on the outer surface of the clava. of a spiracle relative to the prepectus for 
Both sensilla are on the apical claval inferring possible sister-group relation- 
segment in Zealaromma (Fig. 71). The sen- ships of mymarommatids. They also noted 
silla are near the dorsal margin of the clava that the pronotum had a notch in its 
in Mymaromma (Figs 58, 59) and Zealar- posterior margin at about one third of its 
omma (Fig. 71), but near the midline or height from the dorsal margin (Vilhelmsen 
even more ventrally in Mymaromella and Krogmann 2006, fig. 1), but did not 
(Figs 62, 63, 65, 67, 69). Males lack s4-type comment that this position is similar to that 
sensilla from the apical flagellar segment, of the mesothoracic spiracle in most para- 
but the preceding two or three segments sitic Hymenoptera other than the Chalci- 
have a single s4-type sensillum distally doidea. Our survey shows that all mymar- 
depending on whether the female of the ommatoids lack a mesothoracic spiracle 
species has two or three such sensilla on and an external prepectus, but that there is 
the clava (cf. Figs 65 and 66, 67 and 68, 71 a pronotal notch in many Mymaromma 
and 73). We observed strong, basally (Figs 16, 82: pn) and Mymaromella (Figs 97, 
curved sensilla projecting from the outer 99, 101). We did not observe a pronotal 
or dorsal surfaces of the clava of some notch in Zealaromma (Figs 103, 108) or any 
Cretaceous females, but are uncertain Cretaceous representative, though the lat- 
whether they are s2-type or unusually long eral structure of the pronotum was not 
s4-type sensilla. clearly visible in many amber inclusions. 



Volume 16, Number 1, 2007 



67 



Vilhelmsen and Krogmann (2006) further 
stated that the pronotum is rigidly associ- 
ated with the mesopleuron in M. anom- 
alum. Our survey showed that the prono- 
tum extends to the mesopleuron ventrally 
in mymarommatids and does not appear to 
be moveable relative to the mesothorax, 
but in many species and specimens the 
posterior margin is separated from the 
mesopleuron over its dorsal third to half 
(Figs 13, 82, 97, 99). Vilhelmsen and Krog- 
mann (2006) also described and illustrated 
an internal structure near the posterior 
margin of the pronotum that they postu- 
lated was the prepectus fused to the 
pronotum. Our study did not include 
internal features, but we concur with their 
interpretation that the structure parallel to 
the posterolateral margin of the pronotum 
(Vilhelmsen and Krogmann 2006, fig. 4) 
likely is the prepectus. They concluded that 
the structure is fused with the pronotum. If 
so, it is not homologous with a posterolat- 
eral pronotal inflection sensu Gibson (1985). 
Remnants of membrane on the dorsal 
and posterior margins of the putative 
prepectus (Vilhelmsen and Krogmann 
2006, figs 4, 5) suggest that this is where 
membrane from the mesoscutum and 
anterior margin of the mesopleuron attach, 
respectively, providing the pronotum with 
the flexibility to be separated from the 
mesopleuron dorsally (Figs 13, 53, 82, 
97, 99) but at the same time retaining 
structural continuity between the pronotum 
and mesothorax. Even if separation of the 
pronotum from the mesopleuron dorsally 
is only an artifact of drying, their separa- 
tion shows their margins are not rigidly 
interlocked by a tongue and groove mech- 
anism as in taxa with a posterolateral 
pronotal inflection. 

Vilhelmsen and Krogmann (2006) noted 
that the mesopleuron and metapleuron are 
fused together in M. anomalum, which was 
hypothesized as an autapomorphy of My- 
marommatidae by Gibson et al. (1999). Our 
survey shows that the meso- and meta- 
pleura are separate sclerites in Mymaro- 



mella (Figs 95-99, 101) and in those fossil 
taxa where the feature is visible. Except for 
a very short distance immediately below 
the hind wing (Fig. 80: mms), the sclerites 
are fused in Zealaromma (Figs 103, 108) and 
in most Mymaromma. Mymaromma sp. 9 has 
the meso- and metapleura completely 
separated (Fig. 82: mms) and they are 
separated over about their ventral half in 
Mymaromma sp. 7 (Fig. 87: mms). In My- 
maromella, there is often a tiny notch in the 
posterior margin of the mesopleuron 
(Figs 97, 98: mn) or a slight widening of 
the suture between the meso- and meta- 
pleuron at a similar height as the notch on 
the pronotum. Based on this positional 
similarity, the notch on the posterior 
margin of the mesopleuron could be 
a remnant of the metapleural spiracle. 

The dorsal surface of the mesosoma of 
mymarommatids we examined is very 
similar to that described for M. anomalum 
by Vilhelmsen and Krogmann (2006) ex- 
cept for species-specific sculptural differ- 
ences. However, structure of the meta- 
thoracic-propodeal complex differs con- 
spicuously among Mymaromma, Mymaro- 
mella and Zealaromma. As described by 
Vilhelmsen and Krogmann (2006), the 
aperture of the propodeal spiracle con- 
tinues as a slender peritreme dorsally to 
the anterior margin of the propodeum in 
M. anomalum (cf. Fig. 84: spp). The spirac- 
ular aperture (Figs 80, 82, 84, 100, 105: spa) 
is the more or less enlarged ventral portion 
of the spiracular peritreme that overlies the 
actual opening of the propodeal spiracle 
(Figs 84, 100, 105, 109: sp). They inter- 
preted the spiracular peritreme as the 
antecostal suture, which separates the 
metanotum from the propodeum. A slit- 
like spiracular peritreme is characteristic of 
both Mymaromma and Mymaromella, but its 
direction differs in the two genera and this 
determines whether the postalar sensillum 
and prespiracular sensilla appear to be 
located on the metathorax (Fig. 84: pas, 
pss) or on what appears to be the ante- 
rolateral angle of the propodeum (Fig. 98: 



68 



Journal of Hymenoptera Research 



pas, pss). In Mymaromella, the spiracular 
peritreme is directed anterodorsally and 
the metanotum and propodeum are fused 
mesal to the spiracular peritreme, though 
sometimes a carina is present that may 
represent the line of fusion (Figs 98, 102: c). 
Consequently, a slit separates the dorsal 
margin of the metapleuron from the lateral 
margin of the metanotum /propodeum and 
in lateral view the peritreme and dorsal 
margin of the metapleuron together form 
a relatively evenly convex arc (Figs 97- 
102). The prespiracular sensilla (Figs 98: 
pss; 100, 102) are on the metanotum/ 
propodeum above the spiracle and the 
postalar sensillum is anterior to these, often 
also obviously on the metanotum /propo- 
deum (Fig. 98: pas). However, this position 
for the postalar sensillum is not so obvious 
in species that have the anterolateral angle 
of the metanotum /propodeum narrowly 
attenuated (cf. Figs 99, 100), and in M. sp. 
17 the sensillum appears to be disasso- 
ciated from the metanotum /propodeum 
(Fig. 102: pas). Mymaromella sp. 17 is the 
only mymarommatid we observed having 
a single prespiracular sensillum (Fig. 102), 
though this observation is based on only 
a single clean specimen. Additional indi- 
viduals are required to determine whether 
the number of prespiracular sensilla are 
variable in M. sp. 17 or the loss of one 
sensillum is correlated with what appears 
to be a more highly reduced "metanotum" 
in this species (Fig. 102: "no 3 "). In Mymar- 
omma, the spiracular peritreme is directed 
dorsally, apparently continuous with the 
antecostal suture, and the metanotum and 
metapleuron are fused together to form 
a single Pi -like sclerite anterior to the 
peritreme (Fig. 84). In lateral view, the 
peritreme and dorsal margin of the meta- 
pleuron converge dorsally to form an acute 
angle (Figs 82-84), and the prespiracular 
sensilla are on the metathorax near the 
presumed line of fusion between the lateral 
margin of the metanotum and dorsal 
margin of the metapleuron (Fig. 84: pss). 
The postalar sensillum (Fig. 84: pas) ap- 



pears to originate from the dorsal margin 
of the metapleuron anterior to the pre- 
spiracular sensilla. In very clean specimens 
(Fig. 85: no 3 ?; Vilhelmsen and Krogmann 
2006, fig. 11), the sensillum is seen to 
originate from a small, more or less tri- 
angular region that is differentiated above 
the dorsal margin of the metapleuron. The 
posterior edge of the differentiated region 
projects slightly so that the postalar sensil- 
lum sometimes appears bilobed (Vilhelm- 
sen and Krogmann 2006, fig. 11). 

The different locations of the postalar 
and prespiracular sensilla in Mymaromma 
and Mymaromella show that their different 
metathoracic-propodeal structures are not 
a result of a simple shift in direction of the 
spiracular peritreme, and that one propo- 
deal structure was not derived directly 
from the other. If the peritreme simply 
shifted direction this should not affect 
position of the sensilla on the body relative 
to the different sclerites. 

The metanotum or the anterodorsal 
angle of the metanotum /propodeum ap- 
pear to be separated quite widely from the 
base of the hind wing in Mymaromma 
(Figs 83, 84: hwb) and often in Mymaro- 
mella (Figs 95, 101, 102), though sometimes 
in Mymaromella there actually is a slender, 
inconspicuous intervening region (Fig. 
100). The metanotum in other Hymenop- 
tera, including Mymaridae, is more or less 
truncate laterally (Fig. 11: no 3 ). Typically, 
the metanotum extends to the inner margin 
of the base of the hind wing and the 
anterodorsal margin of the propodeum 
extends to the posterior margin of the hind 
wing (Fig. 11: no 3 , hwb, pro). In Mymar- 
idae, there are setiform sensilla at the 
extreme anterolateral angle of the propo- 
deum below the base of the hind wing 
(Fig. 11). Furthermore, the mymarid meta- 
notum (Figs 11, 12) typically has a single 
seta near its anterior margin sublaterally 
and three more lateral setae, one of the 
setae being somewhat more medial than 
two setae at the extreme lateral margin of 
the metanotum. In some mymarids, such 



Volume 16, Number 1, 2007 



69 



as Mi/mar, the somewhat more medial seta 
originates from the inner surface of the 
metanotum (Fig. 12). Mymarommatids do 
not have a sensillum in the position of the 
sublateral metanotal seta of mymarids, but 
at least some have one, and apparently 
sometimes two, setiform suprapleural sen- 
silla (Figs 85, 98: sps) between the meta- 
pleuron and scutellar-axillar complex an- 
terior to the postalar sensillum. We postu- 
late that the prespiracular sensilla (Figs 84, 
85, 98, 101, 102: pss) in mymarommatids 
are of metanotal origin based on their 
positional homology with similar sensilla 
in mymarids (Figs 11, 12). We also postu- 
late that the postalar sensillum is of 
metanotal origin because of its position in 
most Mymaromella (Figs 98, 100: pas). Al- 
though the postalar sensillum appears to 
originate from the dorsal margin of the 
metapleuron in Mymaromma (Fig. 84: pas), 
we suggest that the thickened region from 
which it originates (Fig. 85: no 3 ?) actually 
is lateral remnant of the metanotum that 
remains near the base of the hind wing 
(Fig. 84: hwb) and that became disasso- 
ciated from the dorsal part of the metano- 
tum (Fig. 84: no 3 ) when the metapleuron 
and metanotum (Fig. 84: pl 3 , no 3 ) fused 
together. We are less certain of the origin of 
the suprapleural sensillum (Figs 85, 98: 
sps), but it too likely is metanotal. 

In Zealaromma, the propodeal spiracle is 
posterior to the anterior margin of the 
propodeum (Figs 105, 109: sp) and in Z. 
valentinei the dorsal margin of the meta- 
pleuron forms a more or less evenly convex 
arc (Fig. 105: pl 3 ). These features are more 
similar to the structure of Mymaromella 
(Figs 98, 100, 102) than Mymaromma 
(Figs 82, 84). Otherwise, the two species 
of Zealaromma have structures of the meta- 
thoracic-propodeal complex that differ 
from each other and from the other two 
genera. Zealaromma insulare has the meta- 
notum (Fig. 109: no 3 ) fused laterally with 
both the metapleuron and propodeum so 
that smooth cuticle completely separates 
the spiracular aperture from the anterior 



margin of the composite structure 
(Fig. 109), whereas sutures separate the 
metanotum from both the metapleuron 
and propodeum in Z. valentinei (Figs 105: 
no 3 , pl 3 , pro; 106). In Z. valentinei, the 
posterior margin of the metapleuron is 
carinate dorsal to the propodeal spiracle 
and it overlies a smooth band that extends 
from the spiracular aperture to the in- 
tersection of the metapleuron, metanotum 
and propodeum (Fig. 105). The linear 
smooth region is similar to the spiracular 
peritreme in Mymaromma and Mymaromella 
except that the region is sclerotized. Al- 
though the metathoracic-propodeal struc- 
tures appear to be quite different in Z. 
insulare (Figs 105, 106) and Z. valentinei 
(Figs 109, 110), in both species the postalar 
sensillum is widely separated from the 
prespiracular sensilla (cf. Figs 105, 110: pas, 
pss). This shared feature suggests that the 
metathoracic-propodeal structure of Z. in- 
sulare evolved from a Z. valentinei-Uke 
structure through fusion of the metano- 
tum, metapleuron and propodeum. It also 
suggests that what appears as a laterally 
truncate, independent metanotum in Z. 
valentinei (Fig. 105: no 3 ) is not structurally 
homologous with the laterally truncate, 
independent metanotum of other Hyme- 
noptera (cf. Figs 11, 12 with Fig. 105). This 
conclusion is based on our hypothesis that 
the postalar sensillum is of metanotal 
origin and our observation that the sensil- 
lum sometimes appears to be disassociated 
from the metanotum /propodeum in My- 
maromella because of elongation and nar- 
rowing of the anterolateral corner of the 
metanotum /propodeum. This transforma- 
tion series is illustrated by Figs 98 — > 100 — > 
102, and we suggest that the anterior 
position of the postalar sensillum in Zealar- 
omma evolved through a similar trans- 
formation series. If so, the superficially 
laterally truncate margins of the metano- 
tum of Z. valentinei are not the "true" 
margins of the metanotum. Dissections of 
Z. valentinei are necessary to determine 
whether the postalar sensillum is actually 



70 



Journal of Hymenoptera Research 



separated from the "metanotum" or 
whether the apparently laterally truncate 
metanotum (Figs 105, 106) extends anteri- 
orly as a slender band beneath the dorsal 
margin of the metapleuron and bears the 
postalar sensillum at its apex. 

Although the structure of the metatho- 
racic-propodeal complex of Z. valentinei 
may not be directly ancestral to the 
structures that characterize Mymaromma 
or Mymaromella, both of the latter struc- 
tures likely evolved from a mymarommatid 
that had an independent metanotum. The 
structures characteristic of Mymaromma 
and Mymaromella could both be derived 
from such a hypothetical structure. Fusion 
of the lateral margin of an independent 
metanotum with the dorsal margin of the 
metapleuron in one lineage would result in 
the structure characteristic of Mymaromma 
(cf. Figs 105, 84), whereas fusion of the 
posterolateral margin of an independent 
metapleuron with the propodeum in an- 
other lineage would result in the structure 
characteristic of Mymaromella (cf. Figs 105, 
98). As noted below, at least some Creta- 
ceous representatives appear to have a pro- 
podeal spiracle in the same approximate 
position as in Mymaromella and a line 
extending dorsally from the spiracle 
(Fig. 187: sp), but it is uncertain whether 
this line represents a peritreme or only 
a smooth band. Furthermore, the presence 
or absence of an independent metanotum 
could not be determined from the amber 
inclusions. Both Mymaromma and Mymar- 
omella possess a slit-like spiracular peri- 
treme (Figs 80, 82, 84, 97-102), which is not 
present in Zealaromma (Figs 105, 109). This 
suggests that the common ancestor of 
extant mymarommatids had both an in- 
dependent metanotum and a slit-like spi- 
racular peritreme, but that the peritreme 
was lost in the common ancestor of Z. 
valentinei + Z. insulare. 

We only observed a single seta dorsally 
between the meso-/ metapleuron and scu- 
tellar-axillar complex anterior to the post- 
alar sensillum in Z. insulare and Z. valenti- 



nei (Figs 105, 110: dms?). Because this seta 
is very close to the base of the hind wing 
and is quite obvious we tentatively consid- 
er it as homologous with the seta on the 
dorsal margin of the metapleuron rather 
than with the suprapleural sensillum of 
Mymaromma and Mymaromella. However, 
dirt or position of wings prevented obser- 
vation of the presence or absence of the 
different sensilla in many specimens and 
further study is necessary to document 
their distribution accurately in all species 
of Mymaromma and Mymaromella. 

Another feature of the metathoracic- 
propodeal complex that differs among 
mymarommatids is the presence or ab- 
sence a metapleural pit, and its position 
when present. We did not observe a meta- 
pleural pit in species of Zealaromma 
(Figs 103, 108), whereas species of Mymar- 
omma have quite a distinct pit that is 
comparatively close to the propodeal spi- 
racle (Figs 80-83: pl 3 p). Species of Mymar- 
omella are variable in presence or absence 
of a metapleural pit (cf. Figs 95-97, 99, 101) 
and it is so small that often it is visible only 
with SEM (Figs 97, 101). Therefore, the 
apparent absence of a metapleural pit in 
some species of Mymaromella may be 
because of the angle of view or dirt 
concealing the minute hole. When visible, 
the pit is at least midway between the 
propodeal spiracle and the ventral margin 
of the metapleuron (Figs 97, 101). We are 
uncertain of the presence or absence and 
relative position of a metapleural pit in 
fossil mymarommatids because of the 
difficulty in observing such a tiny feature. 
However, P. agapa appears to have an 
unusually large and distinct metapleural 
pit that is quite close to the propodeal 
spiracle (Fig. 187: pl 3 p, sp). A few species 
of Mymaromma also have a tiny pit on the 
mesopleuron (Fig. 87: pl 2 p) at a similar 
height as the larger metapleural pit 
(Fig. 87: pl 3 p), but this requires clean 
specimens and SEM for observation. 

The final variable feature of the propo- 
deum is its structure posteriorly. In My- 



Volume 16, Number 1, 2007 



71 



maromma, the posterior margin extends as 
a n-like flange over (Figs 78, 79) and on 
either side of the petiolar insertion (Figs 81, 
83). The flange normally conceals the 
propodeal foramen in dorsal (Figs 78, 79) 
or lateral (Fig. 80) view. If the metasoma is 
dissected from the mesosoma (Figs 86, 87) 
the propodeal foramen is seen to project 
slightly as a circular ring (sometimes also 
visible if mesosoma observed from pos- 
terolateral view, Fig. 81). The ventral edge 
of each side of the propodeal flange abuts 
the posterodorsal margin of the slightly 
protruding metacoxal foramen so that the 
two form a rigid pincer-like structure 
around a deep pit (Figs 80, 86: isp; 81, 83, 
87). Although the posterior propodeal 
structures of Mymaromella and Zealaromma 
are superficially quite different from that of 
Mymaromma, thev differ only by lacking 
the propodeal flange dorsally above the 
petiolar insertion. Consequently, the pro- 
tuberant, somewhat tubular propodeal 
foramen is more readily visible (Figs 90- 
97, 99, 101, 103, 104, 107). We could not 
determine the exact structure of the pro- 
podeum in fossils, but they appear to have 
structures (Figs 184, 191) similar to those 
Mymaromella that lack a distinct vertical 
flange on either side of the foramen (cf. 
Fig. 94). In Mymaromella and Zealaromma, 
the rim of the metacoxal foramen and the 
flange on either side of the propodeal 
foramen form the same pincer-like struc- 
ture as in Mymaromma (cf. Figs 92, 94, 104 
with Figs 81, 87: pf). Although usually less 
noticeable, a similar pincer-like structure is 
formed between the dorsal margin of the 
mesocoxal foramen and the anteroventral 
angle of the metapleuron, and sometimes 
between the anteroventral margin of the 
mesopleuron and posterolateral margin of 
the pronotum (Figs 80, 86: isp). The adap- 
tive function of the pincer-like structures is 
unknown. In most individuals, the ante- 
roventral projection of the metapleuron 
and the ventral margin of the propodeal 
flange appear to extend only to the outer 
edge of the rim of the meso- and metacoxal 



foramina, respectively, mesal to an oblique 
trough in the rim of the coxae dorsolater- 
al^ (Fig. 87: ct). The trough likely cradles 
the dorsal, constricted part of the coxa 
(basicoxite) that inserts into its respective 
foramen (Figs 81, 104, 107: be). In Zealar- 
omma, the propodeal flange projects ven- 
trally as quite a strong digitiform process 
(Figs 104, 107). If the metacoxa is raised, 
the propodeal process would articulate 
within a basal groove of the metacoxa on 
the inner side of the basicoxite (Figs 104, 
107). The lateral margin of the posteriorly 
protruded metacoxal foramen would like- 
wise articulate within a basal groove on the 
outer side of the basicoxite (Fig. 107). 
Consequently, the ventrallv projecting pro- 
podeal process and posteriorlv projecting 
metapleural process may help stabilise or 
help control movement of the metacoxa, at 
least in Zealaromma. 

Extant mymarommatids have onlv a sin- 
gle seta on the propodeum near the 
spiracle (Figs 84: prs; 80-83, 95-103, 109). 
Mymaridae usually also have only a single 
propodeal seta (Fig. 11: prs), though there 
is virtually no information concerning the 
distribution of propodeal setae in Hvme- 
noptera and quite likely the number of 
setae is at least partlv correlated with body 
size. 

Wings: What we described as the distal 
part of the forewing stalk in mymaromma- 
tids has its anterior and posterior margins 
curved dorsally. In Cretaceous mymarom- 
matids (Figs 185, 194, 195, 205) and in 
many extant species the recurved margins 
form only quite a short, U-shaped gutter 
basal to the widened disc surface 
(Fig. 125). Some extant species have the 
margins abutting along a longer distance to 
form more of a complete tube (Figs 124, 
131). A slightly thicker basal part of the 
stalk (Figs 124: stp; 167) forms a "proxi- 
mal" portion that in dorsal view has 
a single campaniform sensillum and seta 
distally (Figs 125-128: cs, mc). We consider 
the proximal part of the forewing stalk to 
be composed of remnants of the wing base 



7 ? JOURNAL OF HYMENOPTERA RESEARCH 

and venation similar to those of Chalcidoi- gutter in Mymaromma sp. 10 (Fig. 125). A 
dea (see discussion of suprafamilial rela- cross-section of the forewing immediately 
tionships). We interpret the part of the distal to the long seta reveals a more or less 
anterior convex band that folds under the S-shaped folding pattern (Fig. 126). In 
wing as the costal cell (Figs 127, 129: cc; cross-section, the posterior band is thin 
130, 132), the region between the putative and forms a deeply concave fold that 
costal cell and the distal seta and campani- comprises the retinaculum (Fig. 126: ret), 
form sensillum as the marginal vein whereas the anterior band is tubular and 
(Figs 95, 127, 129, 185: mv; 130, 132), and folds dorsally upon itself so as to abut the 
the distal part of the putative marginal vein posterior band. The tubular portion ap- 
that appears bulbous or curves slightly pears to be subdivided into two parts by 
away from the wing margin in slide a thin septum and the larger portion is 
preparations or in fossils as the remnant filled with some substance. If this sub- 
of the stigmal vein (Figs 185: stv; 194). The stance is dried haemolymph and if the 
retinaculum is the posterior band that in smaller portion that appears empty is 
ventral view is concave (Figs 127, 132: ret), a trachea (Fig. 126: tra?), then this further 
A slender band that extends toward the supports the contention that the anterior 
base of the wing from the base of the tubular portion is a remnant of a vein, 
putative marginal vein, which separates Members of Mymar have even more 
the costal cell from the retinaculum, we conspicuously pedunculate forewings 
interpret as the submarginal vein (Figs 95, (Fig. 2) than mymarommatids. The anteri- 
127, 132, 185: smv). In slide preparations, or margin of the forewing (Figs 4, 5: am) is 
the regions described above as the margin- curved dorsally and is folded over to its 
al and stigmal veins often appear to be posterior margin so as to form a longitudi- 
filled with air (Fig. 167), which supports nally divided tube distal to where the hind 
the hypothesis that these represent vein wing attaches to the retinaculum. The 
remnants. The length of the putative margin is folded over immediately beyond 
marginal vein is variable (cf. Figs 95, a single long seta (Figs 3, 4: mc) and there 
127, 130, 132), but our survey was in- are several campaniform sensilla (Fig. 4: 
sufficient to determine whether length is cs) distal to the seta, unlike the single, more 
a generic or only a specific feature. A line basal sensillum of mymarommatids. In 
of "dots" are visible near the anterior Mymar, the long seta is homologous to 
margin of what we interpret as the costal the distal macrochaeta of other mymarids 
cell in some slide preparations (Fig. 167). and the campaniform sensilla likely are all 
These dots are in the same position as a line that remain of a reduced stigmal vein, 
of spicules on the costal cell of some Consequently, the forewing of Mymar is 
species when studied with SEM (Figs 95, quite similar to that of mymarommatids 
130). Other species have scattered spicules except for position of the campaniform 
on the costal cell (Figs 129, 132). A distinct sensilla relative to the long seta. The 
line of dots on the forewing basal to the sensilla are positionally correct in Mymar, 
marginal vein in some Cretaceous fossils but not in mymarommatids, if they are 
(Figs 185, 201) suggests that a straight line stigmal in origin and the long seta delimits 
is the groundplan state, but our survey of the apex of the marginal vein. The fore- 
extant species was insufficient to deter- wings of Mymar are additionally similar to 
mine whether the arrangement of spicules most mymarommatids in having long 
could be informative for differentiating marginal setae that arise from within the 
supraspecifc taxa. wing periphery (Fig. 7). In mymaromma- 
The anterior and posterior margins of tids, the setae are inserted conspicuously 
the forewing are curved into a U-shaped within the wing periphery only in those 



Volume 16, Number 1, 2007 



73 



species with long marginal setae (Figs 113- 
119, 121, 163, 165, 168). The deep insertion 
of the marginal setae is not apparent when 
these are short (Figs 120, 122, 164). The 
functional significance of a deep insertion 
for the setae likely is to minimize their 
flexion at the wing margin when the wing 
is moved through the air so as to help keep 
the setae in the same plane as the wing 
surface. This would result in the disc and 
projecting setae together forming a larger 
wing "surface". The forewings of mymar- 
ommatids are unique in having a mesh-like 
pattern on the disc (Figs 163-165, 195, 211) 
that is formed by lines of raised membrane 
on both the upper and lower surfaces 
(Fig. 123). Although the pattern is not 
distinct in some amber fossils (Figs 183, 
205), this appears to be an artifact of 
preservation. The pattern likely is less 
distinct in amber fossils because the resin 
fills the depressions between the lines of 
raised membrane similar to using glycerine 
to fill small surface irregularities in amber. 
The mesh-like reticulations may serve to 
provide strength to the relatively large 
wing disc so that it is not deformed as it 
is pushed through the air. Similarly, the 
longitudinal folds of the disc membrane 
that result in a more or less corrugated 
forewing in many mymarommatids 
(Figs 113-115) may help wing rigidity. 
The longitudinal folding affects wing 
shape to some extent because the fore- 
wings usually appear more elongate-lanceo- 
late in species with deeper folds than in 
species with less distinct folding (cf. 
Figs 113-115, 163 with Figs 116, 118, 166). 
Consequently, shape and relative dimen- 
sions of the forewings can be affected by 
method of specimen preservation, such as 
slide versus dry mounting. 

The length and thickness of both the 
marginal and discal setae are quite variable 
among mymarommatids (cf. Figs 113-122). 
Species of MymaromeUa tend to have short, 
spine-like discal setae (Figs 166, 168), 
whereas those of Mymaromma (Fig. 163) 
and Zealaromma (Figs 164, 165) have lon- 



ger, more hair-like setae, but there is 
considerable variation. The forewings of 
all but one species of Mymaromma are 
characterized by the presence of a conspic- 
uously long posterobasal marginal seta 
that is separated by several short setae 
from long marginal setae apically 
(Figs 113-115, 163). All the posterobasal 
marginal setae are quite long in Mymar- 
omma sp. 10, but this may be correlated 
with its very long and conspicuous discal 
setae (Fig. 116). Many species of Mymar- 
omeUa have the posterobasal setae all short 
(Figs 117, 120, 166), though some have 
quite a long posterobasal seta (Figs 118, 
119) similar to species of Mymaromma. The 
forewing marginal setal pattern of Z. 
insulare (Figs 122, 164) is modified similar 
to MymaromeUa sp. 20 (Fig. 121), whereas 
Z. valentinei has a forewing marginal setal 
pattern that is unique among extant my- 
marommatids. The forewing of Z. valentinei 
has three or four quite long basolateral 
setae basally on the posterior margin 
(Fig. 165), which is similar to some Creta- 
ceous mymarommatids (Figs 188, 194). 
Other Cretaceous species (Fig. 195) and 
some Tertiary (Fig. 211) fossils have a sin- 
gle conspicuously long posterobasal seta, 
and one Tertiary species has uniformly 
short posterobasal setae (Figs 207-209) (see 
further below). 

The stalk-like hind wing of mymarom- 
matids (Figs 135, 136) terminates in a C- 
like structure formed from a single, curved, 
socketed hamulus (Fig. 135: ham) and an 
opposing process that appears to be a pro- 
jection of the wing (Figs 134, 135: op). 
Consequently, the opposing process is 
a functional analogue, but probably not 
homologous with what Basibuyuk and 
Quicke (1997) called "modified-erect se- 
tae". Between them, the hamulus and 
opposing process grasp the hind margin 
of the forewing. The apex of the hamulus 
inserts into the retinaculum (Figs 130: ret; 
129, 133) and the apex of the opposing 
process is appressed against the dorsal 
surface of the wing (Figs 134: op; 133). This 



74 Journal of Hymenoptera Research 

functional complex is quite similar to that hind wings, apparently strong coupling 
of Mi/mar, in which the stalk (Fig. 2) has system between the fore- and hind wings, 
two socketed hamuli and additional distal ratchet slide mechanism, and other mod- 
projections that grasp the forewing (Fig. 5) ifications of the forewing discussed above 
(in some species the wing continues as may all serve to reinforce the forewing to 
filament beyond the hamuli). The hind prevent its deformation and hold it at the 
wing stalk of mymarommatids is com- necessary angle so that the individual can 
posed of a tubular vein along the anterior balloon. 

margin and sometimes a slender band of Legs: As first mentioned by Gibson 
membrane posteriorly (Figs 135, 136: mb), (1993), extant mymarommatids have two 
though distinct membrane usually is not different structures of the protibial spur or 
evident. The stalk has long setae along its calcar sensu Basibuyuk and Quicke (1995). 
leading margin and these setae project Species of Mymaromella (Figs 137, 139a: ca) 
within the retinaculum when the wings and Zealaromma (Fig. 140: ca) have a corn- 
are joined (Figs 129, 130), perhaps serving paratively long and curved, apically bi- 
to sense position of the hind wing and/or furcate calcar, whereas those of Mymar- 
to further position the hind wing relative to omnia (Figs 142, 144a, 145a: ca) have a short, 
the forewing. The process of the hind wing simple or needle-like calcar. The protibia 
is bifurcate in most if not all species also has a strong, socketed seta ventroapi- 
(Figs 134-136). At least some species have cally on either side of the calcar that 
longitudinal striations on the dorsal sur- originates from a tube-like elevation of 
face of the posterior band of the proximal the cuticle (Figs 138, 140, 141a, 144a, 145a, 
part of the forewing. The striations likely b: ps). Because of their position and 
act as "tracks" along which the bifurcation structure, under some angles of view one 
of the hind wing process slides (Figs 133, of these socketed setae can be mistaken for 
134). Furthermore, at least some species the calcar in Mymaromma (cf. Figs 142, 143, 
have short, distally projecting denticles on 144a: ca, ps); however, the calcar originates 
the posterior margin of the retinaculum from a concave region that is continuous to 
(Fig. 132: ret) below much of the region we the apex of the protibia (Figs 144a, 145b). 
interpret as the marginal vein. These There is also a much smaller second pro- 
denticles may function as a ratchet, en- jection within the concave region of some 
abling the hamulus and hind wing to slide species of Mymaromma (Fig. 145b). Extant 
distally in the retinaculum when the fore- mymarommatids lack meso- and metati- 
wing is moved downward, but impeding bial spurs, but there are two strong, 
movement of the hamulus if it is slid along socketed setae that originate from tube-like 
the posterior margin of the retinaculum elevations of the cuticle ventroapically on 
when the forewing is moved upward. Such the mesotibia (Figs 139b, 141b, 144b: ps) 
a ratchet structure could be used to help and metatibia (Figs 139c, 141c, 144c: ps) 
maintain the fore- and hind wing complex similar to the protibia. The two ventroapi- 
at a specific angle relative to the body for cal setae project beyond the apex of the 
extended periods of time. It is unknown respective tibiae and usually diverge dis- 
whether the forewings can be rotated for tally. We refer to the two ventroapical setae 
the correct movements necessary to pro- on the tibiae as "pseudospurs" because 
duce lift during upward and downward they resemble the articulated spurs of 
arcs of the wing. It may be that mymar- parasitic Hymenoptera with true tibial 
ommatids actually do not fly, but simply spurs. Because of their size, we were 
use their comparatively large forewing unable to determine whether spur-like 
surfaces as "kites" and are blown passively projections visible on the meso- and meta- 
in wind currents. If the latter, the rod-like tibiae of some fossil mymarommatids are 



Volume 16, Number 1, 2007 75 

true spurs or pseudospurs. We suspect males of Z. insulare retain a seta on Mt 7 

they are pseudospurs because there are (Fig. 154) that in Mymaromma and Mymar- 

more than two spines on the metatibiae of omella is associated with the spiracle 

some fossils similar to extant mymarom- (Figs 149-152). 

matids (Figs 139c, 144c). Lin (1994, fig. 3) illustrated the male 

The only other leg feature we found to genitalia of a species identified as M. 

differ among mymarommatids was the anomalum. Triapitsyn and Berezovskiy 

presence of small "bumps" on the poste- (2006) stated that this was a misidentifica- 

rior surfaces of the femora (Figs 143-145), Hon of their new species, M. i/pt, and gave 

particularly the mesofemur (Fig. 144) of another illustration of the genitalia (fig. 9) 

both species of Zealaromma. The function of based on their specimens. Both illustra- 

the bumps and whether or not they are tions show a posteriorly directed Y-shaped 

campaniform sensilla is unknown. structure over a larger medial structure. 

Basibuyuk et al. (2000) examined the The medial structure tapers posteriorly 

sensilla of the orbicula of the tarsal claws of and is divided apically; anteriorly it has 

an unidentified mymarommatid species in lateral rods that extend for a distance 

their review of that structure in Hymenop- greater the medial rod of the Y-shaped 

tera. We did not attempt to extend their structure. The two drawings appear quite 

survey to determine whether the character different from our SEM microphotographs 

states they documented for the species of the male genitalia of Mymaromma sp. 7 

vary in Mymarommatidae. (Figs 158-160), but the differences likely 

Metasoma: Individuals of Zealaromma are mostly because of the different meth- 

lack cerci and have four setae in a row ods used to prepare and study the genitalia 

near the posterior margin of the syntergum (slide mounting vs. SEM) rather than 

(Figs 153, 155). Mymaromma and Mymar- species differences. Microscope slide pre- 

omella have cerci and these usually are parations of the male genitalia of M. sp. 7 

almost flat, subcircular and with four long and of a specimen of M. anomalum from 

setae (Figs 150: cer; 158). The number of Japan also show a medial Y-shaped struc- 

cercal setae is reduced (Fig. 151) in some ture (Fig. 170: voa) and two longer and 

species of both genera and rarely the cercus stronger paramedial processes (Fig. 170: 

is almost indistinguishably integrated with aea) similar to those in the drawings of 

the tergal surface except for being some- Lin (1994) and Triapitsyn and Berezovskiy 

what concave (Fig. 152: cer). However, (2006). Additionally, there are a pair of 

there is always at least one seta that is very slender, obliquely angled, somewhat 

more or less obviously associated with the sinuate structures (Fig. 170: paa?) exterior 

cereal depression on either side of two to the paramedial processes, which extend 

paramedial syntergal setae. In such in- to the same level as the medial rod of the Y- 

stances, the syntergum appears to have shaped structure (Fig. 170). A photograph 

four setae in a row (Fig. 152) rather than of the apex of the male metasoma of a male 

the two paramedial setae (Figs 150, 151, of M. ypt sent to us by Serguei Triapitsyn 

158) that are otherwise characteristic of (UCRC) shows that it also has the slender 

Mymaromma and Mymaromella. This sug- oblique structures. The oblique structures 

gests that the cerci were lost in Zealaromma extend anteriorly to the same level as in 

through their fusion with the tergal surface Fig. 170, but posteriorly they appear to 

and that the two outer syntergal setae of articulate with the anteriorly directed arms 

Zealaromma are homologous with cereal associated with the sclerotized fl-shaped 

setae in Mymaromma and Mymaromella. structure (Fig. 170: syn?). 

Individuals of Zealaromma also lack meta- Zealaromma males differ conspicuously 

somal spiracles (Figs 153-155), though from those of Mymaromma and Mymaro- 



76 Journal of Hymenoptera Research 

mella because they have elongate-digiti- fig. 7). We therefore interpret the longer 
form processes that project laterally from paramedial rods visible in slide prepara- 
between the syntergum and hypopygium tions of the male genitalia of Zealaromma 
(Figs 154, 156: par). Each process has a long and Mymaromma as the apodemes of the 
terminal seta and the processes point in aedeagus (Figs 169, 170: aea). A pit in each 
different directions in some specimens (cf. smaller anterior lobe of the aedeagus 
inserts in Figs 155, 156), which indicates (visible with SEM) may represent the base 
they are articulated or membranous basal- of the respective apodeme. The medial rod 
ly. The processes resemble the exserted of the Y-shaped process in males of both 
cerci of some Chalcidoidea (e.g. Torymi- genera (Figs 169, 170: voa) appears to be 
dae), but there are several setae on the cerci continuous with the two ventrolateral 
of chalcids whether these are exserted or volsellae/digitae (Figs 161, 162, 169: dig), 
plate-like, similar to mymarommatids In the slide of Z. valentinei, the medial rod 
(Figs 150, 158) and some other parasitic appears to consist of two appressed apo- 
Hymenoptera (Fig. 157). We consider the demes rather than just a single apodeme 
elongate processes of the male genitalia of (Fig. 169) (also suggested in Lin 1994, 
Zealaromma to be parameres. The genitalic fig. 3). Based on positional similarity, we 
structure of Zealaromma is similar to that of suggest that the internal, obliquely angled, 
male Maamingidae, which was described slender rods of male Mymaromma (Fig. 170: 
and illustrated by Early et al. (2001, figs 12, paa?) may be homologous with the internal 
13). The genital complex of Maaminga rangi rod-like portions of the parameres of male 
Early et al. (2001) apparently lacks a phal- Zealaromma (Fig. 169: paa). We are un- 
lobase, but has a medial, apically divided certain as to the structural homologues of 
aedeagus (Fig. 157: aed), a volsella with the two darker, c-shaped regions in 
a spined digitus (Fig. 157: vol, dig) ven- Fig. 170. The smaller, apical c-shaped re- 
trolaterally on either side of the aedeagus, gion (Fig. 170: anp?) may be a sclerotized 
and a digitiform paramere with two anal plate, which separates the anus from 
terminal setae (Fig. 157: par). The digiti- the genital complex (Fig. 159: an, anp). The 
form parameres of M. rangi widen anteri- lateral longitudinal processes associated 
orly and articulate with the aedeagal- with the larger c-shaped structure appear 
volsellar complex basally (Fig. 157). In to be lateral, anteriorly extended portions 
microscope-slide preparations of the male of the syntergum (Fig. 170: syn?). The 
genitalia of Z. valentinei, the bases of the larger c-shaped structure consists of lateral, 
externally visible parameres (Fig. 169: par) slightly convex bands and a straight, trans- 
appear be continuous with slender rods verse ventral band (Fig. 170). The former 
(Fig. 169: paa) that extend anteriorly lateral may simply be the edges of one of the 
to the much stronger paramedial rods tergites and the latter the edge of a sternite, 
(Fig. 169: aea). The slender rods curve but further study is necessary to resolve 
toward the aedeagal-volsellar complex at this. 
a level near the end of the medial Y-shaped 

structure (Fig. 169: voa), where they possi- EXTINCT FAUNA 

bly articulate with the complex. Except for Tertiary taxa.— The only mymarommatid 

the externally projecting parameres, struc- genus described from Tertiary amber is 

ture of the genitalia observed in the slide Palaeomymar Meunier. The type species, P. 

preparation of Z. valentinei is quite similar succini Meunier (1901), was based on 5 of 

to that of Mymaromma sp. 7 (cf. Figs 169, the original 13 males and 3 females that 

17 °)' Duisburg (1868) discussed and illustrated. 

The aedeagus of male chalcids has We did not locate any amber inclusion we 

paired aedeagal apodemes (Gibson 1997, consider as part of their material (see 



Volume 16, Number 1, 2007 77 

"Neotype designation" under Palaeomy- and a forewing that has the posterobasal 

mar) and therefore base our interpretation marginal seta obviously longer than sever- 

of P. suceini on their descriptions and al setae distal to it (Fig. 211). The female 

illustrations. also has seven funicular segments and an 

Although Duisburg (1868) did not de- unsegmented clava. Based only on these 

scribe his specimens in detail he stated features it is likely that the species, if 

why he thought they probably belonged to extant, would be assigned to Mymaromella. 

Mi/mar Curtis. The diagnosis he gave for Two pieces of amber, both labelled as 

Mi/mar included 13-segmented antennae ZMUC: 16-1/961, each has a single male, 

for males, 9-segmented antennae for fe- One is insufficiently preserved to deter- 

males, and 4-segmented tarsi. He said that mine its relevant features, but the other 

the fossils differed from Mymar primarily male has a long posterobasal marginal seta 

in wing features and stated specifically that and lacks a long calcar. If extant, this male 

the long marginal setae began near the probably would be assigned to Mymar- 

apical half of the forewing, as he illustrated omnia. Another two males (ZMUC: 8-19/ 

(Duisburg 1868, fig. II). The forewing of P. 1954, 1890-108) and two females (ZMUC: 

suceini illustrated by Meunier (1901, fig. 12) 28-3/1958, 16-1/1961) are characterized by 

also shows the marginal setae along the the absence of a long protibial calcar, 

posterior margin gradually increasing in uniformly short setae along the posterior 

length apically, i.e., without a conspicuous- margin of the forewing basally (Figs 207- 

ly long basal seta. Meunier (1901) de- 209), and a 6-segmented funicle in the 

scribed five rather than four tarsal seg- females (Fig. 179). Because of these fea- 

ments for P. suceini, but this discrepancy is tures we believe that this third species is 

not surprising because Duisburg (1868, fig. the one that Duisburg (1868) illustrated 

II) illustrated, and thus likely examined, and Meunier (1901) named as P. suceini. 

a female at only 85x magnification. All Although Duisburg (1868) and Meunier 

known mymarommatids have five tarsal (1901) described and Meunier (1901, 

segments. Meunier (1901, fig. 13) only de- fig. 13) illustrated a 13-segmented male 

scribed and illustrated the 13-segmented antenna for P. suceini, males may only 

antenna of a male and provided no in- have 12 antennal segments. The left anten- 

formation to confirm or contradict Duis- na of the male in ZMUC: 8-19/1954 is 

burg's comparison of the female antenna clearly visible in ventral view. The apical 

with that of Mymar. The habitus line four segments are all about the same length 

drawing of Duisburg (1868, fig. II) appears and form a differentiated clava because 

to show only five funicular segments, but they are slightly thicker and more broadly 

known female mymarommatids have ei- attached than 6 more slender funicular 

ther six or seven segments. Based on the segments (cf. Fig. 177). The apical flagellar 

stated similarity of the antenna to Mymar, segment illustrated by Meunier (1901, 

the flagellum likely consisted of six funic- fig. 13) is very short and it may only be 

ular segments and an unsegmented clava. a subsection of the apical segment that is 

We examined 10 Baltic amber mymar- differentiated by a whorl of projecting 

ommatid inclusions, 8 from ZMUC, 1 from setae. The ZMUC: 8-19/1954 male in 

GZG, and 1 from NHRS. The inclusions ventral view also appears to have the 

from ZMUC contained five males and propleura fused into a carapace, the mand- 

three females that likely constitute three ibles lack any distinct dorsal tooth or angle, 

species. The first species, represented by and although the labiomaxillary complex is 

a male (ZMUC: 17-5/1963) and a female quite broad the labium appears to be only 

(ZMUC: 1-5/1967), is characterized by about as wide as the maxilla. The genitalia 

a long, curved protibial calcar (Fig. 212) project ventrally from the metasoma and 



78 Journal of Hymenoptera Research 

lateral digitiform processes (parameres) Schliiter and Kohring, 1990). The left wing, 

are not evident. If our observations are examined with a compound microscope, 

correct, the ZMUC amber inclusions in- appears to have a posterobasal marginal 

dicate that Tertiary mymarommatids had seta that is obviously longer than several 

more diverse combinations of features than more distal short setae, though it is shorter 

the extant fauna. No extant mymaromma- than the posterobasal seta in Fig. 118 (the 

tid has a short calcar (characteristic of right wing also has a "line" apparently 

Mi/maromma) and all the setae along the projecting from its posterobasal margin 

posterior margin of the forewing short that superficially appears as a seta, but it 

basally (characteristic of many Mymaro- is longer than that of the left wing and may 

mella but not Mymaromma). be an artifact). The mandibles definitely are 

The GZG specimen is a male positioned bidentate, similar to those of extant My- 
dorsal side up in a thin piece of amber maromella and Zealaromma (cf Figs 47, 54), 
glued to a microscope slide. It has a long, and the vertex appears to have ocelli under 
curved protibial calcar and the left fore- some angles of light when examined with 
wing, viewed with the compound micro- a dissecting microscope. The body cuticle 
scope, appears to have a very long poster- is translucent in lateral view so that 
obasal seta. Also visible are the dorsolon- presence or absence of a suture between 
gitudinal and dorsoventral flight muscles the meso- and metapleuron is not evident, 
plus the mesotergal-trochanteral muscles. The exact structure of the propodeum 
Using a compound microscope, the meso- cannot be determined, but it apparently 
tergal-trochanteral muscle of one side does not have a strong propodeal flange 
clearly attaches to the apex of an obliquely because the basal constriction of the first 
projecting, rod-like axillar phragma (cf. petiolar segment is visible in lateral view 
Vilhelmsen and Krogmann 2006, fig. 13). (cf. Fig. 99). Schluter and Kohring (1990, 
The NHRS specimen, also a male in an figs 1, 2) described and illustrated a 12- 
amber block glued to a microscope slide, is segmented antenna for P. duerrenfeldi, in- 
positioned in a somewhat dorsolateral eluding seven funicular and three claval 
view. It also appears to have a long, curved segments. The terminal claval segment of 
calcar, but the forewing marginal setal the right antenna was drawn with its basal 
pattern is not clear because the forewings half bulbous and its apical half more 
are crossed over basally. narrowly digitiform. The clava is more 

The only other described Tertiary my- distinctly 4-segmented when examined 
marommatid, P. duerrenfeldi, is based on with a compound microscope because 
a single male from Sicilian amber. Sicilian a transverse line (suture) divides the two 
amber is dated at about 5 mya compared to subsections of the terminal segment (cf. 
Baltic amber at about 44 mya (Grimaldi Fig. 177). Because the flagellum has seven 
and Engel 2005). The lateral habitus draw- distinct funicular segments we believe the 
ing of P. duerrenfeldi provided by Schluter antennae to be 13-segmented with a 4- 
and Kohring (1990, fig. 1) illustrated a com- segmented clava. Consequently, the female 
paratively long and straight (needle-like) of P. duerrenfeldi probably has 7 funicular 
protibial calcar and a broad forewing that segments and a 1-segmented clava. 
is evenly rounded apically. Our study of Cretaceous taxa.—We examined type ma- 
the holotype confirmed that the calcar is terial of all the described Cretaceous 
long but that it is also curved distally. The species except for P. japonicum Fursov et 
forewings are most similar to those of al. (2002). The unique female of P. agapa 
extant Mymaromella because of their shape differs from females of other described 
and the presence of short, spine-like discal mymarommatid species in several re- 
setae (represented by dots in fig. 1 of spects. Most conspicuously, it lacks the 



Volume 16, Number 1, 2007 



79 



highly derived head structure that char- 
acterizes other mymarommatids. The ver- 
tex, genae and temples appear to be 
uniformly sclerotized and finely, trans- 
versely strigose (Fig. 186). The sculpture 
is similar to that of the cheeks of some 
extant members (Figs 13, 16, 30). The 
vertex lacks a hyperoccipital band of 
pleated membrane and the convex vertex 
merges smoothly into the posterior surface 
of the head (Figs 183, 186), which is 
medially concave and with the occipital 
foramen near its center (Fig. 186). The head 
structure of Galloromma bezonnaisensis is 
uncertain (see below), but our study of 
type material and /or published illustra- 
tions indicate that all other described 
Cretaceous species had the same derived 
head structure as extant mymarommatids 
(Fig. 198; Kozlov and Rasnitsyn 1979, 
figs 8, 9; Fursov et al. 2002, fig. 1). 

As noted by Kozlov and Rasnitsyn 
(1979), the female of P. agapa also has 
a different mandibular structure than other 
mymarommatids. Although closed in the 
holotype, in lateral view the mandibles are 
comparatively thin and the outer surface is 
convex. The apex of the right mandible 
broadly overlaps the left mandible almost 
to its base and each mandible appears to be 
tapered to a point. Because they are closed, 
any internal dentition is not visible. There 
is a distinct transverse region between the 
dorsal surface of the mandibles and the 
oral margin, but we could not determine 
whether there is a labium within this 
apparent "cavity". Extant mymaromma- 
tids are characterized by exodont mand- 
ibles. Each has a comparatively broad 
outer surface with the apex outcurved 
and with two or three externally visible, 
asymmetrical teeth. Furthermore, when 
closed, the apices of the mandibles do not 
overlap and their dorsal margins abut or 
even overlap the oral margin of the head 
capsule (Figs 18, 19, 25-31, 33-39, 47-49, 52, 
54). The mandibles are not clearly visible in 
most fossils, but they are fully open and 
obviously exodont in the holotype of P. 



mandibulatus (Figs 197, 198). The left man- 
dible is completely exposed and has a blunt 
subapical ventral tooth and a much smaller 
subbasal tooth (Fig. 198: sbt). Except for the 
tiny subbasal tooth, the exposed mandible 
of P. mandibulatus is quite similar to that 
described for Mymaromella and Zealar- 
omma. The holotype (PIN: 3311/450) and 
a paratype (PIN: 3311/448) of P. scnonicus 
that we examined have the mandibles 
closed, but their apices are widely separat- 
ed and we agree with the description of 
Kozlov and Rasnitsyn (1979) that the apex 
of the mandibles are bent outwards slight- 
ly. Both specimens appear to have the 
labium separate from the maxillae 
(Fig. 203), but we were unable to distin- 
guish palpi. Bidentate mandibles that are 
widely separated from each other and that 
do not meet in the middle was also given 
as a generic feature of Archaeromma by 
Yoshimoto (1975), which we confirm for A. 
minutissimum and A. nearcticum. The mand- 
ibles of P. japonicum were described as 
having two sharp teeth, which almost 
certainly are exodont based on Fursov et 
al. (2002, fig. 1). The exact structure of 
the mandibles of G. bezonnaisensis is un- 
known. 

The female of P. agapa and extant 
mymarommatids also appear to differ in 
relative position of the toruli. In P. agapa, 
when the head is viewed in profile the 
toruli are near the center of the head and its 
large eyes (Kozlov and Rasnitsyn 1979, 
fig. 10). In extant mymarommatids the 
toruli are near the dorsal margin of the 
eyes, regardless of whether these are re- 
duced in size (Fig. 47) or are as large 
(Figs 25, 33) as those of P. agapa (Fig. 186). 
The holotypes of P. mandibulatus and P. 
senouicus are similar to that of P. agapa 
because they have large eyes (Figs 198, 
203) and the toruli appear to originate at 
about mid height when the head is viewed 
from the left side. However, the apparent 
placement of the toruli can be affected by 
the angle of viewing in amber inclusions, 
as discussed below. 



80 



Journal of Hymenoptera Research 



In addition to the cephalic differences, 
the holotype female of P. agapa has 
a distinctly 13-segmented antenna. When 
the specimen is viewed from its left side 
(Fig. 183), the left flagellum is in profile; 
the right antenna is partly concealed by the 
head but the apical three claval segments 
are exposed in dorsal view. The left 
antenna has seven slender funicular seg- 
ments and four much wider segments that 
form a comparatively loosely associated 
clava (Fig. 180: cl; Kozlov and Rasnitsyn 
1979, fig. 10). The exposed claval segments 
of the right antenna each have a pair of 
quite long and robust sensilla, one above 
the other, projecting from the outer surface. 
Each of the sensilla appear to be strongly 
bent basally and originate near the middle 
of the respective segment (Fig. 181a). The 
claval sensilla on the outer surface of the 
left antenna are not as distinct because of 
the angle of view, but at least the penulti- 
mate segment has two sensilla that origi- 
nate near its middle, one almost at the 
dorsal margin and one midway between 
the midline and ventral margin (Fig. 181b). 
Because of their length these sensilla re- 
semble s2-type sensilla of extant mymar- 
ommatids, but one or both could be 
unusually long s4-type sensilla. The claval 
segments of the left antenna also have 
several very short sensilla along the ventral 
margin (Fig. 181b) that are similar to s3- 
type sensilla of extant mymarommatids (cf. 
Fig. 72). 

Females of other described mymarom- 
matids have a clava composed of 1-4 
segments, but if the clava is 3- or 4- 
segmented then the segments are much 
more compacted than in P. agapa. The clava 
is more tube-like with at least the third and 
fourth segments separated only by quite an 
obscure transverse suture (Figs 192, 196, 
200, 204). Kozlov and Rasnitsyn (1979) 
described the female flagellum of P. seno- 
nicus as having seven funicular and three 
claval segments, and they illustrated 
(fig. 8) a paratype in dorsal view with 
three claval segments. Our study of this 



paratype in ventral view at 200x magnifica- 
tion revealed two distinct basal claval 
segments and a terminal segment that is 
nearly as long as the combined length of the 
two basal segments (Fig. 204). At 400 X 
magnification the terminal segment also 
appears to be subdivided by a transverse 
suture, suggesting four claval segments 
(one or two very fine transverse lines also 
appear to subdivide the "fourth" segment). 
On one side of the clava is a strong, basally 
bent sensillum that extends from the apex of 
each of the two basal claval segments and 
from the "third" segment at the apparent 
suture line. The apical "fourth" segment 
has an additional three strong sensilla on 
the same side. The other side of the clava 
also appears to have strong sensilla, but 
they are less distinct and we are uncertain of 
their number and position. Kozlov and 
Rasnitsyn (1979, fig. 9) illustrated and de- 
scribed a 4-segmented clava for the holo- 
type female of P. mandibulatus. Our study of 
the holotype in ventral view shows that the 
right antenna has seven funicular segments 
and only three distinct, coalesced claval 
segments (Figs 199, 200), though possibly 
there is a small and very poorly differenti- 
ated terminal segment. Seta-like sensilla are 
visible projecting from the claval segments 
(Fig. 200), but not clearly enough to be 
confident of their number. 

Yoshimoto (1975) stated that females of 
Archaeromnia have seven funicular and four 
claval segments. Females identified as 
either A. minutissimum or A. nearcticum in 
the CNC have a distinct clava because it is 
longer and slightly thicker than any funic- 
ular segment, but the claval segments are 
strongly compacted. Consequently, the 
sutures and the number of segments 
comprising the clava are indefinite. At 
least three claval segments are visible 
for some females. The female in CNC: 
CAS-1113 has distinct sutures differentiat- 
ing two basal claval segments and a termi- 
nal part that forms about half the length 
of the clava. This specimen and the female 
in CNC: CAS-119 also has other fine, 



Volume 16, Number 1, 2007 



81 



transverse lines on the terminal claval 
"segment" that suggest possible additional 
segmentation. The basal two claval seg- 
ments of CAS-1113 each have two long, 
basally bent sensilla on one side; there are 
other long sensilla on the terminal part but 
we are uncertain of their exact number. 
The female in CNC: CAS-343 has six long, 
basally bent sensilla on one side of its 
compact clava, whereas other females 
usually have variably long, basally curved 
sensilla on one or both sides. Yoshimoto 
(1975) also described Protooctonus masneri 
as having seven funicular and four claval 
segments. Our study of the holotype 
revealed that the compact clava has only 
three distinct segments (Fig. 196). Al- 
though the clava appears to have more or 
less uniformly long seta-like sensilla in 
some angles of view (Fig. 196), in other 
angles the sensilla are longer on one side 
than the other. 

The specimen that Kozlov and Rasnitsyn 
(1979) described as the male of P. agapa 
differs in several respects from the female. 
The most significant difference is that the 
posterior surface of the head is collapsed 
within the head capsule in a manner 
similar to extant mymarommatids having 
the occipital plate rotated anteriorly. The 
similarity includes a Pi -shaped occipital 
margin and a comparatively wide, in- 
folded ventral margin that together form 
an abruptly margined, acute angle lateral- 
ly. Furthermore, the occipital foramen 
appears to be at the ventral margin of the 
head. The mandibles, although slender 
relative to those of extant mymaromma- 
tids, also appear to have the exterior 
surface slightly outcurved apically, or at 
least the exterior surface is somewhat 
concave medially. The two mandibles are 
angled toward each other such that their 
apices extend slightly beyond the oral 
margin and almost meet medially. The left 
mandible, visible in lateral view, is ex- 
panded ventrally and is differentiated into 
a small, acutely angled submedial tooth 
and a longer, slender, apical tooth, as was 



described and illustrated by Kozlov and 
Rasnitsyn (1979, fig. 11). The antennae are 
13-segmented and typical for male mymar- 
ommatoids, i.e., filiform with all the flagel- 
lar segments having a whorl of long setae 
medially to subapically. The apical four 
antennal segments, particularly the apical 
two, are slightly more strongly coalesced to 
form an inconspicuously differentiated 
clava. Kozlov and Rasnitsyn (1979) also 
noted that the eyes are somewhat larger 
and the first petiole segment markedly 
shorter in the male than in the female; 
furthermore, the ocelli are unusually large. 
Although described from the same amber 
deposit, the differences between the un- 
ique female and male of P. agapa suggest 
that they could represent the opposite 
sexes of two different taxa. 

Species described in Archaeromma and 
Protooctonus are about 75 mya, whereas 
Palaeomymar senonicus, P. mandibulatus and 
P. japonicum are about 85 mya and P. agapa 
is about 95 mya. Only Galloromma bezon- 
naisensis, described from amber dated at 
about 100 mya, is older than P. agapa 
(Table 1). The unique holotype of G. 
bezonnaisensis is in a slender shard of amber 
that is double embedded in clear epoxy 
resin. The first resin block was ground to 
within 0.2 mm of the shard surface (Schlii- 
ter 1978) in which the dorsal surface of the 
specimen and the antennae face up, and 
then was re-embedded so that the speci- 
men is now encased in a comparatively 
large block (12 X 12 X 10 mm) of resin. 
Visibility of the specimen in dorsal view is 
similar to that shown in the illustrations 
provided by Schliiter (1978, fig. 50; plate 5, 
fig. 3; plate 11, fig. 2), though in dorsal 
view artifacts largely conceal all but the 
two antennae and the right forewing 
(Figs 214, 216). In lateral view, a visible 
boundary line between the two resins is so 
close to the fossil that it prevents observa- 
tion of the specimen with the same clarity 
as photographed by Schliiter (1978, plate 5, 
fig. 2). The latter photograph is reproduced 
here as Fig. 215. 



82 Journal of Hymenoptera Research 

The habitus drawings given with the help resolve this problem. In lateral view, 
original description of G. bezonnaisensis the antennae appear to be inserted near the 
illustrate a comparatively thin mandible middle of the eyes as in P. agapa, whereas 
that is acutely tapered both in lateral in dorsal view they appear to be inserted at 
(Schliiter 1978, fig. 49) and dorsal (Schliiter the dorsal margin of the eyes as in 
1978, fig. 50) views. In dorsal view, the mymarommatids with a differentiated oc- 
presumed left mandible (Fig. 217: mnd) cipital plate. Different interpretations can 
projects anteriorly from near the lateral also be given for the comparatively long 
margin of the head capsule. It is compar- ventral surface of the head anterior to the 
atively short and evenly tapered to a point, mesosoma (Schliiter 1978, plate 5, fig. 2). If 
i.e., neither obviously exodont nor curved the "posterior" surface of the head is a flat 
mesally. Because of its position and length, or sunken occipital surface then the posi- 
tive apices of the two mandibles presum- tion of the mesosoma indicates that the 
ably would not meet if they were closed, occipital foramen is near the ventral mar- 
However, the head is not clearly visible gin of this region, similar to extant mymar- 
(Fig. 217) and it is possible that only the tip ommatids. However, based on the lateral 
of the mandible is exposed (cf. Fig. 193). habitus photograph in Schliiter (1978), and 
We cannot confirm the structure of the our own observations, the mesosoma ap- 
mandible drawn in lateral view by Schliiter pears to attach to the head from under the 
(1978, fig. 50), but in the photograph "ventral" surface (Fig. 215). If so, this 
(Fig. 215) it appears to be similar to the supports an hypothesis that both the 
mandible P. agapa. The photograph apparent anterior and "posterior" surfaces 
(Fig. 215) and lateral habitus drawing in constitute the face, and that the head is not 
Schliiter (1978) also show a high-triangular structured as for extant mymarommatids. 
head that dorsally is acutely angled. What The sex of the specimen was not stated and 
appears to be the posterior surface is is ambiguous based on antennal structure, 
obliquely angled and almost straight, and Schliiter (1978) described the antenna as 
therefore could be a flat or collapsed being 14-segmented with 8 funicular and 4 
occipital surface that is acutely angled claval segments. Our study shows that 
relative to a more obviously convex frontal there are four distinct, broadly appressed 
surface. However, if the "posterior" sur- claval segments that are separated by 
face is a flat or collapsed occipital plate oblique sutures (Fig. 214), but only seven 
then the antennal toruli are almost contig- funicular segments (Figs 214, 216). The 
uous with the posterior limit of the vertex claval structure indicates a female antenna. 
(Fig. 215), which is unknown for any other Furthermore, male mymarommatids usu- 
mymarommatid. Such a high placement of ally have most of the flagellar segments of 
the toruli would also suggest the absence similar length and /or these widened me- 
of ocelli, which all other Cretaceous my- dially because of the characteristic medial 
marommatids apparently have. Based on whorl of setae (cf Figs 171, 172 and 
our study of other amber inclusions (see Figs 176, 175), whereas at least the basal 
below) we suspect that the high-triangular four segments of the left flagellum of 
head is partly an artifact of preservation the holotype of G. bezonnaisensis are com- 
and the angle of view, and that both the paratively short and uniformly widened 
anterior and apparent "posterior" surfaces distally (Figs 214, 216). However, the ter- 
comprise the face. If so, the antennal toruli minal segment of the flagellum appears 
are near the center of the head, similar to P. to be smaller than the penultimate seg- 
agapa and at least some other Cretaceous ment (Schliiter 1978, fig. 50) and, as noted 
mymarommatids. The habitus line illustra- by Schliiter (1978), the flagellar segments 
tions in Schliiter (1978, figs 49, 50) do not have quite long and conspicuous setae 



Volume 16, Number 1, 2007 83 

(Fig. 216: fs). These last two features are antenna at the base of the sixth funicular 
more characteristic of a typical male segment and the left half of the head and 
flagellum than that of a female, though mesosoma are partly obscured. This female 
some Cretaceous amber fossil females have is similar to that of P. agapa because in 
quite distinct setae (Figs 196, 204). The dorsal view the vertex is transversely 
illustrations of the antennae in lateral view strigose and uniformly convex, the poste- 
given in Schliiter (1978, fig. 49; plate 5, rior margin of the head lacks an abrupt 
fig. 2; plate 11, fig. 1) show a much more margin and is shallowly concave, and the 
distinctly filiform flagellum (Fig. 215), temples are comparatively long (Fig. 213). 
which is also characteristic of a male In ventral view, the lower posterior surface 
antenna. However, the flagellum likely of the head is distinctly concave between 
appears filiform in lateral view (Fig. 215) the mouthparts and occipital foramen, 
because the claval segments are com- Although the exact structure of the labio- 
pressed rather than cylindrical in cross- maxillary complex is not visible, the 
section. Consequently, their perceived mandibles are open so that their apices 
width depends on the angle of view project anteriorly, distinctly beyond the 
(Fig. 214; cf. Figs 181a, 181b). A short, dark head capsule (Fig. 193). They are conspic- 
region projecting posteriorly from the uously curved and, based on length and 
gaster could be the apex of ovipositor the distance between their bases, it is 
sheaths (Fig. 215: ovs?), but because of obvious that if closed one mandible would 
the condition of the specimen this cannot cross over the other. Each mandible is 
be confirmed. Furthermore, the gaster ap- thicker basally and strongly tapered to 
pears to be oval in cross-section (Fig. 215). a slender point. Although the base of the 
Although gastral shape is not diagnostic, right mandible is obscured by what ap- 
males usually have the gaster flattened pears to be head tissue, a small subbasal 
whereas females more commonly have it tooth is clearly visible on the left mandible 
oval in cross-section. The first petiolar (Fig. 193: sbt). This female also resembles 
segment is only slightly longer than the P. agapa because it has a 13-segmented 
second segment in lateral view (Fig. 215) antenna with four distinct claval segments, 
and therefore is very similar to the structure Both antennae project obliquely in the 
of P. agapa (Fig. 184), though petiolar amber block so that the segments are 
structure does not seem to be sexually foreshortened in dorsal or ventral view 
dimorphic or a generic feature (see further (Fig. 182), but because of the view it is 
below). Finally, the setae along the poster- definite that the claval segments are artic- 
obasal margin of the forewing are all quite ulated. In ventral view, the middle two 
long and of a similar length (Fig. 216). segments of the clava each have one, and 
In addition to the described taxa dis- the apical claval segment two, long, robust, 
cussed above, we studied undescribed basally bent sensilla on either side 
mymarommatoids from GPPC and AMNH (Fig. 182, black arrows). Because of the 
Burmese amber. Burmese amber is dated at oblique view of the antenna it is difficult to 
about 95-105 mya (Grimaldi and Engel be certain where on each segment the 
2005) and therefore the fossils are at least sensilla arise, though the ventral sensilla 
as old as P. agapa and G. bezoimaitensis. The of the middle two claval segments appear 
GPPC material consisted of four inclusions to originate from the apical margin of the 
containing three females and one male. The respective segment (Fig. 182). The basal 
best preserved female (GPPC: HY17A) is in claval segment and the apical funicular 
a clear piece of amber and is visible in both segment also have what appear to be 
dorsal and ventral view, though in dorsal straight sensilla projecting from the re- 
view a vertical crack cuts through the left spective segments (Fig. 182, white arrows). 



84 



Journal of Hymenoptera Research 



These may be a different type or the same 
type as the other sensilla, but differ in 
appearance because of the angle of view. 
The protibial calcars are curved distally 
and have a short inner tine subapically, 
and because of their length resemble the 
long calcars of P. agapa (Fig. 183, insert). 
The relative position of the toruli is un- 
known because the head cannot be ob- 
served in lateral view. We are also un- 
certain of the marginal setal pattern be- 
cause the forewings are partly twisted over 
the body, but there appear to be about 15 
subequally long setae along the posterior 
margin basal to the longer distal marginal 
setae. If accurate, this setal pattern differ- 
entiates the species from P. agapa, which 
has three or four (the two wings differ) 
comparatively long basal setae followed by 
several shorter setae basal to the longer 
distal marginal setae (Fig. 188). The fore- 
wings of P. mandibulatus (Fig. 202), P. scno- 
nicus (Fig. 205) and P. japonicum (Fursov et 
al. 2002, fig. 1) all have a single, long, 
posterobasal seta followed by several 
shorter setae. Because of preservation, the 
forewing marginal setal pattern was not 
visible for all examined Canadian Creta- 
ceous amber specimens. Most specimens 
appear to have a variable number of short 
setae distal to a single long posterobasal 
seta. The holotype of A. nearcticum (CNC: 
CAS-76) has about five short setae distal to 
the long seta, a pattern also possessed by 
one female (CNC: CAS-845) and one male 
(CNC: CAS-748) identified as A. minutissi- 
mum. Another female identified as A. 
minutissimum (CNC: CAS-1160a) clearly 
has nine short setae distal to the long 
posterobasal seta. However, one female 
paratype of A. nearcticum (CNC: CAS-598) 
has three quite long and three shorter setae 
basal to the much longer distal marginal 
setae (Fig. 194), which is similar to the setal 
pattern of the P. agapa female (Fig. 188). 
Consequently, the forewing marginal setal 
pattern either is quite variable in Canadian 
Cretaceous species or the fossils represent 
several species. Regardless, the presence or 



absence of a single long posterobasal seta 
likely is species-specific rather than corre- 
lated with fossil age or lineage. 

A second Burmese female (GPPC: 
HY17B) is also in clear amber but is not 
visible in dorsal view because of the 
dimensions of the amber block and a crack 
above the specimen. In ventral view, the 
antenna has seven funicular segments and 
at least three claval segments that form 
a compact tube. The apical "segment" 
constitutes nearly half the length of the 
clava, but it might be subdivided. Project- 
ing from along one side of the clava are 
four very long and thin, suberect, seta-like 
sensilla (one each on basal two claval 
segments and two on apical half of clava). 
At least the apical two funicular segments 
have similar long setae. Robust, curved 
sensilla are not clearly visible, but shorter, 
more prone setae project from the other 
side of the clava. The protibial calcars are 
curved and an inner tine is not visible. 
Because of this the calcar resembles one of 
the curved setae of the fine comb of the 
basal tarsal segment of extant taxa (cf. 
Fig. 137: fc). The apparent absence of an 
inner tine might be because of the angle of 
view, though a tine is not visible on either 
calcar using either a binocular or com- 
pound microscope. This specimen demon- 
strates that the angle of view can make a big 
difference for interpreting structure within 
amber inclusions. In ventral view, the head 
is obliquely angled with the mandibles 
toward the viewer. In this view, the frontal 
surface of the head is comparatively long 
and thin and the toruli appear to be very 
high on the head, slightly above the dorsal 
margin of the eyes. However, in lateral 
view the head is lenticular and the toruli 
are near the center of the face (the eyes are 
not evident so the position of the toruli 
relative to these is not known). The head is 
observed slightly from the posterior in 
lateral view. The center of the head appears 
to be collapsed and the dorsal and lateral 
posterior margins Pl-like margined. What 
appear to be two, apically tapered, thin 



Volume 16, Number 1, 2007 85 

mandibles project anteriorly, though we segment two, robust, basally curved sen- 
could not determine their structure. The silla along one side. The sensillum of at 
posterior margin of the forewing has seven least the second and third claval segments 
comparatively short marginal setae that originates near the middle of the respective 
gradually increase in length distally. The segment, whereas the two sensilla of the 
third female (GPPC: HY17C) is clouded by apical segment are near the basal and 
sphericles within the amber block and is apical third. At least the second and third 
not clearly visible, but in dorsal view the claval segments have a robust sensillum on 
posterior surface of the head resembles an the other side projecting apically from the 
occipital plate because it appears to be respective segment. The head is quite long 
semicircular and slightly concave. The in frontodorsal view (Fig. 189), but in 
male (GPPC: HY17D) is in clear amber, lateral view it is more or less lenticular 
but only partly visible in lateral view. In with the toruli near its center and that of its 
this view, the head is similar to that of the large eyes (Fig. 190). A uniformly convex 
GPPC: HY17B female except that it is more vertex is not as evident as in P. agnpa, but 
highly convex and the toruli apparently are the vertex is rounded without any in- 
near the center of its large eyes and face, dication of an acute angle or pleated 
The head resembles that of G. bezonnaisen- membrane. In lateral view, the posterior 
sis (Fig. 215) because the antennae origi- surface of the head appears to be concave 
nate at the widest point, though it is not with the occipital foramen near its center 
nearly as acutely triangular. Two apically (Fig. 190). The forewings have 13 or 14 
tapered, comparatively thin and curved comparatively short, subequally long mar- 
mandibles project distinctly beyond the ginal setae along the posterior margin 
head. basally (cf. Fig. 191), similar to the GPPC: 
We examined seven specimens in five HY17A female. The mesosoma, in some- 
Burmese amber inclusions from the what dorsolateral view, has a large circle 
AMNH, of which only one was a female, very near the anterior margin of the 
This female is in a block (AMNH: B-0107) propodeum that may be the spiracle, and 
of resin-embedded amber (see Grimaldi et a groove below it that extends to the 
al. 2002) together with a male. The mand- posterior margin of the propodeum. We 
ibles project anteriorly, distinctly beyond are uncertain of the exact structure of the 
the head, and their apices overlap even propodeal foramen, but in lateral view the 
though they are only partly open (Fig. 189). anterior (basal) flange of the first petiolar 
The apex of right mandible clearly extends segment is clearly visible posterior to what 
beyond the midline of the head (Fig. 189) appears to be a slightly protruding orifice 
and, if closed, would extend to the base of (Fig. 190). The male in the same amber 
the other. In dorsal view, each mandible is piece (Fig. 191) and the female of P. agapa 
tapered apically and curved toward the (Fig. 184) have similar propodeal struc- 
other (Fig. 189) and in lateral view is quite tures. The male is best observed from the 
thin (Fig. 190). Only the left mandible is side opposite the female, which presents 
completely exposed and its inner margin a dorsolateral view of its body (Fig. 191). 
lacks a distinct subbasal tooth, though it The posterior of the head is concave or 
has a subbasal angulation because the base more or less collapsed. A transverse ridge 
is broader than the tapered apical portion, or angulation extends across the head 
The antenna is 13-segmented, including above the occipital foramen, with the 
a clava composed of four broad, distinctly surface ventral to the ridge much more 
separated segments (Figs 189, 190). In strongly concave or collapsed than the 
lateral view, the first three claval segments surface dorsal to it. The surface dorsal to 
each have one, and the apical claval the transverse ridge is somewhat similar to 



86 Journal of Hymenoptera Research 

an occipital plate of extant mymaromma- similar forewing marginal setal pattern, 
tids, but it is much more transverse, Observed from its left side, the propodeum 
concave rather than flat, and lacks evident appears to have a long seta posteromesal to 
pleating or transverse ribs of sculpture, a large circular spiracle similar to extant 
Furthermore, the front of the head imme- members. Observed from the right side, 
diately above the toruli is also slightly a short line extends dorsally from the 
collapsed so that there is a transverse ridge propodeal spiracle, which could be the 
between the eyes. In Chalcidoidea, a col- spiracular peritreme, and forms a convex 
lapsed head, particularly the posterior arc with the dorsal margin of the pleuron 
surface of the head, is common for air- similar to the structure of Mymaromella. 
dried, weakly sclerotized specimens. The From either side, a complete metapleural 
left mandible of the male is partly visible suture and distinct metapleural pit are also 
and it is similar to that of the female except visible. The head, seen in ventral view, is 
for apparently having an inner tooth not collapsed but is distinctly concave 
subbasally (Fig. 191). The presence or similar to that of the P. agapa female. In 
absence of a small inner tooth could be dorsal view, the toruli are near the center 
a sexual or taxon-specific feature or it of its head and large eyes and its man- 
could simply depend on the angle at which dibles are as described for the AMNH: B- 
the mandible is viewed, particularly if the 0107 male. The metasoma also has at least 
inner tooth or angulation projects external- four long setae projecting from its apex, but 
ly, i.e., is somewhat exodont rather than the exact number of setae and their 
flat and in the same plane as the rest of the position is uncertain because they project 
mandible. The wings of the male are more from somewhere under a large overhang- 
clearly visible than for the female and the ing tergite. A third Burmese male in one of 
right forewing has 14 comparatively short two inclusions labelled as AMNH: Bu-184, 
setae along its posterior margin basally seen in dorsolateral view, has the posterior 
(Fig. 191). Both the female and male also surface of the head deeply collapsed such 
have the first petiolar segment conspicu- that the temples appear very long, with 
ously longer than the second petiolar parallel horizontal margins facing dorsally. 
segment (Figs 190, 191). As for other extant The parallel margins extend anteriorly 
and extinct mymarommatids, the male has beyond a small dorsal region that has the 
the anterior margin of the base of the ocelli, and extend ventrally as very thin 
forewing folded under and there is a dis- lateral walls relative to a lower, flat, 
tinct, short marginal vein with at least one horizontal surface. The head structure is 
seta projecting from near its apex. The hind similar to mymarommatids with an occip- 
wing is reduced, apically bifurcate, and ital plate collapsed into the head except for 
clasped to the forewing as described for the differentiated dorsal region bearing the 
extant members. The male also has what ocelli and the carinate posterior margins of 
appears to be a large propodeal spiracle the frontal plate facing dorsally. If it could 
and a very long, apically bifurcate protibial be viewed in profile, the head likely would 
calcar (Fig. 191, insert). have a shape similar to that of G. bezonnai- 
The forewing setal pattern and petiolar scnsis (Fig. 215). Because of the angle of 
structure shared by the AMNH: B-0107 view the mandibles are not visible and we 
female and male suggest that they are the are uncertain of the forewing marginal 
opposite sex of the same species and setal pattern. The specimen has a very 
a different species than P. agapa and G. conspicuous posterior scutellum and a dis- 
bezonnaisensis. Another Burmese male tinct oval region over the base of the 
(AMNH: Bu-160) likely represents a third forewing next to the concave posterolateral 
specimen of the species based on a very margin of the mesoscutum. This structure 



Volume 16, Number 1, 2007 87 

appears to be the same as what we Bu-997 male, the leading margin of the 

interpret as the humeral plate in extant wing curves posteriorly distal to the row of 

mymarommatids. Remnants of muscle dots so that the wing appears constricted 

tissue, including the dorsolongitudinal and there is a distinct, brownish circle from 

and dorsoventral flight muscles, are visi- which a seta projects (cf. Fig. 185: mc), 

ble. A slender, vertical band of muscle similar to extant mymarommatids 

tissue immediately posterior to the dorso- (Figs 127, 128: mc). Kozlov and Rasnitsyn 

ventral flight muscle of the right side (1979) stated that the tip of the forewing 

almost certainly is the mesotergal-meso- vein (equivalent to the brownish circle in 

trochanteral depressor. A separate amber AMNH: Bu-997) was markedly widened 

block under AMNH: Bu-184 has another backwards in the P. agapa female (Fig. 185), 

more poorly preserved male. This male has and that the vein has one medial and three 

the posterior surface of its head conspicu- apical setae, of which two are short. The 

ously concave or collapsed and a smaller, long distal seta in P. agapa (Fig. 185: mc) 

slightly collapsed subtriangular vertical and the AMNH: Bu-997 male is also visible 

region that bears the ocelli. The concave in the holotype of P. mandibulatus (Fig. 201) 

posterior surface is reticulate ventrally, but and several other amber specimens. The 

about its dorsal half is transversely striate. AMNH: Bu-997 male, P. agapa (Fig. 185) 

The head is therefore similar to extant and P. mandibulatus (Fig. 201) also have 

members with a collapsed occipital plate, a longitudinal brownish line extending 

though without a distinct occipital plate, proximally from the seta for nearly one- 

Although the mandibles are not clearly third the distance to the body, which we 

visible in ventral view they do not appear interpret as the submarginal vein (cf. 

to be exodont. The forewing marginal setae Figs 95, 185: smv), and faint, almost spec- 

also are not clearly visible. A fifth male tral parallel lines along the posterior 

(AMNH: Bu-997) has the posterior of the margin that we interpret as the retinacu- 

head deeply collapsed behind the ocelli lum. The holotype of P. mandibulatus very 

and the toruli apparently near the middle clearly has a slender retinaculum that 

of the head. The forewings have seven comprises about the posterior half of its 

setae along the posterior margin basally forewing basally, and a stalk-like hind 

that are obviously shorter than the more wing with a long seta basally and an apical 

distal marginal setae. These seven setae process inserted into the retinaculum 

differ in length, the more basal and apical (Fig. 201). The sixth Burmese male 

setae being longer than the medial setae. (AMNH: Bu-479) is visible from both sides, 

The left forewing is exceptionally well with the head either in fronto- or postero- 

preserved. Adjacent to the body is a brown- lateral view. In frontolateral view, the 

ish sclerite that is acutely angled distally, toruli are within the dorsal third of the 

which we interpret as the humeral plate, head, almost at the level of the dorsal 

This is separated by a narrow hyaline margin of the eyes. The dorsal and lateral 

"break" from a slender brownish band posterior margins of the head appear to 

that extends distally along the leading be abrupt, though in posterolateral view 

margin of the wing. Along this brown line the structure and the extent to which the 

is a row of "dots" similar to slide prepara- posterior surface of the head is collapsed is 

tions of some extant species (Fig. 167). not clear. Neither the mandibles nor 

Although not mentioned by Kozlov and the forewing marginal setal pattern are 

Rasnitsyn (1979), the forewings of P. agapa visible. 

(Fig. 185) and P. mandibulatus (Fig. 201) In addition to Burmese amber, we 

also have a row of distinct dots basally examined three males from New Jersey 

along the leading margin. In the AMNH: amber (about 90 mya vide Grimaldi and 



Journal of Hymenoptera Research 



Engel 2005). All the specimens (AMNH: 
NJ-179, AMNH: NJ-686a, AMNH: NJ-1005) 
have the posterior of the head Pi -like 
carinate, with the posterior surface col- 
lapsed similar to extant members with the 
occipital plate rotated anteriorly (Fig. 15). 
Two of the specimens (NJ-179, NJ-1005) 
have the protibial calcars visible. They are 
slightly curved and obviously shorter than 
those of the Burmese amber specimens. 
The left calcar of the NJ-1005 specimen has 
a fine inner tine originating quite far from 
its apex (cf. Fig. 140). Although the speci- 
mens are not well preserved, NJ-179 and 
NJ-686a appear to have bidentate, exodont 
mandibles. At least NJ-179 has ocelli, 
a suture separating the meso- and meta- 
pleura, and a structure of the propodeal 
foramen similar to that described for P. 
agapa and the Burmese amber fossils. The 
forewings of NJ-686a are not visible, but 
NJ-179 and NJ-1005 have a conspicuously 
long posterobasal marginal seta separated 
by several very short setae from long 
marginal setae apically (Fig. 195). Conse- 
quently, the three New Jersey Cretaceous 
amber specimens appear to be more 
similar to P. japonicum, P. mandibulatus, P. 
senonicus and at least some Canadian 
Cretaceous amber specimens than to most 
Burmese amber fossils. 

Alonso et al. (2000) stated they had 
a male from lower Cretaceous Spanish 
amber (about 115 mya vide Grimaldi and 
Engel 2005) that is intermediate in struc- 
ture between Mymarommatidae and My- 
maridae. The specimen has a 2-segmented 
petiole but apparently lacks all the other 
features of Mymarommatidae. Among oth- 
er features, they described a forewing with 
a submarginal vein, a small alar membrane 
and short marginal setae, and an 11- 
segmented flagellum lacking multiporous 
plate sensilla. Because of these features 
they suggested that the specimen might be 
considered as belonging to the sister-group 
of Mymaridae + Mymarommatidae. They 
did not illustrate the specimen and we 
were unable to obtain it for study. 



SYSTEMATICS 

The distribution of 28 variable features 
that are potentially informative for estab- 
lishing monophyly and relationships of 
supraspecifc taxa in Mymarommatoidea 
is given in Table 2. Mymaromma sp. 9 is 
treated separately from other Mymaromma 
because it differs in several features from 
other examined species of the genus. 
Zealaromma insulare and Z. valentinei are 
also treated separately because these two 
species differ in several features. The 
matrix codes and character states are 
described in Appendix III. State desig- 
nates the hypothesized plesiomorphic state 
for 22 of the characters (1-3, 6-11, 14-24, 
26-28). Polarity decisions for these char- 
acters are based on observed states in 
Galloromma and /or other parasitic Hyme- 
noptera. Characters 4, 5, 12, 13 and 25 
represent features of extant mymaromma- 
tids that are too small to be observed 
confidently in fossils and for which other 
Hymenoptera cannot be used to hypothe- 
size polarity. Our hypotheses of generic 
relationships and character-state evolution 
discussed below are illustrated by Fig. 1. 

Suprageneric classification. — As far as can 
be determined, the oldest mymaromma- 
toids that we examined had external 
structures of the mesosoma and metasoma 
similar to extant species. They also had the 
same highly derived forewing and hind 
wing structures of extant mymarommatids 
(cf. Figs 129, 167 with 185, 194, 201) and 
likely lacked meso- and metatibial spurs, 
though we cannot be certain whether tibial 
spines visible in some fossils are pseudo- 
spurs or true tibial spurs. Tubular meso- 
tergal-trochanteral muscles originating 
from rod-like axillar phragmata are visible 
in some Tertiary and Cretaceous (Fig. 206) 
fossils, but our study was insufficient to 
evaluate other internal features that Vil- 
helmsen and Krogmann (2006) proposed 
as autapomorphies of Mymarommatoidea 
(propleural and profurcal arms fused, 
prophragma with rods, metafurca absent). 



Volume 16, Number 1, 2007 



89 



r^. ,v- p-. o i—i ■— < o o 



rv. (v. r^. q 



O O 



P-- P- P- O O O i— i O 



P-. P-. P-. O 





O 


^H 


CN 


cs 


{N 


o 


i— i 


i— i 


r \ 




o 




^ 






o 




ri 


O 


o 


o 


o 


O 


o 


o 


- 


(N 


p.. 


P-. 


p-. 


o 


- 


- 


o 


o 


s 


o 


o 


o 


o 


o 


o 


- 


- 


GC 


o 


r^. 


rv. 


1— I 


C3 


ta 


o 


o 












C 


r^ 






In 


o 


(^. 


r^- 


1 — 1 


i — i 


i — i 


o 


o 










o 










£ 


p.. 


P-. 


r^. 


<N 


cn 


o 


1 — 1 


- 


LT 


o 


o 


O 


O 


i — i 


i — > 


- 


o 


s 


o 


o 


- 


o 


o 


o 


^H 


1 — 1 


r<-, 


P" 


P-. 


p- 


1 — 1 


o 


o 


o 


o 


r i 


p- 


P" 


P-- 


1 — 1 


r0 


o 


o 


o 










o 












o 


o 


i — i 


o 


o 


o 


^H 


o 


1 ' 














o 




— 


o 


,_, 


IS 


nj 


,_ 


,_ 


CO 


ro 


' 






£ 


£ 






r; 


C 




o 


H 


ppj 


CD 


n 


<N 


P"i 


ro 






o 














X 


f^« 


P" 


p.. 


o 


- 


- 


- 


p— 


>• 


r*-« 


o 


O 


o 


o 


o 


-—I 


O 




o 


o 


O 


rH 


1 — 1 


o 


1 1 


rH 










O 






o 






r^- 


p- 


p.. 


,_| 


o 


o 


T— < 


| 1 


ir. 








o 










-T 


r-. 


P-. 


P-. 


T-H 


o 


o 


O 


o 


« 


o 


P-- 


° 


o 


o 


o 


- 


o 


cs 


O 








01 

-2 






s 


L- 










3 








01 










U3 






r, 


CJ 










£ 






Cfl 


5- 


fC 


c 


03 


-2 


(0 


,__ 


rG 




G 
c 

Q 
X 


£ 

g 

Q 


£ 

5 

0) 


£ 
£ 


"a! 

£ 

o 


£ 
£ 


C 
•4= 

c 


C 

£ 



Fh 


£ 
o 




(H 


03 


q 


ns 


Fh 


10 


to 


H 


Q 


x; 
u 


13 

Oh 


£ 


-5 
N 


13 
> 


£ 


£ 




<j 


< 


2 


N 


S 


S 



Contrary to Gibson et al. (1999), fusion of 
the propleura into a carapace (14: 1) and 
fusion of the mesopleuron, metapleuron 
and propodeum below the spiracle (15: 1, 
16: 1-3) are not groundplan features. 
Furthermore, the holotype of P. agapa and 
some Burmese amber fossils lack the bi- 
zarre head structure, exodont mandibles, 
and female claval structures of extant 
mymarommatids. These Cretaceous fossils 
indicate that the mymarommatoid ground- 
plan structure included a uniformly scler- 
otized head capsule (1: 0), non-exodont 
mandibles (2: 0), and a 13-segmented 
female antenna with 7 slender funicular 
segments and 4 wider apical segments that 
formed a loosely associated clava because 
the segments were distinctly separated (10: 
0; Figs 180, 189, 190). The fossil record 
indicates a transformation series in which 
the four claval segments first became 
closely associated to form a compact tube 
(10: 1; Figs 71, 192, 196, 200, 204) prior to 
the apical two segments, and ultimately all 
the segments, fusing together to form 
a single claval segment in Tertiary and 
most extant mymarommatids. Cretaceous 
females that we classify in Archaeromma 
only have three distinct claval segments 
(Figs 196, 200, 204), but some species may 
have had a compact 4-segmented clava. 

The cephalic, mandibular, and antennal 
structures of the holotype of P. agapa 
strongly support classifying this species 
in a separate genus that constitutes the 
sister group of all other mymarommatoids. 
The genus could simply be classified as 
one more genus in Mymarommatidae or 
because of its postulated relationships with 
other mymarommatids be recognized as 
a higher taxon equivalent to all other 
mymarommatids. This higher taxon could 
be a separate subfamily in Mymarommati- 
dae or a family equal to Mymarommatidae. 
Either of the latter two classifications imply 
a sister-group relationship and a greater 
morphological "gap" between it and the 
other recognized genera, which would not 
be implicit if the genus was simply 



90 



Journal of Hymenoptera Research 



Galloromma 




1 2 10 
1 1 1 



Archaeromma 



8 15 17 20 24 26 27 
1111111 




6 16 22 

-0-+0 

1 3 2 



Zealaromma insulare 



Zealaromma valentinei 



4 5 6 12 13 16 17 18 22 

rfOXl # # Q#-D- Mymaromella 

111112 112 



28 

1 



14 23 
1 1 



11 22 
1 2 


Palaeomymar 




Mymaromma 


8 16 19 


1 SP- 9 


1 1 1 


3 6 ' 11 15 25 
L #-0#-00#- Mymaromma 
111111 



Fig. 1. Hypothesized relationships and character-state evolution among genera of Mymarommatoidea. Extant 
taxa in italics. Symbols used: • = synapomorphic feature, = homoplastic feature, □ = feature derived 
within taxon. 



classified as one of several genera within 
Mymarommatidae. For these reasons, we 
prefer to classify the genus containing P. 
agapa separately within a higher taxon. 
Because the bellows-like head of mymar- 
ommatids is unique within Hymenoptera, 
and so bizarre, we prefer not to change the 
existing morphological family concept of 
extant Mymarommatidae. We therefore 
recognize two families in Mymarommatoi- 
dea, one of which is extinct. Below, we 
establish a new family for the genus 
containing P. agapa and, except for G. 
bezonnaisensis, classify all other described 
Cretaceous, Tertiary and extant mymar- 
ommatoids in Mymarommatidae. This 
family classification will have to be re- 
evaluated following study of the male from 
lower Cretaceous Spanish amber that was 
discussed by Alonso et al. (2000). If they 
described the features of this male accu- 
rately, then it could represent the sister 
taxon of Gallorommatidae + Mymaromma- 
tidae. 

A remaining quandary is whether the 
genus containing P. agapa is new or 
congeneric with Galloromma. The head 



structure of the unique specimen of G. 
bezonnaisensis cannot be determined. Be- 
cause its head is conspicuously high-tri- 
angular (angular) in profile (Fig. 215), it 
seems unlikely that the vertex and temples 
are smoothly rounded into the occiput as 
in P. agapa unless the apparent shape is an 
artifact of preservation and /or the angle of 
view. Our study of other amber fossils 
shows that the angle of view can affect 
apparent head shape. As discussed above, 
the mesosoma appears to attach to the 
"ventral" surface of the head (Fig. 215) 
and, if so, indicates G. bezonnaisensis had 
a uniformly sclerotized head capsule. The 
structure of the single visible mandible is 
uncertain, but in lateral view (Fig. 215) it 
appears to be thin and tapered (cf. 
Fig. 190). Although the sex of the holotype 
is also uncertain, in dorsal view the 
flagellum appears to have 4 distinct claval 
segments (Fig. 214). The premise that the 
specimen is a female may be supported by 
the gaster not being flattened and by what 
might be ovipositor sheaths projecting 
posteriorly from the gaster (Fig. 215: 
ovs?). The posterior margin of the forewing 



Volume 16, Number 1, 2007 



91 



has several comparatively long setae ba- 
sally (Fig. 216) similar to P. agapa (Fig. 188) 
and some undescribed Burmese amber 
fossils (Fig. 191) that we consider to belong 
to the same genus. Finally, the petiolar 
segments of G. bezonnaisensis (Fig. 215) and 
P. agapa (Fig. 184) are of similar length, 
though this is not a generic feature. 

We base our morphological concepts of 
the taxon containing P. agapa mostly on the 
holotype of P. agapa, supplemented by 
similar specimens in Burmese amber (par- 
ticularly GPPC: HY17A and AMNH: B- 
0107). Additional, better preserved speci- 
mens from French Bezonnais amber are 
required to definitively resolve the generic 
status of G. bezonnaisensis relative to P. 
agapa. Because the two taxa could be 
congeneric, we transfer P. agapa to Gallor- 
omma and classify Galloroimna and its two 
currently described species, G. bezonnaisen- 
sis and G. agapa (Kozlov and Rasnitsyn) n. 
comb, in the family Gallorommatidae n. 
fam. We classify other extinct and extant 
mymarommatoids in Mymarommatidae 
(Table 1), as justified below. 

Classification of Mymarommatidae. — No 
known autapomorphies support mono- 
phyly of ArcJiaeromma and our study was 
insufficient to differentiate males from 
those of Tertiary and extant mymaromma- 
tid genera. We distinguish ArcJiaeromma 
only by females having three or four 
compact claval segments (9: 0/1). The 
maxillae and labium may have been 
completely or almost completed separated 
in Archaeromma (Fig. 203), but this struc- 
ture undoubtedly is symplesiomorphic. 
Furthermore, some species of Mymaromella 
have the maxillae and labium differentiat- 
ed over much of their length (Fig. 44). Of 
the extant genera, monophyly of Zealar- 
omma is well supported by the absence of 
metasomal spiracles (26: 1) and cerci (27: 
1), presence of bumps on the posterior 
surface of the femora, particularly the 
mesofemur (24: 1), and likely by scleroti- 
zation (secondary) of the spiracular peri- 
treme (20: 1). The 2-segmented female 



clava (9: 2) of Zealaromma is enigmatic. 
The absence of both s4-type and s3-type 
sensilla from the basal claval segment 
(Fig. 71) might indicate that this segment 
evolved through autapomorphic second- 
ary subdivision of a 1-segmented clava. 
Alternatively, the female claval structure 
could indicate Zealaromma constitutes 
a clade that is basal to Palaeomymar and 
other known Tertiary and extant species. 
Presence of external parameres in male 
Zealaromma (28: 0; Figs 154-156) and their 
absence from males of Mymaromma and 
Mymaromella (28: 1) indicates that Zealar- 
omma is at least the sister taxon of these 
two extant genera. The absence of para- 
meres recorded for Palaeomymar (28: 1) is 
based on only a single fossil and requires 
verification. The separate metanotum of Z. 
valentinei (16: 0; Figs 105, 106) may repre- 
sent a third uniquely retained symplesio- 
morphy among extant mymarommatids. 
The two different structures of the meta- 
thoracic-propodeal complex that character- 
ize Mymaromma (16: 1; Figs 82-84) and 
Mymaromella (16: 2; Figs 97-102), and the 
autapomorphic structure of Z. insulare (16: 
3; Figs 109, 110), are hypothesized to have 
evolved through independent fusions of 
the metanotum with one or both of the 
metapleuron and propodeum. Among ex- 
tant species, the posterior margin of the 
forewing of Z. valentinei also uniquely has 
several quite long setae basally (22: 0; 
Fig. 165). This setal pattern is similar to 
that of Galloromma (Fig. 188) and some 
Archaeromma (Fig. 194), suggesting a fourth 
uniquely retained symplesiomorphy 
among extant mymarommatids. We con- 
sider the structures discussed above as 
strong evidence that Zealaromma represents 
a monophyletic clade basal to Palaeomymar, 
Mymaromma and Mymaromella. The two 
known species of Zealaromma share a more 
internalized galea with Mymaromma (8: 1) 
and a completely fused meso- and meta- 
pleuron (15: 1) with most Mymaromma. 
They also have four supraclypeal setae (5: 
0) and lack a metapleural pit (17: 1) similar 



92 Journal of Hymenoptera Research 

to most Mymaromella, but we consider is indicated as the sister species of all other 

these shared states either as symplesio- Mymaromma based on the features dis- 

morphies or independently derived (see cussed above and is uniquely characterized 

further below). by an extremely long, slit-like spiracular 

The best evidence for monophyly of aperture (21: 1; Fig. 82: spa). 
Mymaromma is the presence of a fl-like The monophyly of Mymaromella is not 
propodeal flange (19: 1; Figs 81, 87) that well established. Fusion of the metanotum 
surrounds the insertion of the first petiolar posterolaterally to the propodeum (16:2) 
segment into the propodeal foramen, may support monophyly, but the exact 
Monophyly may also be supported by structure of the metanotal-propodeal corn- 
fusion of the metanotum laterally with plex of P. succini and other fossils is 
the metapleuron (16: 1), which results in unknown. The presence of paramedial 
the spiracular peritreme forming an abrupt setae on the occipital plate (4: 1; Figs 45, 
angle with the dorsal margin of the 46) and a metapleural pit about midway 
metapleuron (Figs 82-84). Although the between the ventral margin of the pleuron 
holotype of G. agapa lacks this structure and propodeal spiracle (18:1; Fig. 97: pl^p) 
(Fig. 187), the spiracular peritreme is not could also support monophyly, but again 
clearly visible in P. succini or most other these features could not be observed 
fossils. Species of Mymaromma, except for confidently in fossil mymarommatids and 
M. sp. 9, share two other apparently polarity is uncertain. Among extant my- 
autapomorphic features — tridentate marommatids, Mymaromella is most vari- 
mandibles (3: 1; Figs 25-30) and a broad able in the number and position of s4-type 
labium relative to the maxillae (7: 1; sensilla. Females of Zealaromma (Fig. 71) 
Figs 18-21). Individuals of M. sp. 9 have and Mymaromma (Figs 58, 59) have two s4- 
bidentate mandibles (Fig. 31) and a com- type sensilla (12: 0) near the dorsal margin 
paratively narrow labium (Fig. 31, insert), of the outer surface (13: 0), whereas 
Mymaromma sp. 9 is also the only known Mymaromella females have two or three 
species of Mymaromma having the meso- such sensilla (12: 1; Figs 62, 63, 65, 67, 69) 
and metapleuron completely separated by more or less medially (13: 1; Figs 63, 65, 67, 
a suture (15: 0; Fig. 82), but undoubtedly 69) or sometimes in the dorsal third 
this is symplesiomorphic. A very fine (Fig. 62). These character-state distributions 
suture also divides the meso- and meta- suggest that the more ventral position of the 
pleuron ventrally in M. sp. 7 (Fig. 87), sensilla is synapomorphic for Mymaromella 
which indicates fusion of the meso- and and that presence of a third sensillum 
metapleuron in Mymaromma evolved in- supports a monophyletic subgroup of 
dependency to similar fusion in Zealar- Mymaromella. These hypotheses could be 
omnia (Figs 103, 108). Finally, M. sp. 9 is the tested by a phylogenetic analysis of the 
only observed species of Mymaromma species of Mymaromella and by determining 
having lateral setae on the first petiolar the number and position of s4-type sensilla 
segment (state 25: 0; cf. Figs 92, 93, 104). in fossil taxa. At least the apical three claval 
However, because of the condition of segments of female Galloromma have long, 
specimens we were unable to survey this basally curved, robust sensilla (Figs 181, 
feature comprehensively and are not con- 182). Because of their length, these sensilla 
fident of the character-state distribution most closely resemble s2-type sensilla of 
given for Mymaromma and Mymaromella. extant mymarommatids (Figs 63, 70), but 
We did not observe petiolar setae in any further study is necessary to determine if 
fossil mymarommatid, but are uncertain any of the sensilla originate from circular 
whether this is because the setae are absent depressions. If so, they likely are homolo- 
or too small to be visible. Mymaromma sp. 9 gous with s4-type sensilla. 



Volume 16, Number 1, 2007 93 

The relationships of Palaeomymar with could be established more confidently by 
Mymaromma and Mymaromella are also not further study to provide missing data, 
well substantiated. Palaeomymar is recog- particularly the exact structure of its meta- 
nized as a separate genus because P. succini thoracic-propodeal complex, 
has a short, straight protibial calcar (23: 1; Based on the above analysis, we current- 
Fig. 210) similar to Mymaromma (Figs 142, ly consider the likely generic relationships 
144a, 145a), but lacks the D-like propodeal of Mymarommatidae as Archaeromma + 
flange (19: 0) that supports monophyly of (Zealaromma + (Mymaromella + (Palaeomy- 
Mymaromma. Individuals of P. succini also mar + Mymaromma))) (Fig. 1). 
have all the setae along the posterior 

margin of the forewing short basally (22: Mymarommatoidea 
2; Figs 208, 209) similar to some species of 

Mymaromella (Figs 117, 166). A single con- Description— Antennae geniculate, in- 
spicuously long posterobasal marginal seta serted subcontiguously at or above middle 
(22: 1) is possessed by some species of of face; flagellum without multiporous 
Archaeromma (Figs 195, 202), other Tertiary plate sensilla. Female antenna with 9-13 
fossils (Fig. 211), some Mymaromella segments including a variably structured 
(Figs 118, 119), and known Mymaromma clava composed of 1-4 segments; clava 
(Figs 113, 114, 163) except M. sp. 10. with short, peg-like sensilla along ventral 
Because of this character-state distribution, midline (Figs 72, 181b), long, robust, basal- 
we consider the lack of a conspicuously ly curved sensilla on dorsal and /or outer 
long posterobasal marginal seta to be surface (Figs 58, 63, 181, 182), and at least 
derived in P. succini and some Mymaro- extant members with 2 or 3 short, basally 
mella. If this shared feature results from curved lanceolate sensilla originating from 
common ancestry then the reduced fore- depressions on outer surface (Figs 59, 
tibial calcar shared by P. succini and 65, 69, 70). Male antenna 12- or 13- 
Mymaromma must be independently de- segmented, with the apical 2-4 segments 
rived. We consider it more likely that a long often somewhat coalesced as an indistinct 
posterobasal marginal seta was lost in- clava, but the flagellum more or less 
dependently from P. succini and within filiform; fl 9 and fl 10 or fl 8 — fl 10 each with 
Mymaromella and that a reduced protibial short, basally curved lanceolate sensillum 
calcar supports monophyly of Palaeomymar originating from depression distally. Ocelli 
+ Mymaromma. A phylogenetic analysis of present or absent. Pronotum with postero- 
the extant species of Mymaromella to dorsal margin not extending to base of 
establish the groundplan state of the forewing and its posterior margin not 
marginal setae could test these hypotheses, rigidly interlocked with anterior margin 
Monophyly of Palaeomymar + Mymaromma of mesopleuron. Propleura forming ex- 
may also be supported by fusion of the posed lateral and ventral portions of 
propleura into a carapace (14: 1; Fig. 18), propectus, with ventral margins abutting 
though further study is necessary to verify medially or fused to varying extent. Pro- 
character-state distribution in Palaeomymar sternum vertical, largely concealed between 
as well as Mymaromma and Mymaromella. A posterior margin of propleura and base of 
6-segmented funicle characterizes females procoxae. Functional mesothoracic spiracle 
and males of Mymaromma goethei (11: 1; absent. Prepectus not externally visible. 
Figs 174-176) and at least females of P. Tegula absent. Mesoscutum with distinc- 
succini (Fig. 179). We consider that this five scabrous sculpture, without notauli or 
similarity likely results from independent median mesoscutal sulcus; scutellum corn- 
loss of a funicular segment in the two taxa. posed of convex anterior scutellum and 
The phylogenetic position of P. succini slightly concave, transverse, longitudinally 



94 



Journal of Hymenoptera Research 



strigose posterior scutellum; anterior scu- 
tellum without differentiated dorsal axil- 
lar regions. Metanotum an independent 
sclerite or fused with one or both of meta- 
pleuron and propodeum. Mesopleuron 
and metapleuron separated by oblique 
suture or partly to entirely fused. Forewing 
humeral plate bare. Forewing pedunculate, 
stalk-like basally and disc broadly spatu- 
late to lanceolate; disc membrane with 
mesh-like pattern formed by raised linea- 
tions on both surfaces; marginal setae, 
when long, arising distinctly from within 
periphery of disc apically; venation strong- 
ly reduced within basal quarter of wing, 
consisting of submarginal, marginal and 
very short stigmal vein delimited apically 
by a campaniform sensilla and projecting 
seta. Hind wing stalk-like, terminated by 



single hamulus and opposing projection 
that together form pincer-like structure. 
Legs with tarsi 5-segmented; protibial 
calcar straight and simple or curved and 
bifurcate; meso- and metatibiae of at least 
extant members without tibial spurs; me- 
socoxa with basicoxite reduced to small 
dorsal lobe projecting into mesocoxal fora- 
men, and mesotrochantinal lobe not exter- 
nally visible. Metasoma 8-segmented with 
first two segments tubular; post-petiolar 
segments with terga broadly overlapping 
sterna without differentiated laterotergites. 
Female with hypopygium extending to 
apex of metasoma and therefore often 
concealing ovipositor, but apex capable of 
wide separation from syntergum; oviposi- 
tor non-telescoping, extended ventrally 
from gaster by rotation. 



KEY TO FAMILIES AND GENERA OF MYMAROMMATOIDEA 



1. Cretaceous; female flagellum with 7 funicular segments and 3- or 4-segmented clava 

(Figs 180, 182, 189, 190, 196, 199, 200, 204) 2 

Tertiary or extant; female flagellum with 6 or 7 funicular segments and 1 -segmented 
clava (Figs 173-175, 178, 179) or 7 funicular segments and 2-segmented clava 
(Figs 71, 172) (Mymarommatidae, part) 3 

2(1) Head capsule uniformly sclerotized and with vertex smoothly rounded into concave 
occiput (Fig. 186); mandibles not exodont (Figs 189, 190, 191, 193), thin, curved 
mesally and broadly overlapping when closed; clava of female consisting of 4 

distinctly separated segments (Figs 180, 182, 189, 190) 

Galloromma Schliiter (Gallorommatidae n. fam.) 

Head capsule with hyperoccipital band of pleated membrane separating frontal plate 
from flat occipital plate (cf. Figs 13, 14) or frontal plate abruptly angled or hood-like 
relative to occipital plate if this rotated anteriorly (Fig. 198, cf. Fig. 15); mandibles 
exodont (Figs 197, 198), comparatively broad with concave outer surface and apices 
not meeting medially when closed (cf. Figs 31, 47-49); clava of female consisting of 
3 or 4 segments, but these separated by linear sutures and forming compact tube 
(Figs 192, 196, 199, 200, 204) Archaeromma Yoshimoto (Mymarommatidae) 

3(1) Protibial calcar straight, simple (Figs 142, 144a, 145a); extant taxa with posterior margin 
of propodeum extending as H-like flange (Figs 86, 87) and concealing propodeal 

foramen in dorsal (Figs 78, 79) and lateral (Figs 80, 83) view 4 

Protibial calcar apically curved and bifurcate (Figs 137-139a, 140); extant taxa with 
posterior margin of propodeum extending as variably distinct flange only laterally 
on either side of short, protruding, tubular foramen (Figs 90-97, 104, 107) 5 

4(3) Posterior margin of forewing with several short setae separating conspicuously long 
basal seta from longer apical setae (Figs 113-115); propodeum with Pl-like flange 
(Figs 86, 87) concealing foramen in dorsal and lateral view; mesopleuron and 
metapleuron usually partially (Fig. 87) or completely fused (Fig. 80), only very rarely 
completely separated by oblique suture (Fig. 82). Extant Mymaromma Girault 



Volume 16, Number 1, 2007 



95 



Posterior margin of forewing with all setae short basally (Figs 207-209); propodeum 
without distinct H-like flange, the foramen projecting as short tube anterior to 
petiolar insertion {cf. Fig. 94); mesopleuron and metapleuron separated by oblique 

suture (cf. Figs 97, 99). Extinct (Baltic amber) Palaeomymar Meunier 

5(3) Mesopleuron and metapleuron fused together (Figs 103, 108); Mt 7 without spiracles 
(Fig. 154); mesofemur with row of bumps on posterior surface (Fig. 146b); female 

clava 2-segmented (Fig. 172) Zealaromma n. gen. 

Mesopleuron separated from metapleuron by diagonal suture (Figs 95- 97, 99, 101); 
Mt 7 with spiracles (Figs 149-152, 158); mesofemur without bumps on posterior 
surface; female clava 1-segmented (Fig. 178) Mymaromella Girault 



Gallorommatidae n. fam. 

Type genus. — Galloromma Schliiter, pres- 
ent designation. 

Diagnosis. — Distinguished from Mymar- 
ommatidae by the following features. Head 
capsule uniformly sclerotized and sculp- 
tured, with convex vertex smoothly round- 
ed into medially concave occiput (Figs 186, 
213); occipital foramen originating from 
middle of occiput. Mandible not exodont 
(Figs 189, 190, 193); when closed, apices of 
mandibles broadly overlapping (Fig. 189) 
and dorsal margin distinctly separated 
from oral margin of head capsule. Female 
antenna with 13 distinct segments, includ- 
ing 7 funicular segments and 4 larger 
apical segments that are distinctly separat- 
ed to form a loose clava (Figs 180, 182, 
189, 190, 214). 

Galloromma Schliiter 

Galloromma Schliiter, 1978: 74-76. Type species: 
Galloromma bezonnaisensis Schliiter; original 
designation. 

Description. — In addition to the diagnos- 
tic features given above for the family, 
other features that are not visible in all 
inclusions but that likely characterize Gal- 
loroiniua are as follows. Head with ocelli 
(Fig. 186); eye large, with numerous om- 
matidia (Fig. 186); toruli near level of 
center of head and eyes (Fig. 183); mandi- 
ble with subbasal tooth (Figs 191, 193). 
Protibial calcar conspicuously long, 



curved, and apically bifurcate (insert, 
Figs 183, 191). Propleura divided medially. 
Mesopleuron and metapleuron separated 
by complete suture; metapleuron with 
large pit about midway between propodeal 
spiracle and ventral margin of pleuron 
(Fig. 187). Propodeum with spiracle near 
posterior margin of scutellum (Fig. 187) 
and without distinct Pi -like flange sur- 
rounding foramen (Figs 184, 187). Fore- 
wing disc spatulate (Fig. 183) with 
two longitudinal folds, the membrane with 
comparatively numerous, scattered, 
short setae; posterior margin with setae 
all of similar length basally (Figs 191, 216) 
or with at least three moderately long 
setae proximal to shorter setae (Figs 188, 
194). 

Included species. — Galloromma agapa (Kozlov 
and Rasnitsyn), n. comb. Palaeomymar agapa 
Kozlov and Rasnitsyn, 1979: 414-415. Holotype 
9, PIN (examined). 

Galloromma bezonnaisensis Schliiter, 1978: 74- 
76. Holotype 9, ZMB (examined). 

Mymarommatidae 

Type genus. — Mymaromma Girault, 1920: 
38. 

Diagnosis. — Distinguished from Gallor- 
ommatidae by the following features. Head 
capsule with D-shaped hyperoccipital 
band of pleated membrane differentiating 
flat occipital plate from convex frontal 
plate (Figs 13, 14, 42); occipital foramen 



96 



Journal of Hymenoptera Research 



originating at ventral margin of occipital 
plate. Mandibles exodont; when closed, 
their apices not meeting and dorsal mar- 
gins abutting oral margin of head capsule 
(Figs 28-31, 47-49). Female antenna with 
9-11 distinct segments, including 6 or 7 
funicular segments and 1 or 2 claval 
segments (Figs 172-175, 178, 179) or if with 
3 or 4 claval segments then these forming 
a compact tube with only basal two 
segments separated by distinct suture 
(Figs 192, 196, 200, 204). 

Remarks.— Huber (2005) discussed the 
derivation and gender of the generic 
names. 

Archaeromma Yoshimoto 

Archaeromma Yoshimoto, 1975: 503. Type spe- 
cies: Ooctonus minutissimus Brues; designated 
by Yoshimoto (1975: 503). 

Protooctonus Yoshimoto, 1975: 511. Type species: 
Protooctonus masneri Yoshimoto; original des- 
ignation, n. syn. 

Diagnosis. — Females are distinguished 
from other Mymarommatidae by their 
compact clava consisting of at least three, 
sometimes possibly four, coalesced seg- 
ments. 

Included species. — Archaeromma japonicum 
(Fursov, Shirota, Nomiya and Yamagishi), n. 
comb. Palaeomymar japonicum Fursov, Shirota, 
Nomiya and Yamagishi, 2002: 52-54. Holotype 
S, Y. Shirota collection, Laboratory of Evolu- 
tional Ecology, Faculty of Agriculture and Life 
Sciences, Hirosaki University, Japan. 

Archaeromma mandibulatum (Kozlov and Ras- 
nitsyn), n. comb. Palaeomymar mandibulatus 
Kozlov and Rasnitsyn, 1979: 413-414. Holotype 
9, PIN (examined). 

Archaeromma masneri (Yoshimoto), n. comb. 
Protooctonus masneri Yoshimoto, 1975: 511-512. 
Holotype 9, CNC (examined). 

Archaeromma minutissimum (Brues). Ooctonus 
minutissimus Brues, 1937: 44; combination by 
Yoshimoto (1975: 503). Holotype 9, ROMT 
(examined). 

Archaeromma nearcticum Yoshimoto. Archae- 
romma nearctica Yoshimoto, 1975: 506. Holotype 
9, CNC (examined). 



Archaeromma senonicum (Kozlov and Rasnit- 
syn), n. comb. Palaeomymar senonicus Kozlov 
and Rasnitsyn, 1979: 412-413. Holotype 9, PIN 
(examined). 

Remarks. — Our generic placement of A. 
japonicum is tentative because this species 
is based on a unique male and we 
distinguish Archaeromma from other genera 
only by the claval structure of females. We 
classify it in Archaeromma because of its age 
(about 85 mya) and because the illustra- 
tions provided with the original descrip- 
tion are very similar to specimens from 
New Jersey amber that we assign to 
Archaeromma. 

When Yoshimoto (1975) established Pro- 
tooctonus he stated that the forewing and 
female antenna "shows great similarity" to 
that of mymarommatids. He classified the 
genus in Mymarinae (Mymaridae) because 
he considered it had a subpetiolate gaster. 
The female holotype is crushed and the 
detached head is behind the mesosoma 
(Fig. 206). The dorsal surface of the meso- 
soma is clearly visible and structure and 
sculpture of both the mesoscutum 
(Fig. 206: msc) and scutellum (Fig. 206: 
asc + psc) is typical for mymarommatids. 
There is also a rod-like structure projecting 
obliquely from the anterolateral margin of 
the scutellum that almost certainly is an 
axillar phragma (Fig. 206: axp) and the 
toruli are subcontiguous and inserted high 
on the head (Fig. 206). In dorsal view, the 
posterior surface of the head (Fig. 196: 
ocp?) appears detached from the vertex 
and in lateral view the region has concen- 
tric lines dorsally (Fig. 206) similar to 
a hyperoccipital band of pleated mem- 
brane. What might be a 2-segmented 
petiole (Fig. 206: pt|, pt 2 ) lies below the 
detached head anterior to a coxa. For these 
reasons, we transfer Protooctonus from 
Mymaridae to Mymarommatidae. We also 
synonymise Protooctonus under Archae- 
romma because of the similar antennal 
structure of the type species (cf. Figs 192, 
196). The two males designated as para- 
types of P. masneri by Yoshimoto (1975) are 



Volume 16, Number 1, 2007 



97 



not conspecific with the holotype female. 
The males represent an unidentified spe- 
cies of Mymaridae based on the presence 
of long, raised, multiporous plate sensilla 
on the flagellum (diagnostic of Chalcidoi- 
dea) and widely separated antennal toruli 
(diagnostic of Mymaridae). 

Palaeomymar 1 Meunier 

Palaeomymar Meunier, 1901: 289. Type species: 
Palaeomymar succini Meunier; original desig- 
nation and monotypy. 

Diagnosis. — Distinguished from other 
Mymarommatidae by the following com- 
bination of features. Head with ocelli. 
Mandible bidentate. Labiomaxillary plate 
with labium about as wide as maxilla. 
Female antenna 9-segmented with 6-seg- 
mented funicle and 1 -segmented clava 
(Fig. 179); male antenna 12-segmented 
with 6-segmented funicle and 4-segmented 
clava {cf. Fig. 177). Propleura fused into 
carapace. Protibial calcar short and straight 
(Fig. 210). Forewing disc with hair-like 
setae; posterior margin with short marginal 
setae over basal half (Figs 207-209). Pro- 
podeum posteriorly without complete, (T- 
like flange, the foramen protruding as 
short tube (cf. Fig. 94). Male genitalia 
lacking externally projecting parameres. 

Included species. — Palaeomymar succini 
Meunier, 1901: 289-290. 

Remarks. — As discussed above, Duisburg 
(1868) originally had 13 males and 3 
females of which Meunier (1901) examined 
five males when he established P. succini. 
He obtained the type material from the 
"Musee Provincial de Koenigsburg" (Al- 
bertus Universitat in Konigsberg). Accord- 
ing to Grimaldi and Engel (2005) most of 
the material stored in the museum was lost 
or destroyed during World War II, al- 
though a surviving portion is preserved in 
GZG. Search of the GZG Baltic amber 
collection by its curator, M. Reich, failed 
to locate any material identified as P. 
succini, and our survey of their unidenti- 
fied Baltic amber Hymenoptera failed to 



recover any mymarommatids. The slide- 
mounted male labelled "33059 Kl." from 
the GZG "Klebs" collection (see discussion 
of Tertiary fossils) could not have been one 
of the five males examined by Meunier 
(1901). He stated that "embrowning" of the 
forewings prevented observation of the 
structure of the "nervus ulnerus" and both 
wings of the Klebs collection male are 
perfectly clear within the amber piece. 

Because we distinguish three genera for 
the extant species, all of which are classi- 
fied in Palaeomymar, it is necessary to 
designate a neotype for P. succini to fix 
use of the generic name. Based on corre- 
lated structure of the protibial calcar, fore- 
wing marginal setal pattern, and 9-seg- 
mented antenna in females, we identify in 
the ZMUC collection what we believe are 
associated sexes of the species that was 
discussed and illustrated by Duisburg 
(1868) and Meunier (1901). We designate 
a male as neotype because the original type 
series of Meunier (1901) consisted only of 
males and the neotype male is the only 
specimen that under a dissecting micro- 
scope clearly shows a short, straight 
(needle-like) protibial spine (Fig. 210: ca?). 
This spine almost certainly is the calcar, 
but absence of a long, curved calcar is 
indicated even if it is only a pseudospur [cf. 
Fig. 143: ps). The marginal setae of the 
forewings are also visible in the male 
selected and clearly show that the basal 
setae are all very short (Figs 207, 208). It is 
easier to differentiate P. succini from other 
species in Baltic amber based on females 
than males because of additional antennal 
features. We therefore also provide de- 
scriptive features for the ZMUC female 
labelled as 16-1/1961, which is the best 
preserved of all the specimens. It is almost 
the same size as the neotype male and also 
lacks a long protibial calcar. 

Neotype designation: Neotype data: Fos- 
sil male, deposited in ZMUC with the 
following label data: two white labels with 
"Vesterhav, S. Nielsen, Skjoldelev, Min. 
Mus. [Mineralogisk Museum] 1890-108" 



98 



Journal of Hymenoptera Research 



and "Mymarommidae Mymaromma: O.B. 
1964", plus our red neotype label with 
"Neotype Palaeomymar succini Meunier, 
1901 designated by Gibson, Read and 
Huber". 

The maximum size of the amber block 
containing the neotype is 5.5 X 3.75 X 
3.25 mm, though it narrows slightly in all 
dimensions in the direction toward the 
dorsal surface of the specimen. The sides of 
the block are smooth except one side has 
a missing section, which partly obscures 
observation of the specimen in ventrolat- 
eral view. The specimen is obliquely 
angled with the head directed more or less 
dorsally (Fig. 207) when the block is set on 
its side with the missing section. 

Neotype description. — Male. Body length 
about 0.35 mm excluding antennae (length 
probably underestimated slightly because 
of foreshortening). Forewing 0.36 mm long 
X 0.12 mm wide; marginal fringe with 
longest setae as long as maximum wing 
width, the exact number of marginal setae 
uncertain but with about 15 shorter setae 
along posterior margin basally (these setae 
increasing in length distally) and about 35- 
40 much longer setae over apical half; disc 
membrane longitudinally convoluted. 
Head with ocelli. 

Female description. — Antenna (Fig. 179) 
with combined length of pedicel and 
flagellum subequal to body excluding 
second petiolar segment and gaster; scape 
4x as long as wide and 1.6x as long as 
pedicel; relative length of pedicel: funicle: 
clava = 1: 4: 3; funicle with basal four 
segments tubular, apical two segments 
increasingly more distinctly widened ven- 
trally, and apical segment with dorsal 
attachment to base of clava; clava 3.75x as 
long as wide, the dorsal margin apparently 
without strong bent sensilla. Face very 
finely sculptured, shiny; malar space dis- 
tinct. Eye moderately large, with numerous 
ommatidia. Mandibles comparatively slen- 
der, closed with apices apparently widely 
separated medially. Mesosoma with com- 
plete but fine metapleural suture; propo- 



deum posteriorly without complete Pi -like 
flange, the foramen protruding as short 
tube that is longer ventrally than dorsally; 
in lateral view, first petiolar segment with 
basal constriction clearly visible and in- 
serting into tubular propodeal foramen, the 
segment subequal in length to metacoxa 
and twice as long as second petiolar 
segment. Forewing elongate-spatulate, 
1.4x as long (measured from apex of 
venation) as maximum width, with about 
15 shorter setae basally along anterior 
margin, about 45 much longer setae distal- 
ly along anterior, apical and posterior 
margins, and an uncertain number of 
shorter setae along posterobasal margin. 

Mymaromma Girault, revised status 

Mymaromma Girault, 1920: 38. Type species: 
Mymaromma goethei Girault; original designa- 
tion. Synonymy under Palaeomymar Meunier 
by Doutt (1973: 225). 

Petiolaria Blood and Kryger, 1922: 229. Type 
species: Petiolaria anomala Blood and Kryger; 
by monotypy. Synonymy under Mymaromma 
by Girault (1930: 4). 

Description. — Head with or without ocel- 
li; interorbital region of face with 10 setae, 
including 2 setae paramedially above oral 
margin (Figs 18, 20); occipital plate bare 
except for ventrolateral seta at each corner 
(Figs 14, 23, 42). Mandible usually triden- 
tate with variably large and distinct sub- 
apical angle or tooth projecting above oral 
margin (Figs 26-30), though rarely biden- 
tate (Fig. 31). Labiomaxillary plate with 
labium about twice as wide as maxilla 
(Figs 19-22, 28) except very rarely (Fig. 31, 
insert); maxilla with palpal spine apically 
on stipes and with galea internalized and 
apically pustulate (Figs 19-22). Female 
flagellum with 6 or 7 funicular segments 
and 1 -segmented clava; clava with 2 s4- 
type sensilla on outer surface near dorsal 
margin (Figs 58, 59). Male flagellum 12- or 
13-segmented with single s4-type sensil- 
lum apically on fl 9 and fl 10 (Figs 60, 61). 
Propleura fused into carapace (Fig. 18). 
Mesopleuron and metapleuron usually 



Volume 16, Number 1, 2007 99 

fused (Figs 80, 81, 83), rarely partly and with 1-4 setae. Male genitalia without 

(Fig. 87: mms) or completely separated by externally evident parameres. 
suture (Fig. 82), but with curved furrow 

extending from posteroventral margin of J™ 1 ^ species -Mymaromma anomalum 

., . j- .■ 1 1 - L (Blood and Kryger), rev. comb. Petiolaria auom- 

mesothorax to distinct metapleura pit near ^ Blood ^ £ ^ ^^ Ro!o 

propodeal spiracle (Figs 81, 87). Metatho- ^ BMNH (not examined) 

racic-propodeal complex with meta- Mymaromma buyckxi Mathot, 1966: 236-237, 

pleuron, at least dorsally, distinctly rev . comb. Holotype 9, ISNB (examined). 

smoother than more coarsely sculptured Mymaromma goethei Girault, 1920: 38-39, rev. 

propodeum (Figs 80-83); spiracular aper- comb. Holotype 9, QMBA (examined). 

ture oval to slit-like and connected to Mymaromma mirissimum (Girault), n. comb. 

anterior margin of propodeum by slit-like Mymaromella mirissima Girault, 1935: 3. Holo- 

peritreme, in lateral view the peritreme tv P e 9, QMBA (examined). 

and dorsal margin of metapleuron forming Mymaromma ypt (Triapitsyn and Berezovs- 

acute angle directed between scutellum ki y)' n< comb - ^oAaeomymar ypt Triapitsyn and 

and propodeum (Figs 82-84); metanotum Berezovski Y' 2006: 5-7. Holotype 9, Zoological 

separated from propodeum by peritreme !^ USeum ° f M °^° w S^ University, Moscow, 

, . c , .., . , ,J. nAS Russia (not examined). Paratype 9 and C, CNC 

but fused with metapleuron (Fie. 84); pre- , ,, ;r 

f , ? (examined), 
spiracular setae within angle formed by 

dorsal margin of metanotum and peri- Distribution. — We have seen individuals 
treme (Figs 84, 85) and postalar sensillum and /or species have been reported in the 
at extreme dorsal margin of metapleuron literature from the following areas: Afro- 
(Figs 84, 85). Propodeum posteriorly re- tropical — Congo (Mathot 1966), Gabon, 
flexed into Pi-like flange over petiolar Madagascar, Nigeria. Australasian — Aus- 
insertion, the posteroventral margin of tralia (Girault 1920), Chatham Island, 
flange and projection of metapleuron form- Christmas Island, Hawaii (Beardsley et al. 
ing pincer-like structure (Figs 81, 83, 2000), Indonesia (Ceram), New Caledonia, 
86, 87). Forewing often appearing more or New Zealand, Norfolk Island, Papua New 
less lanceolate, the disc longitudinally Guinea. Nearctic — Canada, Mexico. Neo- 
convoluted with hair-like setae aligned tropical — Bermuda, Brazil, Colombia, 
along folds on dorsal and ventral surfaces Oriental — China (Lin 1994) [Fujian, 
(Figs 113-115), and sometimes densely Yunan], Indonesia [Kalimantan, Sulawesi, 
hairy if setae very long (Fig. 116); posterior Sumatra], Malaysia [Sabah], Nepal, Philip- 
margin usually with basal seta conspicu- pines (Gallego 1986), Taiwan, Thailand, 
ously longer than adjacent short setae Palaearctic — Belgium (Debauche 1948), 
(Figs 113-115). Femora without bumps on Bulgaria (Donev 1982), Czech Republic 
posterior surface. Protibial calcar short, (Kalina 1989), Denmark (Blood and Kryger 
straight and simple (Figs 142, 144a, 145a). 1936), England (Blood and Kryger 1922), 
Metasoma with Mtj longer than Mt 2 but France, Germany (Vidal 2001), Hungary, 
their relative length variable; Mti dorsally Italy (Viggiani 1966), Japan, Norway (Han- 
variable in sculpture, almost smooth sen 1997), Poland (Soyka 1937), Romania 
(Fig. 83) to strongly reticulate (Fig. 88), (Andriescu and Suciu 1963), Russia (Tria- 
and only very rarely with lateral seta; Mt 2 pitsyn and Berezovskiy 2000), South Korea 
finely sculptured, usually almost smooth (Triapitsyn and Berezovskiy 2006), Spain 
(Fig. 83); Mt 7 with spiracle and seta mesal (Askew et al. 2001), Sweden, Switzerland 
to spiracle (Fig. 152); syntergum with cerci, (Ferriere 1948). 

the cercus differentiated as subcircular Ronnrks. — All described species are ex- 
structure (cf. Fig. 150) or variably fused tant, but species likely existed in the 
with tergite as part of its surface (Fig. 152), Tertiary. 



100 Journal of Hymenoptera Research 

Mymaromella Girault, revised status podeum, but separated from metapleuron 
Mymaromella Girault, 1931: 4. Type species: b Y peritreme such that metanotal-propo- 
Mymaromella mira Girault; by monotypy. deal complex form variably long angle 
Synonymy under Mymaromma Girault by directed anteriorly between scutellum and 
Annecke and Doutt (1961: 14) and under dorsal margin of metanotum (Figs 98, 100, 
Palaeomymar Meunier by Doutt (1973: 225). 102); prespiracular setae within dorsolater- 
al angle of metanotal-propodeal complex 
Description.— -Head with or without ocel- anc j postalar sensillum either at extreme 
li; interorbital region of face with 6-10 anterior angle of complex (Figs 98, 100) or 
setae, including 2 (Fig. 36) or 4 (cf. Fig. 56) sometimes appearing anterior to it 
setae above oral margin; occipital plate (Fig. 102). Propodeum posteromedially 
with 2 paramedial setae (Figs 45, 46: ms) in constricted into short tube anterior to 
addition to ventrolateral seta at each reflexed foraminal margin (Figs 90-97, 
corner. Mandible bidentate with both teeth gg r 101), and with posterolateral margin 
acutely angled (Figs 34, 48, 49), the dorsal more or less distinctly reflexed as flange on 
tooth much longer than ventral subapical either side of constriction (Figs 92, 94: pf) 
tooth. Labiomaxillary plate with labium to form pincer-like structure with pro- 
narrow, only about as wide as maxilla jection of metapleuron (Figs 92, 94, 96, 97, 
(Figs 41, 41, 44, 47-49); maxilla with palpal 101). Forewing variable in shape (Figs 117- 
spine subapical and with galea a longer, 121) though often spatulate (Fig. 119) or 
fleshy lobe exterior to spine (Fig. 41). broadly rounded apically (Figs 118, 166), 
Female flagellum with 7 funicular seg- the disc with or without distinct longitu- 
ments and 1-segmented clava; clava with dinal convolutions but with short, spine- 
2 or 3 s4-type sensilla on outer surface near like setae; length of marginal setae highly 
midline (Figs 62, 63, 65, 67, 69). Male variable and posterior margin with 
flagellum 13-segmented with single s4- (Figs 118, 1119) or without (Figs 117, 120, 
type sensillum apically on fly and fl 10 or 166, 121) a conspicuously long basal seta. 
flg-flio (corresponding to number of sen- Femora without bumps on posterior sur- 
silla on female clava) (Figs 66, 68). Pro- face. Protibial calcar long, apically curved 
pleura abutting medially or at least distin- and bifurcate (Figs 137-1 39a). Metasoma 
guished by median line of differentiated with Mt] longer than Mt 2 but their relative 
sculpture (Figs 43, 44). Mesopleuron and length variable; Mt] dorsally more or less 
metapleuron separated by suture (Figs 95- transversely strigose-coriaceous and with 
97, 99, 101), without a distinct metapleural lateral seta within basal half (Figs 90, 92, 
pit though sometimes with a very tiny hole 94); Mt 2 variable in sculpture dorsally, 
about midway between ventral margin of usually transversely strigose; Mt 7 with 
pleuron and propodeal spiracle (Figs 97, spiracle and seta mesal to spiracle 
101), and without curved furrow extending (Figs 149-151); syntergum with cerci, 
dorsally from posteroventral margin of the cercus differentiated as subcircular 
mesothorax. Metathoracic-propodeal com- structure (Figs 149-151) or variably 
plex with sculpture of propodeum and fused with tergite as part of its surface 
metapleuron quite similar even if pleural (cf- Fig. 152), and with 2-4 setae. Male 
sculpture somewhat finer (Figs 95-97, 99, genitalia without externally evident para- 
101); spiracular aperture circular to oval meres. 

and connected to anterior margin of pro- r , A A ,, „ , ■ ,, . , 

, i i- i-i Included species. — Mymaromella chaoi (Lin), n. 

podeum by slit-like peritreme, in lateral comb Falaeomymar chaoi Lin , 1994: 123 _ 124 . 

view the peritreme and dorsal margin of Holotype 9, FAUF (not examined). 

metapleuron forming evenly convex arc Mymaromella cyclopterus (Fidalgo and De 

Figs 95-102); metanotum fused with pro- Santis), n. comb. Palaeomymar cyclopterus 



Volume 16, Number 1, 2007 



101 



Fidalgo and De Santis, 1982: 3-4. Holotype 9, 
MLPA (examined). 

Mymaromella duerrenfeldV (Schliiter and Kohr- 
ing), n. comb. Palaeomymar duerrenfeldi Schliiter 
and Kohring, 1990: 117-118. Holotype $, ZMB 
(examined). 

Mymaromella mira Girault, 1931: 4. rev. comb. 
Holotype 9 photograph, QMBA, USNM (exam- 
ined). 

Distribution. — We have seen individuals 
and /or species have been reported in the 
literature from the following areas: Afro- 
tropical — Ivory Coast. Australasian — 
Australia (Girault 1931). Nearctic — Ca- 
nada (Clouatre et al. 1989), United States. 
Neotropical — Argentina (Fidalgo and De 
Santis 1982), Brazil, Trinidad, Venezuela 
(reported as Palaeomymar by Garcia 2000). 
Oriental — China (Fujian, Guangxi). Pa- 
laearctic — South Korea (Triapitsyn and 
Berezovskiy 2006), Sweden. 

Remarks. — We newly classify M. duerren- 
feldi in Mymaromella because of the rela- 
tively recent age (about 5 mya) of this 
Tertiary species and because it has biden- 
tate mandibles, a long and curved protibial 
calcar, and broadly rounded forewings 
with short discal setae. 

Lin (1994, fig. 4) illustrated the bidentate 
mandibles of M. chaoi as having a short 
dorsal tooth and a long ventral tooth. 
Study of topotypic material shows this to 
be erroneous and the short tooth is ventral 
as in other Mymaromella. 

Mymaromella mira was described from 
a photograph of a female "collected at 
Canterbury, Victoria in April" (Dahms 
1984), which is in the QMBA. According 
to Girault's unpublished manuscript, the 
photograph is the "type" and a "paratype" 
photograph dated "January 4, 1931 (4-1- 
31)" that was sent to the USNM (Dahms 
1984). The USNM photograph (Fig. 168) is 
identical to the QMBA photograph and has 
the data "[?] Girault letter to Gahan Nov. 
1934". Because the two photographs are 
identical this suggests that M. mira was 
based on a unique specimen. The discrep- 
ancy between the collection date of the 



"type" and the date given for the "para- 
type" may simply reflect the date the 
photographs were printed or perhaps is 
a misinterpretation of the month vs. the 
day, i.e., "4-1-31" either representing April 
1 or 4 January, 1931. Close examination of 
the photographs show that there is a com- 
paratively long posterobasal seta on the 
forewings (cf. Fig. 118). 

Zealaromma n. gen. 

Type species. — Mymarommn insulare Val- 
entine, 1971: 331-333; present designation. 

Etymology. — A combination of New 
Zealand, the only country from which 
species of the genus are known, and 
Mymaromma, the genus in which the type 
species was first described. 

Description. — Head with or without ocel- 
li; interorbital region of face with 8 setae, 
including 4 setae above oral margin 
(Fig. 56); occipital plate bare except for 
ventrolateral seta at each corner (Fig. 50). 
Mandible bidentate with both teeth acutely 
angled and dorsal tooth much longer than 
ventral subapical tooth (Figs 52, 54). La- 
biomaxillary plate with labium narrow, 
only about as wide as maxilla (Figs 50, 
52); maxilla with palpal spine apical on 
stipes and with galea internalized and 
apically pustulate (Figs 55-57). Female 
flagellum with 7 funicular segments and 
2-segmented clava (Figs 71, 172); terminal 
claval segment with 2 s4-type sensilla on 
outer surface near dorsal margin (Fig. 71). 
Male flagellum 12-segmented with single 
s4-type sensillum near apex of fl 9 and fl 10 
(Fig. 73). Propleura abutting medially. Me- 
sopleuron and metapleuron fused except 
for short suture below base of forewing, 
without evident metapleural pit, but with 
shallow groove extending dorsal ly from 
posteroventral margin of mesothorax to 
propodeal spiracle (Figs 103, 108). Meta- 
thoracic-propodeal complex with distinct 
smooth region between much more coarse- 
ly sculptured propodeum and meta- 
pleuron ventrally (Figs 104, 107, 108); 



102 Journal of Hymenoptera Research 

spiracular peritreme circular and without (Figs 153-155) and usually bare (Figs 153, 

slit-like peritreme between aperture and 155); syntergum without cerci (Figs 153, 

anterior margin of propodeum; metano- 155). Male genitalia with elongate-digiti- 

tum either independent from both propo- form paramere bearing long terminal seta 

deum and metapleuron (Figs 105, 106) or projecting externally from between synter- 

fused with propodeum and metapleuron gum and hypopygium (Figs 154-156). 

(Figs 109, 110), but with postalar sensillum T , , , . ^ , . 

,„. inP 11n . j , Included species. — Zealaromma insulare (Valen- 

(Fies 105, HO: pas) separated from appar- l . .,__,. ' , 

v ° ,, . . i tine, 1971), n. comb. 

ent metanotum anterior to prespiracular _ , , ,. . _.. „ , , TJ 

,t,. + n <- -,« n \ ^. i Zealaromma valentinei Gibson, Read and Hu- 

setae (Figs 10s, 110: pss). Propodeum , 

posteromedially constricted into short tube 

anterior to reflexed foraminal margin Remarks. — Valentine (1971) stated that he 
(Figs 103, 104, 107, 108), and with postero- recognized five species from mainland 
lateral margin reflexed as distinct vertical New Zealand in addition to Z. insulare, 
flange beside tubular constriction which differed in the form and fusion of 
(Figs 104, 107). Forewing variably distinct- the mesothoracic segments, simple or di- 
ly spatulate (Fig. 165) to lanceolate vided clava, size and shape of the fore- 
(Figs 122, 164), but disc longitudinally wings (one being apterous), and presence 
convoluted and with hair-like setae aligned or absence of ocelli. Noyes and Valentine 
along folds on dorsal and ventral surfaces; (1989) also stated that there were five 
marginal setae comparatively short and undescribed species in New Zealand, in- 
stiff (spine-like) (Figs 122, 164) or very long eluding a brachypterous and apparently an 
and hair-like (Fig. 165), the posterobasal apterous species. We did not locate any 
seta sometimes long but if so then not brachypterous or apterous specimens and 
distinctly differentiated in length from from the material we examined from New 
adjacent setae (Fig. 165). Femora, particu- Zealand recognize only three species: Z. 
larly mesofemur, with bumps on posterior insulare, Z. valentinei and an undescribed 
surface (Fig. 146). Protibial calcar long, species of Mymaromma. This species, M. sp. 
apically curved and deeply bifurcate with 10, is represented by a single female from 
very long inner tine (Fig. 140). Metasoma the South Island (NZAC) and several other 
with Mt! nearly twice as long as Mt 2 ; M^ specimens (Appendix II) that differ from 
with subbasal lateral seta (Fig. 104) and all other mymarommatids in their extreme- 
Mt] and Mt 2 dorsally very finely sculp- ly long and dense forewing discal setae 
tured (Figs 104, 107); Mt 7 without spiracle (Fig. 116). 



KEY TO SPECIES OF ZEALAROMMA 

Forewing with about 10 stiff, relatively short (needle-like) and widely spaced distal 
marginal setae originating from wing margin (Figs 122, 164); head without ocelli; 
metanotum fused with both metapleuron and propodeum (Figs 108- 
110) Zealaromma insulare (Valentine) 

Forewing with numerous fine, very long (hair-like) and closely spaced distal marginal 
setae originating distinctly from within wing periphery (Fig. 165); head with ocelli 
(Fig. 53); metanotum separated from metapleuron and propodeum by sutures 
(Figs 105, 106) Zealaromma valentinei n. sp. 



Volume 16, Number 1, 2007 



103 



Zealaromma valentinei n. sp. 

Type material. — Holotype. Female. "NEW 
ZEALAND BR [South Island: Buller] Lake 
Rotoiti 610m; 26 Dec 1979 - 4 Jan 1980 A.K. 
Walker" / "Holotype Zealaromma valentinei 
Gibson, Read and Huber" (NZAC). Paratypes 
(159, 5c?). NEW ZEALAND, North Island: ND 
[Northland] Waipoua Kauri Forest, 11- 
12.XII.1983, L. Masner, s.s. (lo CNC). TK 
[Taranaki]: Mt. Egmont N.P., 16.XII.1983, L. 
Masner, s.s. (lo CNC). NEW ZEALAND, South 
Island: BR lo, same data as holotype (NZAC). 
Lower Buller R., Norris Ck., 14.X.1970, J.I. 
Townsend, litter 70/158 (29 on one card 
NZAC). Punakiki, 28.XII.1983, L. Masner, s.s. 
(19, let CNC). Tuttys Plateau, Mawhera, 
20.IX.1972, J.S. Dugdale, moss 72/178 (19 
NZAC). MC [Mid Canterbury], Banks Peninsu- 
la, Prices Valley, 11.1981, R.P. Macfarlane, 
malaise trap, edge of native bush (I9 NZAC). 
Banks Peninsula, Prices V, 19-27.X.1988, 
J.W. Early (59, 26" CNC; I9 gold coated for 
SEM). NN [Nelson], Wooded peak Dun Track 
Sdle, 14.IX.1971, G.W. Ramsay, litter 71/110 
(19 NZAC). SL [Southland], Longwood For- 
est, Purakino picnic area, 24-28.1.1999, L. Le- 
Sage, Nothofagus forest, YPT 99-57 (I9, 2J 
CNC). 

Etymology. — Named in honour of Dr. 
Errol Valentine, who first published on 
New Zealand mymarommatids. 

Description. — Length, 0.55 mm (air-dried 
holotype). Body yellow except small tri- 
angular region below base of forewing and 
ocellar triangle dark brown, gaster some- 
times brownish dorsally; hind leg with 
apices of basal three tarsal segments very 
narrowly brown; forewing disc with mid- 
dle third to basal half of mesh-like linea- 
tions and the marginal setae behind this 
region brown (Fig. 165). 

Head with ocelli (Fig. 53). Face with 
supraclypeal region finely strigose, other- 
wise with distinctive mesh-like sculpture, 
the sculpture narrowed toward midline so 
as to more or less form a median carina 
(Fig. 52). Eye moderately large, with about 
20-25 ommatidia in female and 30-35 
ommatidia in male; separated from oral 
margin by distinct malar space of similar 



width as mandible (Figs 52, 54). Antenna 
of female as in Fig. 172. 

Mesosoma with propleura completely 
divided medially (Fig. 112). Forewing as 
in Fig. 165, with 43-48 very long and thin 
(hair-like), evenly spaced marginal setae 
over about apical half of wing, 12-15 much 
shorter setae basally along anterior margin, 
and with 6 or 7 variably shorter setae 
basally along posterior margin, of which at 
least the basal 3 setae are comparatively 
long and at least the distal seta (sometimes 
distal 1-3 setae) are quite short. Metano- 
tum separated from metapleuron laterally 
and from propodeum posterolaterally 
(Figs 105, 106). 

Metasoma of male without seta on Mt 7 
(Fig. 155). 

Remarks. — Body length varies from about 
0.5 mm in air-dried individuals to about 
0.7 mm in some critical-point dried fe- 
males. The difference appears to be mostly 
because the gaster remains inflated in 
critical-point dried specimens. Females 
with the gaster neither shrunken nor 
obviously over-inflated are about 
0.65 mm. The species description of Z. 
valentinei is brief because it mostly includes 
only those features that readilv differenti- 
ate it from Z. insulare. 

SUPRAFAMILIAL RELATIONSHIPS 

One of the objectives of this study was to 
determine whether any additional mor- 
phological evidence could be found to help 
resolve the suprafamilial relationships of 
Mymarommatoidea. Mymarommatoidea 
has been proposed as the sister group of 
either Serphitoidea (Kozlov and Rasnitsyn 
1979) or Chalcidoidea (Gibson 1986). We 
therefore included fossils of Serphitidae in 
our study, but our observations are based 
on only seven Taimyr and nine Canadian 
Cretaceous specimens. Our study of my- 
marommatoids revealed that they are 
much more diverse morphologically than 
thought previously. Consequently, the in- 
ferences made below should all be consid- 
ered as tentative until character-state dis- 



104 



Journal of Hymenoptera Research 



tribution and groundplan features of Ser- 
phitidae are determined more accurately 
through an equally comprehensive study. 

A Mymarommatoidea + Serphitoidea 
sister-group relationship is supported by 
common possession of a similarly struc- 
tured 2-segmented petiole (Kozlov and 
Rasnitsyn 1979), whereas common posses- 
sion of axillar phragmata as the sites of 
origin for tubular mesotergal-mesotro- 
chanteral muscles is the principal evidence 
for a Mymarommatoidea + Chalcidoidea 
relationship (Gibson 1986). Vilhelmsen and 
Krogmann (2006, figs 12-14) showed that 
the prophragma of M. anomalum has rod- 
like structures similar to axillar phragmata. 
They suggested that this apparent "serial" 
homology and the different structure of the 
axillar phragmata in mymarommatids and 
chalcids (tubular vs. flat) might indicate 
that the phragmata evolved independently 
in the two taxa. However, Mymaromma- 
toidea and Chalcidoidea are still the only 
known apocritans having the mesotro- 
chanteral depressor muscle consisting only 
of a notal portion that originates at least 
partly from an axillar phragma. Our study 
shows that Cretaceous mymarommatids 
also had tubular mesotrochanteral depres- 
sor muscles originating from rod-like ax- 
illar phragmata (Fig. 206: axp). Knowledge 
of the structure of this muscle in Serphiti- 
dae would therefore be valuable for in- 
ferring relationships. Serphitids are much 
more highly sclerotized and melanized 
than are mymarommatids and we did not 
see any specimen in which the internal 
musculature was visible. 

As noted by Vilhelmsen and Krogmann 
(2006), the presence or absence of an 
exposed prepectus and the position of the 
mesothoracic spiracle are other phyloge- 
netically informative features for inferring 
relationships of Chalcidoidea. The position 
of the mesothoracic spiracle in Chalcidoi- 
dea is autapomorphic, being at or above 
the dorsal margin of the prepectus between 
the pronotum and mesoscutum (Gibson 
1986, figs 18-29). Other parasitic Hyme- 



noptera have the spiracle ventral to the 
level of the dorsal margin of the pronotum, 
either between the pronotum and meso- 
pleuron or secondarily on the pronotum in 
the same relative position. Furthermore, 
the prepectus or its remnant extends 
dorsally behind the spiracle, between it 
and the mesopleuron (Gibson 1985, 
figs 19-26), except in Stephanidae (Gibson 
1985, fig. 16) and Ichneumonoidea (Gibson 
1985, figs 27, 28). Although a pronotal 
notch was not observed in any fossil 
mymarommatoid or in Zealaromma (Figs 
103, 108), the incision on the posterior 
margin of the pronotum of some mymar- 
ommatids (Figs 16, 82: pn; 99, 101) could 
indicate the position of the mesothoracic 
spiracle before it was lost. If so, the 
ancestor of mymarommatoids had the 
mesothoracic spiracle in the symplesio- 
morphic position for parasitic Hymenop- 
tera. It therefore is not useful for establish- 
ing relationships of Mymarommatoidea, 
but supports the hypothesis of autapomor- 
phy for Chalcidoidea. 

In lateral view, one specimen of Serphi- 
tidae from Taimyr amber (PIN: 3730/31) 
very clearly has a white, globular meso- 
thoracic spiracle ventral to the dorsal 
margin of the pronotum between the 
pronotum and mesopleuron (Figs 220, 
221: sp). The spiracle appears to lie partly 
within the excised posterodorsal margin of 
the pronotum (most clearly visible from 
a somewhat ventral view, Fig. 222: sp). In 
lateral (Figs 220, 221) or ventrolateral 
(Fig. 222) view, about the dorsal third of 
the posterior margin of the pronotum abuts 
the anterior margin of the mesopleuron, 
but ventrally a slender, spindle-shaped 
region (Figs 220-222: pre?) separates the 
posterior margin of the pronotum from the 
mesopleuron between the spiracle and 
procoxa. This intervening region appears 
to be a separate sclerite because its anterior 
margin is delimited by a distinct suture 
and it is on a lower level than the posterior 
margin of the pronotum, because it is 
separated from the mesopectus by a distinct 



Volume 16, Number 1, 2007 



105 



suture at least ventrally, and because its 
sculpture is different from the putative 
pronotum and mesopectus (with longitu- 
dinal rugae compared to rugose pronotum 
and mesopectus). The region could be 
a differentiated part of the mesopectus, 
but more likely it is a relatively long and 
slender prepectus. In lateral view, the 
mesothoracic spiracle extends over the 
anterior margin of the mesopleuron 
(Figs 220-222) and appears to be on 
a slightly higher level than the meso- 
pleuron. This is because the mesopleuron 
posterior to the spiracle is slightly concave. 
The concave region is distinct because it is 
delineated by carinae (Fig. 221: c). Many 
extant platygastroid (Gibson 1985, figs 23, 
24) and some proctotrupoid (Gibson 1985, 
figs 19a, 20a, 25a) taxa also have the 
spiracle on a slightly higher level than the 
mesopleuron because the posterior prono- 
tal inflection extends dorsally under and 
behind the spiracle. This suggests that if 
the slender region of PIN: 3730/31 is the 
prepectus then it likely also extends dor- 
sally behind the spiracle. If so, the pre- 
pectal structure is more similar to that of 
Monomachidae (Gibson 1985, fig. 15) than 
Stephanidae (Gibson 1985, fig. 16). Because 
of the intervening region, the posterior 
margin of the pronotum is somewhat 
sinuate (Figs 220-222). The posterior mar- 
gin of the pronotum is similarly sinuate in 
many Scelionidae (Platygastroidea) having 
a netrion (Gibson 1985, fig. 23). Rasnitsyn 
(1980) hypothesized that the netrion in 
scelionids is the prepectus fused with the 
pronotum. The similarity between the 
putative sclerite of PIN: 3730/31 and the 
netrion of some scelionids suggests that it 
might be fused with the pronotum internal 
to the external suture (cf. Figs 220-222 with 
Gibson 1985, fig. 23 and Masner 1979, 
figs 4-8). If so, it is structurally the same 
as a netrion and therefore a possible 
synapomorphy for Serphitoidea + Platy- 
gastroidea. 

The other serphitids we examined did 
not have such a distinctly differentiated 



mesopectal region, though the specimen in 
PIN: 3311/80 and the largest of the three 
specimens in PIN: 3730/28-30 has an 
obscurely differentiated region between 
the pronotum and mesopectus, at least on 
the right side from direct ventral view. The 
region is less distinct in these two speci- 
mens partly because it is quite smooth, 
similar in sculpture to the pronotum and 
mesopleuron. The posterior margin of the 
pronotum of the holotype of Microserphites 
parinrius Kozlov and Rasnitsyn was also 
stated as having a narrow border (Kozlov 
and Rasnitsyn 1979, fig. 7), which may be 
equivalent to the region we observed in 
PIN: 3730/31. Further study is necessary to 
determine whether all Serphitidae have 
a slender mesopectal region differentiated 
below the mesothoracic spiracle, whether 
this region is a free prepectus or a netrion, 
and whether the posterior margin of the 
pronotum is free from the mesopleuron or 
connected by a posterior pronotal inflec- 
tion. These observations could be compli- 
cated because the pronotum is somewhat 
moveable relative to the mesothorax in 
extant apocritans with a prepectus (e.g. 
Monomachidae). The pronotum could 
overlie and partly or entirely conceal the 
sclerite in some fossils if the slender region 
in PIN: 3730/31 is a free prepectus rather 
than a fused netrion. The margin of the 
sclerite adjacent to the pronotum appears 
to be on a slightly lower level than the 
pronotum in PIN: 3730/31, which suggests 
that the pronotum can override the sclerite. 
Study of other individuals of the same 
species as in PIN: 3730/31 could help 
determine whether the pronotum is move- 
able, and therefore whether the sclerite is 
more likely a prepectus or a netrion. The 
species is readily identified by the presence 
of a distinct setal patch on the mesopectus 
near the procoxa (Figs 220, 222: sep). 

Rasnitsyn et al. (2004, p. 128) stated that 
"monophyly of Khutelchalcididae within 
Chalcidoidea is very likely" based on 
presence of a spiracular excision in the 
pronotum (as an indicator of the chalcid 



106 Journal of Hymenoptera Research 

synapomorphy of a dorsal spiracular posi- son 1985, figs 21, 23, 27, 29). These features 
tion) and a free prepectus. However, if are not evidence of relationships. Gibson 
their interpretation of the unique fossil (1986) showed that structure of the prono- 
comprising the family is correct, then the turn and mesonotum is correlated with 
comparatively large incision in the postero- presence or absence of a free prepectus and 
dorsal margin of the pronotum (Rasnitsyn relative mobility of the pronotum. Exami- 
et al. 2004, fig. 6) shows that the spiracle is nation of the serphitids we could observe 
in the symplesiomorphic position relative in ventral view shows a propleural struc- 
to chalcids. This placement of the putative ture (Fig. 222: ppm) very similar to extant 
spiracle and the vertical structure ventral mymarommatids with the propleura abut- 
to it, which they interpret as the prepectus ting along their entire length (Fig. 112). 
(Rasnitsyn et al. 2004, fig. 6), resemble the Vilhelmsen and Krogmann (2006) correctly 
structure of the serphitid (Fig. 222) in PIN: stated that most Chalcidoidea have the 
3730/31. As noted above, the presence of prosternum partly exposed because the 
a free prepectus is symplesiomorphic and ventral margins of the propleura are more 
does not support monophyly of Khutel- or less divergent. Mymarommatids are 
chalcididae + Chalcidoidea. The thoracic therefore more similar to serphitids than 
structure of Monomachidae, Serphitidae, to chalcids in their propleural structure. 
Khutelchalcididae and some Aculeata sug- However, this structure is shared with 
gest that one or more early lineages of most Hymenoptera excluding Chalcidoi- 
Apocrita had a free prepectus that was dea and the most basal symphytan lineages 
visible ventral to the spiracle depending on (Vilhelmsen and Krogmann 2006). The 
position of a somewhat mobile pronotum. basal symphytan lineages with chalcid-like 
The ancestor of Platygastroidea may have propleura also have the pronotum corn- 
had a similar structure, but it is only in paratively mobile relative to the mesotho- 
Chalcidoidea that the prepectus was en- rax. Further study is required to determine 
larged secondarily to intervene between whether there is a correlation between 
the pronotum and mesopleuron dorsally. propleural structure and pronotal mobility 
This prepectal structure apparently was similar to that of the pronotal-mesonotal 
derived concurrently with the dorsal shift complex. Such a study should include 
in position of the mesothoracic spiracle. Of Aculeata with an independent prepectus 
critical importance in taxa with a prepectus, and a relatively mobile pronotum as well 
whether this is free or secondarily fused as aculeates with a rigidly attached prono- 
with the pronotum, is whether it extends turn. 

behind and above the level of the meso- Most apocritans have a curved, apically 

thoracic spiracle. Unfortunately, this can bifurcate protibial calcar (Basibuyuk and 

usually be determined only by dissecting Quicke 1995). Only Chalcidoidea and 

the pronotum from the mesothorax and Mymarommatoidea have some members 

therefore is not readily apparent in fossils, with this calcar structure and others with 

However, the exact structure of the pre- a short, straight, needle-like calcar. Because 

pectus/netrion in Serphitidae may be de- Chalcidoidea and Mymarommatoidea are 

termined if a fossil is discovered with the both well substantiated as monophyletic 

pronotum or prothorax detached from the taxa, the reduced calcar shared by some 

mesothorax. members of both groups must be conver- 

Mymarommatoidea, Serphitoidea, Platy- gent. For the same reason, the similar 

gastroidea and most parasitic Hymenop- forewing and hind wing structures of 

tera other than Chalcidoidea have a more mymarommatoids and most Mymar (cf. 

ore less gibbous mesoscutum and the Figs 1-7 with 129-136) are certainly con- 

pronotum triangular in lateral view (Gib- vergent. 



Volume 16, Number 1, 2007 107 

Kozlov and Rasnitsyn (1979) did not 1989, Liu and Nordlander 1992). This, 
provide the tibial spur formula for Micro- along with loss of one mesotibial spur 
serphites or Aposerphites Kozlov and Rasnit- from Megalyridae and both mesotibial 
syn, but stated that this was 1:2:2 for spurs from Stephanidae, suggests that 
Serphites Brues. Most of the serphitids we there could be some correlation between 
examined had two distinct meso- and parasitism of wood-boring insects and the 
metatibial spurs, though these were not loss of mesotibial spurs. The single meso- 
visible in all examined inclusions. Nau- tibial spur shared by Chalcidoidea and 
mann and Masner (1985) listed the tibial Megalyridae presumably is convergent, 
spur formula as 1:2:2 for 9 of the 11 families but further study of the number of meso- 
they treated as the "proctotrupoid com- tibial spurs throughout Apocrita is war- 
plex" of families. The formula was listed as ranted for phylogenetic inference. 
1:2:2 or 1:1:1 for Scelionidae and Platygas- The similar forewing venation of most 
tridae, but Austin and Field (1997) stated Chalcidoidea and Scelionidae has often 
that members of the most primitive tribes been cited as possibly indicating a sister- 
of Platygastroidea are plesiomorphic in group relationship between Chalcidoidea 
having two mesotibial spurs. The number and Platygastroidea. The venation is re- 
of mesotibial spurs is also variable in duced to single vein complex near the 
Ceraphronoidea. Members of Megaspili- anterior margin of the wing in most 
dae have two and members of Ceraphro- members of both groups. Typically, there 
nidae one mesotibial spur (Gauld and is a "submarginal", "marginal", "stigmal" 
Bolton 1988). We found that the tibial spur and "postmarginal" vein, and often also 
formula is 1:0:2 for Stephanidae and 1:1:2 a short "uncus" projecting from the stig- 
for Megalyridae and most Chalcidoidea. A mal vein (Huber and Sharkey 1993, fig. 14; 
few Chalcidoidea only have a single meta- Gibson 1997, fig. 5). Rasnitsyn et al. (2004) 
tibial spur, but the exceptions include some stated that the forewing pterostigma was 
of the tiniest Mymaridae, such as species of not prominent and possibly was absent 
Alaptus Westwood and Camptoptera For- from Khutelchalcididae, unlike in an un- 
ster. At least extant mymarommatids lack described fossil they identified as a scelio- 
both meso- and metatibial spurs, though nid from the lowermost Cretaceous. The 
the tibiae have socketed setae ventroapi- photograph of this putative scelionid 
cally (pseudospurs) that can be mistaken shows a costal vein (Rasnitsyn et al. 2004, 
for tibial spurs (Figs 139b, c; 141b, c; 144b, fig. 7: c) and a longitudinal pterostigma 
c: ps). Absence of meso- and metatibial distal to a "pre-pterostigmal break", plus 
spurs from mymarommatids and the loss an oblique r-rs (= stigmal vein) projecting 
of one mesotibial spur from more derived from the pterostigma near its middle. 
Platygastroidea and in Ceraphronidae Another photograph of the same impres- 
might be correlated with very small body sion sent to us by Alex Rasnitsyn (PIN) 
size. If so, the presence of only a single more clearly shows the venation (Fig. 227). 
mesotibial spur in Chalcidoidea could in- The presence of a pterostigma suggests 
dicate that their common ancestor was that the marginal vein of extant scelionids 
very small. This would support the hy- evolved through a gradual narrowing of 
pothesis that the chalcid ancestor was an the pterostigma into an only somewhat 
egg parasitoid rather than a parasitoid of thickened vein-like structure (A. Rasnitsyn, 
a wood-boring insect (Gibson et al. 1999). pers. comm.). However, the groundplan 
However, even though most Cynipoidea venation of extant Platygastroidea is con- 
have two mesotibial spurs, some species of troversial. Masner et al. (in press) reinter- 
Ibalia Latreille (Ibaliidae) have only a single preted the groundplan forewing venation 
mesotibial spur (Ronquist and Nordlander of Scelionidae based on the putative basal 



108 



Journal of Hymenoptera Research 



extant lineages of the group. They noted 
that members of these lineages lack a mar- 
ginal vein. Rather, the submarginal vein 
(R) bifurcates distally before attaining the 
anterior margin of the wing so that the 
stigmal vein (r-rs) originates within the 
membrane and is separated from the 
wing margin by a short R! (Masner et al. 
in press, figs 41-43, 45-47). Based on this, 
they concluded that a marginal vein 
evolved secondarily in Scelionidae and is 
convergent with the marginal vein of 
Chalcidoidea. They did not state explicitly 
how the marginal vein of some 
scelionids evolved, but they illustrated 
a distal extension of R] along the wing 
margin in some scelionids as producing the 
postmarginal vein (Masner et al. in press, 
figs 46, 47). Presumably, secondary proxi- 
mal lengthening of R along the wing 
margin resulted in the marginal vein of 
some scelionids (Masner et al. in press, 
fig. 44). 

Masner et al. (in press, fig. 49) compared 
their interpretation of the groundplan 
venation of Platygastroidea with Cynipoi- 
dea to illustrate presumed symplesio- 
morphic features. The groundplan vena- 
tion of Platygastroidea, as interpreted from 
basal extant lineages, can also be derived 
readily from the venation characteristic of 
Serphitidae. Kozlov and Rasnitsyn (1979, 
fig. 2; Rasnitsyn et al. 2004, fig. 8) de- 
scribed and illustrated a costal vein and 
a distinct pterostigma in Serphites, but 
indicated that the costal vein was missing 
and the pterostigma was indistinct in 
Microserphitcs parvulus (Kozlov and Rasnit- 
syn 1979, fig. 6). A specimen of Serphites 
from Canadian Cretaceous amber (MCZ: 
5330) has a distinct pterostigma but lacks 
a costal vein. Vein R bifurcates before the 
anterior margin of the wing so that R, 
constitutes the anterobasal and anterior 
margins of the pterostigma and r-rs con- 
stitutes the posterior margin of the pter- 
ostigma (Fig. 226). Except for the ptero- 
stigma, this venation is very similar to the 
possible groundplan venation of Platygas- 



troidea. However, the unusually large 
"pterostigma" of serphitids may be be- 
cause of autapomorphic secondary mela- 
nization of the wing membrane in the 
region between R} and r-rs (Fig. 226). 

The precursor of the marginal vein of 
Chalcidoidea is uncertain. The chalcid 
marginal vein may have evolved through 
a narrowing of a pterostigma in a trans- 
formation similar to that proposed for 
Scelionidae by A. Rasnitsyn. Alternatively, 
it may have evolved through anterior 
elongation of vein R along the wing margin 
similar to the transformation series pro- 
posed for Scelionidae by Masner et al. (in 
press). If the latter, the marginal vein of 
chalcids consists only of R, not C+R (Huber 
and Sharkey 1993, fig. 14; Masner et al. in 
press, fig. 48) or Sc 2 +Ri (Bradley 1955) or 
R! (Burks 1938). Regardless of the correct 
interpretation of the groundplan venation 
of extant Platygastroidea, both hypotheses 
suggest that the marginal vein of Scelioni- 
dae is secondarily derived and therefore 
convergent to that of Chalcidoidea. My- 
marommatoidea also have a recognizable 
marginal vein (Figs 127, 129, 185: mv) that 
appears to have a very short stigmal vein 
distally (Fig. 185: stv). These features may 
be synapomorphic for Mymarommatoidea 
+ Chalcidoidea, but if so they are homo- 
plastic relative to the venation of some 
Platygastroidea. 

The presence of a costal vein in some 
fossils assigned to Scelionidae (Fig. 227: cv) 
and the lack of a costal vein in some 
Serphitidae (Fig. 226) indicates that loss of 
this vein occurred independently in the 
two taxa. Both taxa are known from the 
earliest Cretaceous or latest Jurassic (Ras- 
nitsyn et al. 2004) so perhaps it is un- 
remarkable that both families exhibit a di- 
versity of wing venation. However, the 
earliest fossil "scelionids" are impression 
fossils, from which far less morphological 
information can be deduced than from 
amber fossils. The putative scelionids with 
a pterostigma and a costal vein may only 
resemble and not actually be closely 



Volume 16, Number 1, 2007 109 

related to extant Scelionidae. Further re- lanceolate sensilla originating from a de- 
search is necessary to establish this and to pression distally on the claval and funicu- 
resolve the true groundplan venation of lar segments (Figs 75-77), and according to 
extant Scelionidae. John Heraty (UCRC, pers. comm.) they are 
Further study of the external and in- also present in some Aphelinidae (e.g. 
ternal structure of the different types of Cales Howard). We are not aware of any 
claval sensilla of mymarommatids and reports of similarly shaped sensilla in other 
other apocritans is also required. Mymar- Chalcidoidea, although so called basiconic, 
idae (Figs 8-10: s4) have sensilla that in campaniform, capitate peg or multiporous 
external structure are very similar to the s4- grooved sensilla reported in Agaonidae 
type sensilla of mymarommatids (Figs 58, (Ware and Compton 1992, fig. 3), Eulophi- 
60, 62-64, 66, 68, 70, 71: s4). In Mymaridae dae (Veen and Wijk 1985, fig. 9c), Encyrti- 
they have been called "basiconic" (Baaren dae (Baaren et al. 1996, figs 1, 5f), Pter- 
et al. 1999, figs 2F, 4A, 4B, 4F, 5F), omalidae (Miller 1972, figs 3, 4), and 
"grooved peg" (Chiappini et al. 2001, Trichogrammatidae (Voegele et al. 1975, 
figs la, lc, 16) or "sickle-shaped" (Huber pi. Ill figs 1-4; Olson and Andow 1993, 
and Fidalgo 1997, figs 47, 48) sensilla. figs 1, 11) likely are homologous with s4- 
Chiappini et al. (2001) stated that the clava type sensilla. Even though they have 
of Anagrus atomus (L.) (Mymaridae) has a different external structure, they all 
a single such sensillum near its ventral originate from a circular depression and 
margin almost at its middle, and 1-3 usually at least some are positioned distal- 
sensilla distally on the apical three (of five) ly on the funicular segments similar to s4- 
funicular segments. Baaren et al. (1999) type sensilla in mymarids and mymarom- 
reported that in Anaphes listronoti Huber matids. A reduction in the length and 
and Anaphes victus Huber (Mymaridae) all globular enlargement of the apex of a lan- 
but the first of six funicular segments have ceolate s4-type mymarid or mymaromma- 
one or two of the sensilla distally and both tid sensillum would result in a petiolate, 
claval segments each have two sensilla. mushroom-like sensillum similar to those 
There is at least one s4-type sensilla of the other chalcid families listed above, 
ventroapically on the female clava of The presence of lanceolate s4-type sensilla 
Polynema striaticome Girault (Mymaridae) in Mymaridae and Rotoitidae suggest that 
(Isidoro et al. 1996, fig. 8c). Consequently, these are groundplan features of Chalci- 
mymarids appear to differ from mymar- doidea, but that the shape of the sensillum 
ommatids in having the sensilla not only likely was modified secondarilv in other 
on the clava but also distally on some chalcids. Chiappini et al. (2001, figs lc, 18) 
funicular segments (Figs 9, 10: s4) in called a small peg-like sensillum that is 
a position similar to those of the claval ventroapical in position on the clava of 
segments of male mymarommatids (cf. Anagrus atomus (cf. Fig. 8: s4?) a "sunken 
Figs 60, 61, 66, 68). Female mymarids have peg sensillum". Our very limited survey of 
the sensilla ventrally (Fig. 8), dorsally the sensilla in Mymaridae suggests that the 
(Fig. 9) or both ventrally and dorsally sunken peg sensillum in A. atomus may be 
(Huber and Fidalgo 1997, fig. 48), and in nothing more than a modified s4-tvpe 
some specimens they can be present on sensillum (cf. Figs 8, 9). More comprehen- 
a segment of one antenna but missing from sive surveys are required to determine the 
the same segment of the other antenna exact distribution of lanceolate s4-type 
(Huber and Fidalgo 1997). In addition to sensilla in Chalcidoidea, whether such 
mymarids and mymarommatids, individ- sensilla occur in other apocritan groups, 
uals of both described genera of Rotoitidae and whether it is possible to demonstrate 
(Chalcidoidea) have apically projecting, structural transformation series from 



HO Journal of Hymenoptera Research 

which phylogenetic inferences can be buyuk and Quicke 1999, fig. 5d). However, 

made. parasitic Hymenoptera other than Mymar- 

Further study of the different types of ommatidae also have long, thick, basally 
dorsal (s2-type) and ventral (s3-type) claval bent sensilla. Basibuyuk and Quicke (1999) 
sensilla are also warranted for phylogenet- stated that Ceraphronoidea, Platygastroi- 
ic inference. Mymarid females have a row dea (Isidoro et al. 1996, fig. 2) and some 
or rows of specialized trichoid sensilla Proctotrupoidea, particularly Diapriidae 
along the ventral surface of the clava (Basibuyuk and Quicke 1999, fig. 5a), have 
(Fig. 9: s3), which Baaren et al. (1999) distinctive, curved, multiporous setiform 
called "sensilla chaetica". These sensilla sensilla. Maamingidae also have such 
are similar to the s3-type sensilla of sensilla, each recessed into a shallow lon- 
mymarommatoids (Figs 64: s3; 181b). Baa- gitudinal groove (Early et al. 2001) and 
ren et al. (1999) described four different with the apices of some of the sensilla 
types of sensilla chaetica in AnapJics listro- extending beyond the apex of the segment 
noti, including an apical sensillum (their (Fig. 74). Fusion of the lower surface of the 
type 1), which at least some female sensillum within the longitudinal groove 
mymarommatids have (if. Figs 8, 9 with would result in a mps external structure 
Fig. 64: as; 67, 69). They also differentiated very similar to that characteristic of chal- 
three other morphological types of ventral cids (cf. Figs 74, 77). As noted by Basi- 
sensilla chaetica. Isidoro et al. (1996, fig 8d) buyuk and Quicke (1999, p. 53), mps vary 
suggested that a double row of robust in shape, but are "always embedded in 
sensilla ventrally on the clava of female P. a chamber in the antennal integument". 
striaticome are multiporous gustatory sen- A comprehensive survey of the mps 
silla. Although we did not study the structure of chalcids is required to de- 
morphology of the ventral row or rows of termine their diversity more accurately, 
sensilla in mymarommatids in detail, more but the mps of Mymaridae appear to be 
than one structure of s3-type sensilla is quite typical for Chalcidoidea. The mps are 
evident for at least some species (Fig. 72, structurally diverse in Rotoitidae. Those of 
insert). R. basal is (Fig. 76) are somewhat interme- 

Barlin (1978) proposed that the uniquely diate between a typical chalcid-like mps 

structured multiporous plate sensilla (mps) and a s2-type sensillum. The sensilla are 

of Chalcidoidea evolved from bent multi- non-socketed, as for chalcid mps, but they 

porous setiform (s2-type) sensilla that resemble a very thick s2-type sensillum 

became attached to the cuticle. The pore- because only the oval base is confluent 

canal opening in the proximal end of the with the surface and the longer and more 

mps and the unattached distal end of the slender curved portion, rather than just the 

mps were cited as possible evidence, apex, is free above the surface. Further- 

Kozlov and Rasnitsyn (1979) further sug- more, under higher magnification there is 

gested that the thick, basally curved sen- a fine groove around the base of the 

silla on the clava of female mymaromma- sensillum (Fig. 76) reminiscent of a socket, 

tids (Figs 63, 64: s2) might be precursors of The mps of Chiloc micropteron Gibson and 

the mps of Chalcidoidea (Figs 9, 10: mps). Huber (Fig. 75) are more similar to typical 

They also proposed that the latter evolved chalcid-like mps. An undescribed species 

through accretion of the ventral surface of of Rotoita near R. basalis also has chalcid- 

the sensillum to the surface of the segment like mps, except that on the funicular 

(see also, Basibuyuk and Quicke 1999). Our segments at least some of the convex mps 

study supports such an "accretion" hy- have a distinct groove on either side 

pothesis for the origin of the unique, (Fig. 77), and on the clava the mps origi- 

chalcid-like mps (Gibson 1986 fig. 4; Basi- nate more distinctly within a longitudinal 



Volume 16, Number 1, 2007 



111 



depression. Both the funicular and claval 
segments of the undescribed Rotoitn also 
have much more slender, curved sensilla 
that resemble s2-type sensilla except that 
they are non-socketed (Fig. 77: s2/mps). 
Kozlov and Rasnitsyn (1979) partly differ- 
entiated serphitids from mymarommatids 
based on the absence of strong, curved 
sensilla from the clava of serphitids. The 
serphitids we examined also lacked s2-type 
sensilla. The flagellar segments of serphi- 
tids (Figs 218, 219) appear to be more or 
less uniformly covered with suberect, 
curved, trichoid sensilla, though these 
could be multiporous setiform sensilla. 

Even if chalcid mps were derived from 
s2-type sensilla, the wide distribution of s2- 
type sensilla in parasitic Hymenoptera 
indicates their presence is synapomorphic 
for a larger group of taxa than just 
Mymarommatoidea + Chalcidoidea. In 
Chalcidoidea, both sexes have mps and 
they are on the funicular as well as the 
claval segments. Therefore, their immedi- 
ate common ancestor may also have had 
s2-type sensilla in both sexes on all the 
flagellar segments, unlike Mymarommati- 
dae. Rasnitsyn et al. (2004, fig. 5) suggested 
that lines on the flagellar segments of 
Khutelchalcididae were internal apertures 
of long mps and that these were a synapo- 
morphy for Khutelchalcididae + Chalcidoi- 
dea. It is also possible that the lines are 
remnants of elongate s2-type sensilla sim- 
ilar to those in Maamingidae (Fig. 74). 
According to Barlin (1978), the internal 
aperture extends the length of the mps in 
Chalcidoidea and in some subfamilies of 
Braconidae (Ichneumonoidea). Cynipoidea 
also have elongate mps, but with long 
internal apertures only in Cynipidae (Ras- 
nitsyn et al. 2004). According to Barlin 
(1978), the pore canal for entry of the 
dendrites occurs at the proximal end of the 
sensillum in Chalcidoidea but near the 
center of the sensillum in Cynipoidea and 
Ichneumonoidea. Although the irregular 
lines on the flagellar segments of Khutel- 
chalcididae might be elongate internal 



apertures of mps, they do not show the 
unique features of chalcid mps — internal- 
ly, the proximal position of the pore canal 
and, externally, the sensillum being raised, 
ridge-like, above the surface of the cuticle, 
without an encircling groove, and with the 
distal end free (Basibuyuk and Quicke 
1999, fig. 5d). Additional study is neces- 
sary to confirm the presence of mps in 
other fossil impressions of Khutelchalcidi- 
dae, to determine the internal structure of 
the mps in Rotoitidae, and to assess more 
thoroughly the internal and external struc- 
ture of the flagellar sensilla throughout 
parasitic Hymenoptera. A comprehensive 
survey of the setation/sensilla of the 
metanotum and around the wing bases 
throughout parasitic Hymenoptera might 
also provide additional characters and 
transformation series for phylogenetic in- 
ference. 

Except for Serphitidae, mymarommatids 
are unique in having the second metasomal 
segment tubular (Figs 88, 89). The func- 
tional or ecological significance of a tubular 
second metasomal segment is unknown. 
The post-petiolar metasoma (gaster) of 
Mymarommatoidea is quite similar to that 
of Chalcidoidea. Similarities include cerci 
on the syntergum and absence of spiracles 
from all but the penultimate tergite in the 
groundplan of Mymarommatoidea, ter- 
gites smoothly overlapping the sternites, 
the hypopygium being able to separate 
widely from the syntergum in females, and 
an ovipositor that rotates and extends 
ventrally from the gaster rather than being 
extended posteriorly. However, all of these 
features have a much wider distribution in 
Apocrita and at best support the mono- 
phyly of a much larger group of taxa than 
just Mymarommatoidea + Chalcidoidea. 

Kozlov and Rasnitsyn (1979) stated that 
the gastral tergites protrude laterally some- 
what and the sternites form a longitudinal 
fold in Microserphites. A serphitid in 
Canadian Cretaceous amber (MCZ: 5343) 
has what we interpret as a similar gas- 
tral structure. This specimen belongs to 



112 Journal of Hymenoptera Research 

Serphites based on its wing venation and (MCZ: 5330) has an inflated gaster with the 

position of the posterior ocellus. In ventral tergites and sternites separated by a dis- 

view, the gaster is strongly flattened with tinct, white, membranous band (Fig. 225). 

the ventrolateral aspect of each tergite Under some angles of light, the tergites are 

extending on either side of the sternum slightly angled along a straight line so that 

for a short distance before being abruptly there is a slender region along the tergum 

angled back toward the sternum. Conse- above the narrow membranous band 

quently, the gaster is uniformly carinately (Fig. 225: ltt), which we consider as a band 

margined laterally and ventrally has dis- of differentiated laterotergites. Even less 

tinctly differentiated "laterotergites" form- distinct is a slender band laterally along the 

ing a narrow band on either side of the sternites (Fig. 225: 1st) that is similar in 

sternum (cf. Fig. 224: ltt). The abruptly width to the putative laterotergites. This 

folded tergal structure is most distinct for region appears to consist of a slightly 

the basal gastral tergites because both the depressed part of each sternite that the 

dorsal and ventral surfaces are visible, but respective laterotergite would override if 

it is not possible to determine whether the the gaster was not inflated. Based on 

laterotergites articulate with differentiated serphitids with a flattened gaster and 

laterosternites. The gastral sternum is deeply inflexed laterotergites, our assump- 

composed of six sternites that form a uni- tion is that the differentiated lateral part of 

formly low convex surface; the second the sternites can flex so as to lie over the 

gastral sternite is transverse-rectangular laterotergites, and thus constitute "latero- 

and is the largest sternite, though only sternites". Regardless, the membranous 

slightly longer than the first gastral ster- band separating the sternites and tergites 

nite. Several Taimyr amber serphitids is straight in MCZ: 5330 (Fig. 225), unlike 

(PIN: 3311/80, PIN: 3730/28-30 (Fig. 224), the irregular line of separation between the 

PIN: 3731) have gastral structures very tergites and sternites in taxa that lack 

similar to that described for the MCZ: 5343 laterotergites, such as mymarommatids 

specimen. These Taimyr specimens repre- (Fig. 147) and chalcids. The latter two 

sent a species different from MCZ: 5343 Canadian Cretaceous serphitids (MCZ: 

based on a very different length of the 5256 and 5330) suggest that laterotergites 

second petiolar segment. All the Taimyr are present in serphitids, but that they are 

amber specimens had been identified as not obvious when the gaster is inflated. 

"Serphites sp.". Kozlov and Rasnitsyn Observation of a differentiated lateral re- 

(1979) did not describe laterally differenti- gion on the sternites is only possible when 

ated tergites for Serphites or their new the gaster is inflated and the tergites and 

genus Aposerphites. Some fossils we exam- sternites are separated, 

ined and identify as Serphites have the Austin et al. (2005) stated that mono- 

gaster subcircular in cross-section rather phyly of Platygastroidea is supported by 

than flattened. It usually is not possible to two character systems, gustatory sensilla 

be certain whether the tergites have differ- on the claval segments and structure of the 

entiated laterotergites when the gaster is gaster. In Platygastroidea, the gastral ter- 

subcircular. However, one specimen gites and sternites are connected by later- 

(MCZ: 5256) with an inflated gaster has otergites and the spiracles are reduced and 

a whitish line laterally. This line superfi- non-functional. Vanhorniidae (Proctotru- 

cially divides the tergites from the ster- poidea) also lack gastral spiracles but do 

nites, but under some angles of light what not have laterotergites (Naumann and 

appears to be quadrate plates and the true Masner 1985), whereas Ambositrinae (Dia- 

ventral margins of the tergites are visible priidae) have spiracles on the penultimate 

below the whitish line. Another specimen gastral tergite and laterotergites (Masner 



Volume 16, Number 1, 2007 



113 



1961). Most of the serphitids we examined 
were inappropriately preserved or posi- 
tioned to be confident of the presence or 
absence of spiracles, but under some 
angles of light MCZ: 5330 has a small, 
smooth, circular structure laterally on Mt 7 , 
which led Gibson (1985) to code metasomal 
spiracles as present in Serphitidae. There 
also appears to be a circular structure 
laterally on Mt 7 of PIN: 3730/31 that may 
be a spiracle, which is in a transversely- 
oval depression that is margined anterior- 
ly. The hypopygium was extended ven- 
trally in most female serphitids we exam- 
ined similar to mymarommatids and chal- 
cids. Consequently, if the presence of 
metasomal spiracles is verified for other 
serphitids, then their gastral structure is 
more similar to Ambositrinae (Diapriidae) 
than Platygastroidea, which have a tubular, 
telescoping ovipositor extension system. 
The ambositrine gaster differs by having 
Mt 2 composed of two or three fused terga 
(Masner 1993). 

Masner et al. (in press) discussed four 
genera of Scelionidae that apparently lack 
laterotergites, but at least three of the 
genera have longitudinal sublateral keels 
on the tergites. These keels may represent 
vestiges of the lateral fold that differenti- 
ates the laterotergites of other platygas- 
troids. Furthermore, the ventral margins of 
the tergites and the dorsal margins of the 
sternites form straight lines (Masner et al. 
in press, fig. 3), which are separated by 
membrane when the gaster is inflated 
(Masner et al. in press, fig. 1). This "multi- 
segmented carapace" structure is very 
similar to that of other platygastroids with 
laterotergites and of the MCZ: 5330 serphi- 
tid (Fig. 225). Relationships of the four 
scelionid genera that apparently lack later- 
otergites need to be clarified because they 
are all characterized by a single mesotibial 
spur rather than the groundplan two 
mesotibial spurs (see above). 

The MCZ: 5343 serphitid is a male. The 
metasoma in ventral view has symmetrical, 
elongate-digitiform processes that have at 



least one long terminal seta (Fig. 223: par?). 
The processes extend from either side of 
the sternum at the line of junction between 
Ms 7 and Ms s and in posterior view appear 
to originate laterally between the apical 
tergite and sternite. The processes resemble 
what we consider to be parameres in male 
Zealaromma (Figs 154, 156: par) and Maa- 
minga (Fig. 157: par). Rasnitsyn (1988, node 
65) postulated that a tubular male genital 
capsule with both the volsellae and para- 
meres fused with the aedeagus was syna- 
pomorphic for Chalcidoidea + Mymarom- 
matidae + Serphitidae + Platygastroidea 
and, possibly, the extinct family Juraprii- 
dae. This hypothesis appears to be falsified 
by the presence of moveable parameres in 
the male genitalia of Zealaromma and, 
likely, Serphitidae, but the fused condition 
remains a possible synapomorphy for 
Chalcidoidea + Platygastroidea. 

The metasoma of MCZ: 5343 has seven 
distinct, melanized tergites in dorsal view 
and a small, lighter coloured apical eighth 
tergite that in posterior view originates 
somewhat under Mt 7 . Laterally on this 
apical tergite is a circular, somewhat 
convex structure with several projecting 
setae, which likely is a cercus. Two females 
(PIN: 3730/31; MCZ: 5330) also have setose 
digitiform processes (cf. Gibson 2003, 
figs 26, 42) basolaterally on Mt 8 that likely 
are the cerci (Gibson 1985). The left process 
of PIN: 3730/31 appears to have at least 
four and probably six projecting setae. 

The serphitid head in lateral (Fig. 218) or 
frontal (Fig. 219) view has the antennae 
inserted very low on the face close to the 
oral cavity. In well preserved specimens 
there is a distinctly differentiated clypeus 
and the antennal toruli are separated from 
a linear epistomal suture by a distance 
similar to the diameter of a torulus. This is 
typical of the head structure that charac- 
terizes Platygastroidea (Naumann and 
Masner 1985) and contrasts with the 
higher, possibly plesiomorphic position of 
the toruli in Mymarommatoidea and Chal- 
cidoidea. The mandibles of most serphitids 



114 Journal of Hymenoptera Research 

we examined are quite broad with three Other than the common possession of 

long and slender teeth (Fig. 218; Alonso et a 2-segmented petiole, we found no evi- 

al. 2000, fig. 2), though the holotype of dence supporting a Serphitoidea + Mymar- 

Microserphites parvulus appears to have ommatoidea sister-group relationship. The 

only two slender teeth (Kozlov and Ras- petiolar segments are similar in both taxa, 

nitsyn 1979, fig. 7). Distinct maxillary and though in some serphitids they are longi- 

labial palpi are often visible in serphitids, tudinally strigose and /or extensively se- 

unlike in mymarommatids, though this tose, unlike the petiolar segments of my- 

represents another symplesiomorphy. A marommatids. Newly discovered similari- 

feature that may be phylogenetically in- ties, such as medially abutting propleura 

formative is the presence of an acetabular and independent parameres in the male 

carina or at least an anteriorly differentiat- genitalia have much wider distribution 

ed region of the mesopectus ventrally and therefore are indicated as symplesio- 

behind the procoxae (Fig. 220: ace). This morphies. 

carina is absent from mymarommatids Common possession of lanceolate s4- 

(Fig. 86), mymarids and most other chal- type flagellar sensilla and one or two rows 

cids. Although Naumann and Masner of s3-type sensilla ventrally on the clava of 

(1985) indicated that an acetabular carina females could provide additional support 

was absent in Platygastroidea this was an for monophyly of Mymarommatoidea and 

error — absence or presence of an acetab- Chalcidoidea. However, the diverse exter- 

ular carina in Platygastridae and Scelioni- nal and internal structures of different 

dae should have been listed (see Masner claval and flagellar sensilla found by 

1979). Isidoro et al. (1996) in parasitic Hymenop- 
tera demonstrate that a much more com- 

COINCLUblONb prehensive comparative study is required 

Several structural similarities suggest prior to reliable hypotheses. It seems likely 

a close relationship between Serphitoidea that the marginal veins of at least Chalci- 

and Platygastroidea, though at present doidea and Scelionidae are convergent, 

there are no unequivocal synapomorphies though common possession of a marginal 

for the two taxa. Some similarities, such as vein could be synapomorphic for Mymar- 

presence and position of the mesothoracic ommatoidea + Chalcidoidea. Study of the 

spiracle and two mesotibial spurs are male from lower Cretaceous Spanish am- 

obvious symplesiomorphies, but other ber that Alonso et al. (2000) stated has a 2- 

shared features are not studied sufficiently segmented petiole but non-mymaromma- 

for confident hypotheses of character-state tid wing features, and which they postu- 

distribution and polarity in Serphitidae or lated may be the sister taxon of Mymar- 

Apocrita. These features include a similarly ommatidae + Chalcidoidea (Mymaridae), 

structured gaster having laterotergites and could provide valuable information for 

possibly laterosternites, a differentiated resolving character-transformations and 

mesopectal region that might be homolo- relationships. 

gous with a netrion or the ancestral The extinct Cretaceous taxon Khutelchal- 

structure (prepectus exposed ventral to cididae is indicated to be at most the sister- 

mesothoracic spiracle) from which a netrion group of Chalcidoidea rather than a "basal 

evolved, a forewing venation that could be chalcidoid". The pronotum has an incision 

similar to the groundplan venation of in the posterodorsal margin that suggests 

Scelionidae, position of the toruli relative the mesothoracic spiracle was in the 

to the epistomal suture, and similar meso- plesiomorphic position and even though 

somal sclerotization and sculpture, includ- it may have had a slender, externally 

ing an acetabular carina. visible prepectus ventrally, this did not 



Volume 16, Number 1, 2007 



115 



extend dorsally to the mesoscutum as in 
Chalcidoidea. 

Future studies of higher-level relation- 
ships of Chalcidoidea or Mymarommatoi- 
dea should include Maamingidae as well 
as Platygastroidea and Serphitidae as out- 
groups. Inclusion of Maamingidae is war- 
ranted because of the structure of their s2- 
type antennal sensilla (Fig. 74) compared 
to the uniquely structured mps of Chalci- 
doidea, the presence of independent, rod- 
like parameres in the male genitalia, a fore- 
wing venation (Early et al. 2001, fig. 19) 
similar to that of Rotoitn (Gibson and 
Huber 2000, fig. 46), other similarities 
between the taxa discussed by Early et al. 
(2001), and relationships indicated by 
molecular analyses (Castro and Dowton 
2006). 

ACKNOWLEDGEMENTS 

We thank the individuals and institutions listed 
under Materials and Methods for the loan of speci- 
mens on which our study was based. We also thank 
Lars Vilhelmsen and Lars Krogmann for providing us 
with their complete file of SEM images of M. 
anomalum, and Alex Rasnitsyn (PIN), Norman John- 
son (Ohio State University, Columbus) and Lubomir 
Masner (CNC) for discussions concerning serphitid 
and platygastroid morphology. Thomas Schluter 
kindly granted permission to reproduce Fig. 215, Alex 
Rasnitsyn supplied us with Fig. 227 and Bob Wharton 
(Texas A&M University, College Station) provided us 
with a copy of the PhD thesis of Margaret Barlin. 
Klaus Bolte (CNC) photographed several of the amber 
inclusions used to illustrate the text. Andrew Bennett 
and Henri Goulet (CNC), Lars Vilhelmsen, and an 
anonymous reviewer provided many useful com- 
ments for improvement of a previous version of the 
manuscript. This research was conducted as part of 
the Hymenoptera Tree of Life initiative (National 
Science Foundation grants DEB-0334945 and EF- 
0337220). 

LITERATURE CITED 

Andriescu, I. and I. Suciu. 1963. Mymaromma anomala 
(Blood et Kryger), reprezentanta unei familii de 
himenoptere (Chalcidoidea) noua pentru fauna R. 
P. R. Analele Stiintifice ale Universitatii "Al. I. 
Cuza" din lasi (Seria Nona) 9: 251-255 + 1 pi. 

Alonso, J., A. Arillo, E. Barron, ]. Carmelo Corral, J. 
Grimalt, J. F. Lopez, R. Lopez, X. Martinez- 
Delclds, V. Ortuno, E. Penalver, and P. R. Trincao. 



2000. A new fossil resin with biological inclusions 
in lower Cretaceous deposits from Alava (north- 
ern Spain, Basque-Cantabrian basin), journal of 
Paleontology 74: 158-178. 

Annecke, D. P. and R. L. Doutt. 1961. The genera of 
the Mymaridae. Hymenoptera: Chalcidoidea. 
Entomology Memoirs, Department of Agricultural 
Technical Services, Republic of South Africa 5: 1-71. 

Askew, R. R., J. Blasco-Zumet, and ]. Pujade-Villar. 

2001. Chalcidoidea y Mymaromma toidea (Hyme- 
noptera) de un sabinar de Juniperus thurifera L. en 
Los Monegros, Zaragoza. Monografias de la Socie- 
dad Entomoldgica Aragonesa 4: 1-76. 

Austin, A. D. and S. A. Field. 1997. The ovipositor 
system of scelionid and platygastrid wasps 
(Hymenoptera: Platygastroidea): comparative 
morphology and phylogenetic implications. In- 
vertebrate Taxonomy 11: 1-87. 

, N. F. Johnson, and M. Dowton. 2005. 

Systematics, evolution, and biology of scelionid 
and platygastrid wasps. Animal Review of Ento- 
mology 50: 553-582. 

Baaren, J. van., R. Barbier, and J. -P. Nenon. 1996. 
Female antennal sensilla of Epidinocarsis lopezi 
and Leptomastix dactylopii (Hymenoptera: Encyr- 
tidae), parasitoids of pseudococcid mealybugs. 
Canadian journal of Zoology 74: 710-720. 

, G. Boivin, J. Le Lannic, and J.-P. Nenon. 1999. 

Comparison of antennal sensilla of Anaphes victus 
and A, listronoti (Hymenoptera, Mymaridae), egg 
parasitoids of Curculionidae. Zoomorphology 119: 
1-8. 

Bakkendorf, O. 1948. A comparison of a mymarid 
from Baltic amber with a recent species, Petiolaria 
anomala (Micro-Hym.). Entomologiske Meddeleleser 
25: 213-218. 

Barlin, M. R. 1978. Multiporous plate sensilla in parasitic 
Hymenoptera: their ultrastructure and phylogenetic 
relationship. PhD Dissertation, Texas A&M Uni- 
versity, College Station, Texas. 126 pp. 

Basibuyuk, H. H. and D. L. J. Quicke. 1995. Morphol- 
ogy of the antenna cleaner in the Hymenoptera 
with particular reference to non-aculeate families 
(Insecta). Zoologica Scripta 24: 157-177. 

and , 1997. Hamuli in the Hymenoptera 

(Insecta) and their phylogenetic implications. 
Journal of Natural History 31: 1563-1585. 

and , 1999. Gross morphology of multi- 
porous plate sensilla in the Hymenoptera (In- 
secta). Zoologica Scripta 28: 51-67. 

, , A. P. Rasnitsyn, and M. G. Fitton. 



2000. Morphology and sensilla of the orbicula, 
a sclerite between the tarsal claws, in the 
Hymenoptera. Annals ot the Entomological Society 
of America 93: 625-636. 
Beardsley, J. W., J. T. 1 luber, and W. D. Perreira. 2000. 
Mvmarommatoidea, a superfamily of Hymenop- 
tera new for the Hawaiian Islands. Proceedings of 
the Hawaiian Entomological Society 34: 61-63. 



116 



Journal of Hymenoptera Research 



Blood, B. N. and T. P. Kryger. 1922. A new mymarid 
from Brockenhurst. Entomologist's Monthly Maga- 
zine 58: 229-230. 

and , 1936. Pctiolaria anomala Bl. & Kr. 

(Hym., Chalcid.): description of the female. 
Journal of the Society for British Entomology 1: 
115-116 + plate III. 

Bradley, J. C. 1955. The wing-venation of Chalcidoidea 
and of some allied Hymenoptera. Memoires de la 
Societe Royale d'Entomologie de Belgique 27: 
127-138. 

Brues, C. T. 1937. Superfamilies Ichneumonoidea, 
Serphoidea, and Chalcidoidea. Pp. 27-44 in 
Carpenter, F. M, J. W. Folsom, E. O. Essig, A. 
C. Kinsey, C. T. Brues, M. W. Boesel, and H. E. 
Ewing. Insects and arachnids from Canadian 
amber. University of Toronto Studies, Geological 
Series 40: 7-62. 

Burks, B. D. 1938. A study of chalcidoid wings 
(Hymenoptera). Annals of the Entomological Society 
of America 31: 157-161. 

Campbell, B., J. M. Heraty, J.-Y. Rasplus, K. Chan, J. 
Steffen-Campbell, and C. Babcock. 2000. Molecu- 
lar systematics of the Chalcidoidea using 28S-D2 
rDNA. Pp. 59-73 in Austin, A. D., and M. 
Dowton, eds. Hymenoptera: evolution, biodiversity 
and biological control. CSIRO publishing, Colling- 
wood. 468 pp. 

Castro, L. R. and M. Dowton. 2006. Molecular analyses 
of the Apocrita (Insecta: Hymenoptera) suggest 
that the Chalcidoidea are sister to the diaprioid 
complex. Invertebrate Systematics 20: 603-614. 

Chiappini, E. and E. Mazzoni. 2000. Differing mor- 
phology and ultrastructure of the male copulato- 
ry apparatus in species-groups of Anagrus Hali- 
day (Hymenoptera: Mymaridae). Journal of Natu- 
ral' History 34: 1661-1676. 

, C. Solinas, and M. Solinas. 2001. Antennal 

sensilla of Anagrus atomus (L.) (Hymenoptera: 
Mymaridae) female and their possible beha- 
vioural significance. Entomologica, Bari 35: 51-76. 

Clouatre, A., D. Coderre, and D. Gagnon. 1989. 
Habitat of a new Mymarommatidae found in 
southern Quebec, Canada (Hymenoptera: Tereb- 
rantes). The Canadian Entomologist 12: 825-826. 

Dahms, E. C. D. 1984. A checklist of the types of 
Australian Hymenoptera described by Alexandre 
Arsene Girault: III. Chalcidoidea species F-M 
with advisory notes. Memoirs of the Queensland 
Museum 21: 579-842. 

Debauche, H. R. 1948. Etude sur les Mymarommidae 
et les Mymaridae de la Belgique (Hymenoptera 
Chalcidoidea). Memoires du Musee Royal d'Histoire 
Naturelle de Belgique 108: 1-248 + 24 plates. 

Donev, A. 1982. Mymarommidae. Nauchni Trudove, 
Biologiya. Plovdivski Universitet "Paissi Hilde- 
narski" 20: 483-484. 

Doutt, R. L. 1973. The fossil Mymaridae. Pan-Pacific 
Entomologist 49: 221-228. 



Dowton, M. and A. D. Austin. 2001. Simultaneous 
analysis of 16S, 28S, COl and morphology in the 
Hymenoptera: Apocrita - evolutionary transi- 
tions among parasitic wasps. Biological journal of 
the Linnean Society 74: 87-11 1. 

Duisburg, P. von. 1868. Zur Bernstein-Fauna. Schriften 
der Physikalisch-okonomischen Gesellschaft zw Ko- 
nigsberg 9: 23-28. 

Early, J. W., L. Masner, I. D. Naumann, and A. D. 
Austin. 2001. Maamingidae, a new family of 
proctotrupoid wasp (Insecta: Hymenoptera) from 
New Zealand. Invertebrate Taxonomy 15: 341-352. 

Ferriere, C. 1948. LJn curieux Mymaride: Pctiolaria 
anomala Bl. et Kr. Mitteilungen der Schweizerischen 
Entomologischen Gesellschaft 21: 555-556. 

Fidalgo, P. and L. De Santis. 1982. Una nueva especie 
de mimarido de la subfamilia Mymaromminae 
(Insecta, Hymenoptera). Revista del Museo de La 
Plata (Nueva Serie), Zoologia 13 (127): 1-6. 

Fursov, V., Y. Shirota, T. Nomiya, and K. Yamagishi. 
2002. New fossil mymarommatid species, Palaeo- 
mymar japonicum sp. nov. (Hymenoptera: Mymar- 
ommatidae), discovered in Cretaceous amber 
from Japan. Entomological Science 5: 51-54. 

Gallego, V. C. 1986. Notes on the parasitoid complex 
of the coconut leafminer Promecothcca cumingii 
Baly in some Mindanao provinces, Philippines. 
Philippine journal of Coconut Studies 11: 4-7. 

Garcia, J. L. 2000. Nuevos registros genericos para 
Venezuela de Hymenoptera Parasitica. Boletin de 
Entomologia Venezolana 15: 113-117. 

Gauld, I. D., and B. Bolton, eds. 1988. The Hymenoptera. 
Oxford University Press, Oxford. 332 pp. 

Gibson, G. A. P. 1985. Some pro- and mesothoracic 
structures important for phylogenetic analysis of 
Hymenoptera, with a review of terms used for the 
structures. The Canadian Entomologist 117: 
1395-1443. 

. 1986. Evidence for monophyly and relation- 
ships of Chalcidoidea, Mymaridae, and Mymar- 
ommatoidea (Hymenoptera: Terebrantes). The 
Canadian Entomologist 118: 205-240. 

. 1993. Superfamilies Mymarommatoidea and 

Chalcidoidea. Pp. 570-655 in Goulet, H., and |. 
Huber, eds. Hymenoptera, an identification guide to 
families. Agriculture Canada Research Branch 
Monograph No. 1894E, Ottawa. 668 pp. 

. 1997. Chapter 2. Morphology and terminolo- 
gy. Pp. 16-44 in: Gibson, G. A. P., J. T. Huber, and 
J. B. Woolley, eds. Annotated Keys to the Genera of 
Nearctic Chalcidoidea (Hymenoptera). NRC Re- 
search Press, Ottawa, Ontario. 794 pp. 
— . 2003. Phylogenetics and classification of 
Cleonyminae (Hymenoptera: Chalcidoidea: Pter- 
omalidae). Memoirs on Entomology, International 
16: i-v + 339 pp. 

and J. T. Huber. 2000. Review of the family 

Rotoitidae (Hymenoptera: Chalcidoidea), with 



Volume 16, Number 1, 2007 



117 



description of a new genus and species from 

Chile. Journal of Natural History 34: 2293-2314. 

— J. M. Heraty, and J. B. Woolley. 1999. 

Phylogenetics and classification of Chalcidoidea 

and Mymarommatoidea — a review of current 

concepts (Hymenoptera, Apocrita). Zoologica 

Scripta 28: 87-124. 
Girault, A. A. 1920. New genera and species of 

chalcid-flies from Australia. Insecutor Inscitiae 

Menstruus 8: 37-50. 
. 1930. Nezv pests from Australia, VIII. Privately 

printed, Brisbane. 5 pp. 
. 1931. A nezv habit in an old insect. Homo pudicus 

and nezv Eurytomidae. Privately printed, Brisbane. 

4 P p. 
. 1935. Microhymenoptera Aitstraliensis nova, 

mostly Chalcididae. Privately printed, Sydney. 4 

PP- 

Grimaldi, D. A. and M. S. Engel. 2005. Evolution of the 
Insects. Cambridge University Press, Cambridge, 
etc. 772 pp. 

, , and P. C. Nascimbene. 2002. Fossilif- 

erous Cretaceous amber from Myanmar (Burma): 
its rediscovery, biotic diversity, and paleontolog- 
ical significance. American Museum Novitates 3361: 
1-72. 

Hansen, L. O. 1997. Palaeomymar duisburgi (Stein, 1877) 
(Hym., Mymarommatoidea): a species and su- 
perfamily new to the Norwegian fauna. Fauna 
Norvegica, Series B 44: 81-82. 

Huber, J. T. 1987. Premiere mention en Suisse de la 
famille Mymarommatidae (Hymenoptera). Mit- 
teilungen der Schweizerischen Entomologischen Ge- 
sellschaft 60: 82. 

. 2005. The gender and derivation of genus- 
group names in Mymaridae and Mymarommati- 
dae (Hymenoptera). Acta Societatis Zoologicae 
Bohemoslovenicae 69: 167-183. 

and P. Fidalgo. 1997. Review of the genus 

Stephanodes (Hymenoptera: Mymaridae). Proceed- 
ings of the Entomological Society of Ontario 128: 
27-63. 

and M. J. Sharkey. 1993. Chapter 3. Structure. 

Pp. 13-59 in: Goulet, H., and J. Huber, eds. 
Hymenoptera, an identification guide to families. 
Agriculture Canada Research Branch Monograph 
No. 1894E, Ottawa. 668 pp. 

Isidoro, N., F. Bin, S. Colazza, and S. B. Vinson. 1996. 
Morphology of antennal gustatory sensilla and 
glands in some parasitoid Hymenoptera with 
hypothesis on their role in sex and host recogni- 
tion. Journal of Hymenoptera Research 5: 206-239. 

Kalina, V. 1989. Checklist of Czechoslovak Insects III 
(Hymenoptera). Chalcidoidea. Acta Faunistica 
Entomologica Musei Nationalis Pragae 19: 127. 

Konigsmann, E. 1978. Das phylogenetische System der 
Hymenoptera. Teil 3: "Terebrantes" (Unteror- 
dung Apocrita). Deutsche Entomologische Zeits- 
chrift (N. F.) 25: 1-55. 



Kozlov, M. A. and A. P. Rasnitsyn. 1979. On the limits 
of the family Serphitidae (Hymenoptera, Procto- 
trupoidea). Entomologischeskoe Obozrenie 58: 
402-416. [In Russian] 

Lin, N. 1994. First discovery of Mymarommatidae 
(Hymenoptera) from China, with description of 
a new species. Entomotaxonomia 16: 120-126. 

Liu, Z. and G. Nordlander, G. 1992. Ibaliid parasitoids 
of siricid woodwasps in North America: two new 
Italia species and a key to species (Hymenoptera: 
Cynipoidea). Proceedings of the Entomological Soci- 
ety of Washington 94: 500-507. 

Masner, L. 1961. Ambositrinae, a new subfamily of 
Diapriidae from Madagascar and Central Africa. 
Memoirs de I'lnstutute Scientifique de Madagascar, 
Serie E 12: 289-295. 

. 1979. Pleural morphology in scelionid wasps 

(Hymenoptera: Scelionidae) — an aid in higher 
classification. The Canadian Entomologist 111: 
1079-1087. 

. 1993. Superfamily Proctotrupoidea. Pp. 

537-557 in: Goulet, H., and J. Huber, eds. 
Hymenoptera, an identification guide to families. 
Agriculture Canada Research Branch Monograph 
No. 1894E, Ottawa. 668 pp. 
— , N. F. Johnson, and A. D. Polaszek. 2007. 



Redescription of Archaeoscelio Brues and descrip- 
tion of three new genera of Scelionidae 
(Hymenoptera): a challenge to the definition of 
the family. American Museum Novitates 3550: 1-24. 

Mathot, G. 1966. Contribution a la connaissance des 
Mymaridae et Mvmarommidae d'Afrique Cen- 
trale (Hymenoptera Chalcidoidea). Bulletin and 
Annates Socicte Royal d'Entomologie de Belgique 102: 
213-239. 

Meunier, F. 1901. Contribution a la faune des 
Mymaridae ou 'atomes ailes' de l'ambre. Annates 
de la Societe Scientifique de Bruxelles 25: 282-292. 

Miller, M. C. 1972. Scanning electron microscope 
studies of the flagellar sense receptors of Perides- 
mia discus and Nasonia vitripennis (Hymenoptera: 
Pteromalidae) Annals of the Entomological Society of 
America 65: 1119-1124. 

Naumann, I. D. and L. Masner. 1985. Parasitic wasps 
of the proctotrupoid complex: a new family from 
Australia and a key to world families (Hyme- 
noptera: Proctotrupoidea sensu lato). Australia)! 
journal of Zoology 33: 761-783. 

Nikol'skaya, M. N. 1978. Family Serphitidae (Serphi- 
rida). Pp. 1179-1180 in Medvedev, G. S. ed.-in- 
chief. 1987. Keys to the insects of the European part of 
the USSR. Vol. Ill, Hymenoptera, Part 11. Amerind 
Publishing Co., New Dehli 1341 pp. [Lranslated 
from Russian original! 

Noyes, J. S. and E. W. Valentine. 1989. Chalcidoidea 
(Insecta: Hymenoptera) — introduction, and 
review of genera in smaller families. Fauna of 
Nezv Zealand 18: 91 pp. 



118 



Journal of Hymenoptera Research 



Olson, D. M. and D. A. Andow. 1993. Antennal 
sensilla of female Trichogramma nubilale (Ertle and 
Davis) (Hymenoptera: Trichogrammatidae) and 
comparisons with other parasitic Hymenoptera. 
International Journal of Insect Morphology and 
Embryology 22: 507-520. 

Rasnitsyn, A. P. 1980. The origin and evolution of 
hymenopterous insects. Trudy Paleontologicheskogo 
Institute Akademiya Nauk SSSR 174: 1-190. [In 
Russian.] 

. 1988. An outline of evolution of the hyme- 
nopterous insects (Order Vespida). Oriental In- 
sects 22: 115-145. 

. 2002. Superorder Vespida Laicharting, 1781. 

Order Hymenoptera Linne, 1758. Pp. 242-254 in 
Rasnitsyn, A. P., and D. L. J. Quicke, eds. History 
of Insects. Kluwer Academic Publishers, Dor- 
drecht, etc. 517 pp. 
-, H. H. Basibuyuk, and D. L. J. Quicke. 2004. A 



basal chalcidoid (Insecta: Hymenoptera) from the 
earliest Cretaceous or latest Jurassic of Mongolia. 
Insect Systematics and Evolution 35: 123-135. 

Ronquist, F. and G. Nordlander. 1989. Skeletal 
morphology of an archaic cynipoid, Ibalia rufipes 
(Hymenoptera: Ibaliidae). Entomologica Scandina- 
vica, Supplement 33: 1-60. 

, A. P. Rasnitsyn, A. Roy, K. Eriksson, and M. 

Lindgren. 1999. Phylogeny of the Hymenoptera: 
a cladistic reanalysis of Rasnitsyn's (1988) data. 
Zoologica Scripta 28: 13-50. 

Schluter, T. 1978. Zur Systematik und Palokologie 
harzkonservierter Arthopoda einer Taphozonose 
aus dem Cenomanium von NW-Frankreich. 
Berliner Geowissenschaftliche Abhandlungen (A) 9: 
1-150. 

and R. Kohring. 1990. Sie zwergwespengat- 

tung Palaeomymar (Hymenotpera: Proctotrupoi- 
dea: Serphitidae) aus dem mio/pliozanen Simetit 
Siziliens. Berliner Gcowissenschaftliche Abhandlun- 
gen (A) 124: 115-121. 

Soyka, W. 1937. Mymaromma anomala (Kryger) 9 
(Erstbeschreibung des 9) (Mymaridae, Chalc. 
Hym.). Natuurhistorisch Maandblad, Maastricht 26: 
23-24. 

Stein, J. P. E. F. 1877. Drei merkwiirdige Bernstein- 
Insekten. Mitteilungen der Munchener entomolo- 
gischen Verhandlung 1: 28-30. 

Triapitsyn, S. V. and V. V. Berezovskiy. 2000. Super- 
fam. Mymarommatoidea 27. Fam. Mymaromma- 
toidea. Opredcliteli nasckoniikh dalinego vostoka 
Rossii 4: 579. [In Russian.] 

and , 2006. A new species of the genus 

Palaeomymar Meunier, 1901 (Hymenoptera: My- 
marommatidae) from the Russian Far East, with 
notes on other Palaearctic species. Far Eastern 
Entomologist 159: 1-8. 

Valentine, E. W. 1971. Entomology of the Aucklands 
and other islands south of New Zealand. Pacific 
Insects Monographs 27: 327-333. 



Veen, J. C. van. and L. E. van Wijk. 1985. The unique 
structure and functions of the ovipositor of the 
non-paralyzing ectoparasitoid Colpoclypeus floras 
Walk. (Hym., Eulophidae) with special reference 
to antennal sensilla and immature stages. Zeits- 
chrift fiir angewandte Entomologie 99: 511-531. 

Vidal, S. 2001. Entomofauna Germanica. Band 4. 
Verzeichnis der Hautflugler Deutschlands. Chal- 
cidoidea. Entomologische Nachrichtcn und Berichte 
Beiheft 7: 61. 

Viggiani, G. 1966. Ricerche sugli Hymenoptera Chal- 
cidoidea. VI. Generi e specie nuovi per l'entomo- 
fauna italiana (Encyrtidae, Aphelinidae, Mymar- 
ommatidae). Bollettino del Laboratorio di Entomolo- 
gia Agraria 'Filippo Silvestri', Portici 24: 84-105. 

Vilhelmsen, L. and L. Krogmann. 2006. Skeletal 
anatomy of the mesosoma of Palaeomymar anom- 
alum (Blood & Kryger, 1922) (Hymenoptera: 
Mymarommatidae). Journal of Hymenoptera Re- 
search 15: 290-306. 

Voegele, J., J. Cals-Usciati, J. -P. Pihan, and J. Daumal. 
1975. Structure de l'antenne femelle des Tricho- 
grammes. Entoniophaga 20: 161-169. 

Ware, A. B. and S. G. Compton. 1992. Repeated 
evolution of elongate multiporous plate sensilla 
in female fig wasps (Hymenoptera: Agaonidae: 
Agaoninae). Proceedings Nederlandsc Akademie van 
Wetenschappen 95: 275-292. 

Yoshimoto, C. M. 1975. Cretaceous chalcidoid fossils 
from Canadian amber. The Canadian Entomologist 
107: 499-528. 

. 1984. The insects and arachnids of Canada, Part 

12. The families and subfamilies of Canadian chalci- 
doid zoasps. Publication 1760, Canadian Govern- 
ment Publishing Centre, Supply and Services 
Canada, Ottawa. 149 pp. 



Appendix I. Abbreviations used for structures in the 
illustrations. 



ace: 


acetabular carina 


aea: 


aedeagal apodeme 


aed: 


aedeagus 


am: 


anterior margin 


an: 


anus 


anp: 


anal plate 


as: 


apical sensillum of clava 


asc: 


anterior scutellum 


axp: 


axillar phragma 


be: 


basicoxite of metacoxa 


c: 


carina 


ca: 


calcar 


cc: 


costal cell 


cer: 


cercus 


cl: 


clava 


cl n : 


claval segment 


cs: 


campaniform sensillum 



Volume 16, Number 1, 2007 



119 



ct: 


coxal trough 


cv: 


costal vein 


ex: 


coxa 


d: 


dorsal mandibular surface 


dig: 


digitus 


dis: 


digital spine 


dms: 


dorsal metapleural sensillum 


dsc: 


disc of forewing 


dt: 


dorsal tooth of mandible 


fc: 


fine comb 


is: 


flagellar setae 


frp: 


frontal plate of head capsule 


fs: 


flagellar setae 


fu: 


funicle 


fu n : 


funicular segment 


gP : 


gonopore 


ham: 


hamulus 


hp: 


humeral plate 


hwb: 


base of hind wing 


hyp: 


hypopygium 


i: 


inner mandibular surface 


isp: 


intersegmental pit 


lab: 


labium 


lbr: 


labrum 


1pm: 


prementum of labium 


1st: 


laterosternite 


ltt: 


laterotergite 


max: 


maxilla 


mb: 


membrane 


me 


macrochaeta of forewing 


mg: 


mandibular groove 


mms: 


mesopleural-metapleural suture 


mn: 


mesopleural notch 


mnd: 


mandible 


mps: 


multiporous plate sensillum 


ms: 


paramedial setae of occipital plate 


msc: 


mesoscutum 


Mt n : 


metasomal tergite 


mv: 


marginal vein 


mxg: 


galea of maxilla 


mxp: 


maxillary palpus 


mxs: 


stipes of maxilla 


no 3 : 


metanotum 


o: 


outer mandibular surface 


ocp: 


occipital plate of head capsule 


op: 


opposing process of hind wing 


opp: 


pit on occipital plate 


ov: 


ovipositor 


ovs: 


ovipositor sheath 


paa: 


parameral apodeme 


pap: 


process of labiomaxillary complex 


par: 


para mere 


pas: 


postalar sensillum 


pi: 


propodeal flange 


PgP : 


postgenal plate of head capsule 


Pl 3 : 


metapleuron 


pl 2 p: 


mesopleural pit 


pl 3 p: 


metapleural pit 


pn: 


pronotal notch 



ppm: 


propleural margin 


pro: 


propodeum 


pre: 


prepectus 


prs: 


propodeal seta 


psc: 


posterior scutellum 


ps: 


pseudospur 


pss: 


prespiracular sensillum 


pst: 


pterostigma 


pt n : 


petiolar segment 


prp: 


posterior tentorial pits 


ret: 


retinaculum 


sl-4: 


type 1, 2, 3 or 4 flagellar sensillum 


sa: 


subalar area 


sas: 


subalar sensilla of mesopleuron 


sbt: 


subbasal tooth 


sep: 


setal patch 


smv: 


submarginal vein 


sp: 


spiracle 


spa: 


spiracular aperture 


sp P : 


spiracular peritreme 


sps: 


suprapleural sensillum 


std: 


forewing stalk, distal part 


stk: 


forewing stalk 


stp: 


forewing stalk, proximal part 


stv: 


stigmal vein 


syn: 


syntergum 


tg: 


tegula 


tra: 


trachea 


voa: 


volsellar apodeme 


vol: 


volsella 


vt: 


ventral tooth of mandible 



Appendix II. Synopsis of extant material examined. 



1. 



2. 



3. 



Mymaromma Blood and Kryger 
Mymaromma anomalum (Blood and Kryger) (nu- 
merous individuals, CIRAD, CNC, NHRS, 
UCRC): Nearctic - - Canada (ON). Palaearctic 
— Czech Republic, England, France, Germany, 
Hungary, Italy, japan, Poland, Russia, Sweden. 
Mymaromma buyckxi Mathot (79, 4 ]; CNC, ISNB): 
Afrotropical - Congo, Gabon, Madagascar, 
Nigeria. 

Mymaromma goethei Girault (numerous individu- 
als, AN1C, BPBM, CNC, UCRC): Australasian — 
Australia (QLD, NT, WA), ?Hawaii, Papua New 
Guinea. 

4. Mymaromma mirissimum Girault (numerous indi- 
viduals, ANIC, CNC): Australasian — Australia 
(ACT, QLD, NT, WA), Christmas Island, Norfolk 
Island. 

5. Mymaromma ypt (Triapitsvn & Berezovskiy) (339, 
11 J; CNC, UCRC): Oriental -- China (Fujian). 
Palaearctic — Russia, South Korea. 

6. Mymaromma sp. (numerous individuals, CNC, 
UCRC): Oriental: Taiwan, Thailand. 



120 



Journal of Hymenoptera Research 



7. Mymaromma sp. (numerous individuals, CNC): 
Australasian — Indonesia (Ceram), New Caledo- 
nia. Oriental — Indonesia (Sulawesi, Sumatra), 
Malaysia (Sabah), ?Nepal, Philippines (Luzon, 
Negros), Taiwan, Thailand. 

8. Mymaromma sp. (39, 2c?; CNC, UCRC): Neotrop- 
ical — Bermuda, Colombia, Mexico. 

9. Mymaromma sp. (19, 26*; CNC): Neotropical — 
Brazil. 

10. Mymaromma sp. (229, 26c?; ANIC, NZAC): 
Australasian — Chatham Island, New Zealand, 
Norfolk Island. 

Mymaromella Girault 

11. Mymaromella chaoi (Lin) (29; FAUF): Oriental — 
China. 

12. Mymaromella cyclopterus (Fidalgo and De Santis) 
(19, MLPA): Neotropical — Brazil. 

13. Mymaromella mira Girault (209, 35c?; ANIC, 
CNC, UCR): Australasian - - Australia (ACT, 
SA, WA). 

14. Mymaromella sp. (439, 2c?; CNC, USNM): Nearctic 

— Canada (ON), USA (GA, NC, NY, MD, MI, SC, 
VA). 

15. Mymaromella sp. (69; CNC): Nearctic — Canada 
(NB, ON, PQ). 

16. Mymaromella sp. (39; CNC, UCRC): Australasian 

— Australia (ACT). 

17. Mymaromella sp. (119; ANIC, BMNH): Australa- 
sian — Australia (TAS, WA). 

18. Mymaromella sp. (79, 1 J; ANIC, BMNH, UCRC): 
Australasian — Australia (SA, WA). 

19. Mymaromella sp. (79; CNC): Neotropical — Brazil, 
Trinidad, Venezuela. 

20. Mymaromella sp. (19; BMNH): Afrotropical — 
Ivory Coast. 

21. Mymaromella sp. (69; BMNH): Afrotropical — 
Ivory Coast. 

22. Mymaromella sp. (29; CNC): Palaearctic — 
Sweden. 

23. Mymaromella sp. (19; ANIC): Australasian — 
Australia (WA). 

Zealaromma Gibson, Read and Huber 

24. Zealaromma insulare (numerous individuals; CNC, 
NZAC): Australasian — New Zealand. 

25. Zealaromma valentinei (I69, 5c?; CNC, NZAC): 
Australasian — New Zealand. 



Appendix III. Character state summary of 
Mymaromma toidea. 

1. Head capsule: (0) uniformly sclerotized, a single 
structure; (1) with a hyperoccipital band of 
pleated membrane differentiating moveable 
occipital plate from frontal plate. 

2. Mandibular structure: (0) laterally thin, with 
outer surface convex and apices broadly over- 



lapping when closed; (1) laterally thick, with 
outer surface convex and apices not meeting 
when closed (exodont). 

3. Number of mandibular teeth: (0) two; (1) three. 

4. Paramedial setae on occipital plate: (0) absent; 
(1) present. 

5. Number of supraclypeal interorbital setae: (0) 4; 
(1)2. 

6. Ocelli: (0) present; (1) absent. 

7. Width of labium: (0) about as wide as maxilla; 

(1) about twice as wide as maxilla. 

8. Relative position of maxillary palpus and galea: 
(0) palpus originating from subapical lobe on 
inner part of galea; (1) palpus originating 
apically on lobe ventral to galea. 

9. Number of claval segments of female: (0) 4; (1) 3; 

(2) 2; (3) 1. 

10. Structure of multi-segmented clava: (0) seg- 
ments distinctly separated, forming loosely 
associated clava; (1) segments compacted, 
forming tube-like clava. 

11. Number of funicular segments of female: (0) 7; 
(1)6. 

12. Number of s4-type claval sensilla of female: (0) 
2; (1)3. 

13. Position of s4-type claval sensilla of female: 

(0) near dorsal margin; (1) near midline or 
below. 

14. Structure of propleura: (0) abutting, but sepa- 
rated or distinguished medially by distinct line; 

(1) fused into carapace. 

15. Condition of meso- and metapleuron: (0) 
separated by suture; (1) completely fused or 
separated by suture only ventrally. 

16. Condition of metanotum: (0) independent from 
metapleuron and propodeum; (1) fused laterally 
to metapleuron; (2) fused posterolateral^ to 
propodeum; (3) fused to both metapleuron and 
propodeum. 

17. Metapleural pit: (0) present; (1) absent. 

18. Position of metapleural pit when present: (0) 
distinctly nearer spiracle than ventral margin of 
pleuron; (1) about midway between spiracle 
and ventral margin of pleuron. 

19. Propodeal flange: (0) present laterally as vertical 
flange but incomplete dorsally; (1) complete 
laterally and dorsally, fl-like. 

20. Spiracular peritreme: (0) slit-like; (1) evident 
only as slender, smooth band of cuticle. 

21. Spiracular aperture: (0) circular to oval; (1) 
elongate, slit-like. 

22. Pattern of marginal setae along posterior margin 
of forewing: (0) with at least three moderately 
long basal setae; (1) with conspicuously long 
basal seta proximal to several very short setae; (2) 
with all setae very short basally. 

23. Foretibial calcar: (0) comparatively long, curved 
and apically bifurcate; (1) comparatively short, 
straight and simple. 



Volume 16, Number 1, 2007 121 

24. Posterior surface of mesofemur: (0) without 27. Cerci: (0) present, though sometimes 
bumps; (1) with bumps. partly integrated into syntergal surface; (1) 

25. Seta laterally on first petiolar segment: (0) absent. 

present; (1) absent. 28. Male genitalia: (0) with external parameres; ( 1 ) 

26. Metasomal spiracle: (0) present; (1) absent. without external parameres. 



122 



Journal of Hymenoptera Research 




Figs 2-12. 2-7, Mi/mtir schwanni: 2, forewing stalk and hind wing, ventral; 3, wing coupling, dorsal; 4, forewing 
stalk sensilla; 5, distal part of forewing stalk; 6, hind wing apex (a: ventral, b, dorsal); 7, forewing disc, dorsal. 8 
and 9, 9 clava: 8, Eustochus atripennis; 9, M. schwanni. 10, M. schwanni, 3 flagellar segments. 11 and 12, 
dorsolateral view of metanotum, propodeum and base of hind wing: 11, Ptilomymar sp. (insert: enlargement of 
lateral metanotal setae); 12, Mymar taprobanicum. 



Volume 16, Number 1, 2007 



123 




Figs 13-22. 13-15, head capsule: 13, Mymaromella mira (ocp expanded); 14, Mymaromma anomalum (ocp 

normal); 15, Mymaromma buyckxi (ocp indexed). 16 and 17, M. anomalum, articulation of occipital and frontal 
plates: 16, posterolateral; 17, lateral. 18, M. buyckxi, mouthparrs and propleura. 19, M. anomalum, head, 
anteroventral (insert: posterior tentorial pits). 20-22, Mymaromma spp., labiomaxillary complex: 20, Al. anomalum, 
ventral; 21, M. sp. 7, anteroventral; 22, M. sp. 6, anterolateral. 



126 



Journal of Hymenoptera Research 




Figs 42-52. 42, Mymaromma sp. 8, occipital plate and pleated membrane. 43 and 44, Mymaromella spp., 
mouthparts and propleura: 43, M. sp. 15; 44, M. sp. 14. 45, Mymaromella sp. 17, head and mesosoma, dorsolateral. 
46, Mymaromella sp. 16, head, dorsal. 47-49, Mymaromella spp., face and mouthparts: 47, M. sp. 20; 48, M. sp. 14; 
49, M. sp. 17. 50 and 51, Zcalawmma valentinei: 50, head capsule, posterior; 51, articulation of occipital and frontal 
plates, posterior. 52, Zealaromma insulare, face and mouthparts (insert: subtorular setae). 



Volume 16, Number 1, 2007 



127 




Figs 53-64. 53 and 54, Zealaromma ualentinei, head: 53, posterodorsal; 54, frontolateral. 55-57, Zealaromma spp., 
labiomaxillary complex: 55, Z. insulare, ventral; 56, Z. valentinei, anterior; 57, Z. valentinei, lateral. 58-61, 
Mymaromma spp., clava: 58, M. sp. 9 9; 59, M. sp. 10 9; 60, M. goethei ]; 61, M. anomalum J. 62-64, Mymaromella 
spp., clava, outer surface: 62, M. sp. 21 9; 63, M. sp. 17 9; 64, M sp. 17 9 (apical half). 



128 



Journal of Hymenoptera Research 




Figs 65-77. 65-69, Mymaromella spp., clava: 65, M. mira 9, outer; 66, M. mira J; 67, M. sp. 18 9, outer; 68, M. sp. 
18 jj; 69, M. sp. 23 9, outer. 70 and 71, Zealaromma valentinei 9, clava: 70, oblique dorsal; 71, outer. 72, 
Mymaromella sp. 20 9, clava, inner ventral (insert: different structures of s3-type sensilla). 73, Z. insulare <$, clava. 
74-77, 9 funicular segment: 74, Maaminga rangi; 75, Chiloe micropteron; 76, Rotoita basalts; 77, Rotoita sp. nr R. 

basal is. 



Volume 16, Number 1, 2007 



129 




Figs 78-85. 78 and 79, Mymaromma spp., mesosoma, dorsal: 78, M. buyckxi; 79, M. sp. 6. 80 and 81, Mymaromma 
sp. 10: 80, mesosoma, lateral; 81, articulation of mesosoma and petiole, posterolateral. 82 and 83, Mymaromma 
spp.: 82, M. sp. 9, mesosoma, lateral; 83, M. goethei, posterior of mesosoma and petiole, lateral (insert: second 
petiolar segment). 84 and 85, Mymaromma sp. 7, metathoracic-propodeal complex: 84, posterolateral; 85, sensilla. 



130 



Journal of Hymenoptera Research 




Figs 86-94. 86 and 87, Mymaromma sp. 7: 86, mesosoma, ventrolateral; 87; posterior half of mesosoma, 
ventrolateral. 88 and 89, Mymaromma spp., petiole: 88, M. buyckxi, dorsal; 89, M. sp. 6, ventral. 90, Mymaromella 
sp. 21, mesosoma and petiole, posterolateral. 91 and 92, Mymaromella sp. 14: 91, mesosoma, dorsal; 92, propodeal 
apex and petiole, lateral. 93 and 94, Mymaromella minr. 93, mesosoma and petiole, dorsal; 94, propodeal apex and 
petiole, lateral. 



Volume 16, Number 1, 2007 



131 




Figs 95-102. 95 and 96, Mymaramella spp., mesosoma, lateral: 95, M. sp. 21; 96, M. sp. 15. 97 and 98, 
Mymaromella mini: 97, mesosoma, lateral; 98, metathoracic-propodeal sensilla. 99 and 100, Mymarotnella sp. 23: 
99, mesosoma, lateral; 100, metathoracic-propodeal sensilla. 101 and 102, Myniaivmella sp. 17: 101, mesosoma, 
lateral (insert: metapleural pit); 102; metathoracic-propodeal sensilla. 



132 



Journal of Hymenoptera Research 




Figs 103-110. 103-107, Zealaromma valentinei: 103, mesosoma, lateral; 104, posterior of metasoma, lateral; 105, 
metathoracic-propodeal complex, lateral; 106, metathoracic-propodeal complex, dorsolateral; 107, posterior of 
metasoma, dorsolateral. 108-110, Zealaromma insulate: 108, mesosoma, lateral; 109, metathoracic-propodeal 
complex, posterolateral; 110, metathoracic-propodeal complex, posterodorsal. 



Volume 16, Number 1, 2007 



133 




Figs 111-118. Ill, Zealaromma insulare, mesosoma, anterior (head removed). 112, Zealaromma valentinei, 
propleura, anteroventral. 113-116, Mymaromma spp., forewing: 113, M. goethei; 114, M. sp. 9; 115, M. buyckxi; 116, 
M. sp. 10. 117 and 118, Mymaromella spp., forewing: 117, M. sp. 14; 118, M. sp. 21. 



134 



Journal of Hymenoptera Research 




Figs 119-126. 119-121, Mymaromella spp., forewing: 119, M. sp. 18; 120, M. sp. 17; 121, M. sp. 20. 122, 
Zealaromma insulare, dorsal habitus. 123, Zealaromma valentinei, forewing disc reticulate pattern. 124, Mymaromma 
anomalum, proximal part of forewing. 125 and 126, Mymaromma sp. 10: 125, proximal part of forewing; 126, cross- 
section of forewing venation. 



Volume 16, Number 1, 2007 



135 




Figs 127-136. 127 and 128, Mymaromma anomalum: 127, proximal part of forewing stalk, dorsal; 128, apex ol 
forewing venation. 129 and 130, proximal part o\ forewing stalk, ventral: 129, Mymaromma goethei; 130, 
Mymaromella sp. 14. 131-133, M. anomalum: 131, distal part of forewing stalk, ventral; 132, proximal part ot 
forewing stalk, ventral; 133; wing coupling, dorsal. 134, Zealaromma valentinei, wing coupling. 135 and 136, hind 
wing: 135, Mymaromella sp. 23 (insert: apex, Mymaromella sp. 18); 136, Zealaromma valentinei. 



136 



Journal of Hymenoptera Research 




Figs 137-146. 137 and 138, strigil: 137, Mymaromella sp. 17, lateral; 138, Mymaromella sp. 21, ventrolateral. 139, 
Mymaromella sp. 14: a, strigil, lateral (insert: enlargement of calcar base between pseudospurs); b and c, apex of 
tibia, ventral: b, mesotibia; c, metatibia. 140 and 141, Zealaromma valentinei: 140, strigil, lateral; 141, apex of tibia, 
ventral: a, protibia, b, mesotibia, c, metatibia. 142, Mymaromma goethei, strigil, ventral. 143 and 144, Mymaromma 
anomalum: 143, strigil, lateral; 144: apex of tibia, ventral: a, protibia, b, mesotibia, c, metatibia. 145, Mymaromma 
sp. 10: a, strigil, lateral, b, apex of protibia, ventral (calcar foreshortened). 146, Z. valentinei, femur, posterior 
surface: a, profemur; b, mesofemur; c, metafemur. 



Volume 16, Number 1, 2007 



137 




Figs 147-154. 147, Mymaromella sp. 23 9, gaster. 148, Mymaromma sp. 6 9, hypopygium. 149 and 150, 
Mymaromella sp. 18 9: 149, apex of gaster; 150, apical two gastral tergites. 151 and 152, 9 apical two gastral 
tergites: 151, Mymaromella sp. 17; 152, Mymaromma sp. 10. 153 and 154, Zealaromma spp.: 153, Z. valentinei 9, apex 
of gaster; 154, Z. insulare J, penultimate gastral tergite and aedeagus (syntergum concealed under Mt 7 ). 



138 



Journal of Hymenoptera Research 




Figs 155-162. 155 and 156, Zealaromma valentinei S, apex of gaster (insert: paramere, arrows indicate direction): 
155, dorsolateral; 156, ventrolateral. 157, Maaminga rangi ], Mt, s and genitalia. 158-160, Mymaromma sp. 7 <$, apex 
of gaster and genitalia: 158, dorsolateral; 159, ventrolateral; 160, ventral. 161, Mymaromma sp. 6 j, parameres, 
ventral. 162, Mymaromella mini ], genitalia, dorsolateral. 



Volume 16, Number 1, 2007 



139 




® 



Figs 163-170. 163-166, forewing: 163, Mymaromma iimmuihini (Sweden); 164, Zcalmviiima insulare; 165, Z. 
valentinei; 166, Mymaromella cyclopterus (holotype 9). 167, Mytnaromella sp. 14, forewing stalk. 168, Mymarometta 
mira (USNM type photograph). 169 and 170, \ genitalia: 169, Z. insulare; 170, Al. anomalum (Japan). 



138 



Journal of Hymenoptera Research 




Figs 155-162. 155 and 156, Zealaromma valentinei ], apex of gaster (insert: paramere, arrows indicate direction): 
155, dorsolateral; 156, ventrolateral. 157, Maaminga rangi <$, Mt 8 and genitalia. 158-160, Mymaromma sp. 7 J, apex 
of gaster and genitalia: 158, dorsolateral; 159, ventrolateral; 160, ventral. 161, Mymaromma sp. 6 j, parameres, 
ventral. 162, Mymaromella mini ], genitalia, dorsolateral. 



Volume 16, Number 1, 2007 



139 




Figs 163-170. 163-166, forewing: 163, Mymaromma anomalum (Sweden); 164, Zealaromma insulare; 165, Z. 
valentinei; 166, Mi/uimvinclla ci/cloptcrus (holotype 9). 167, M\/iuiuvinclla sp. 14, forewing stalk. 168, Mymaromella 
mira (USNM type photograph). 169 and 170, J genitalia: 169, Z. insulare; 170, At anomalum (japan). 



140 



Journal of Hymenoptera Research 




Figs 171-182. 171 and 172, Zealaromma spp., antenna: 171, Z. insulare <$; 172, Z. valentinei 9. 173-176, 
Mymaromma spp., antenna: 173, M. anomalum 9; 174, M. goethei 9; 175, M. sp. nr goethei 9; 176, M. sp. nr goethei <$. 
177 and 178, Mymaromella sp. 14, antenna: 177, 9; 178, ,> 179, Palaeomymar succini 9, antenna. 180 and 181, 
Galloromma agapa (holotype 9): 180, left antenna; 181, apical three claval segments (a, right clava, dorsal view; b, 
left clava, outer surface). 182, Galloromma sp. 9, clava (GPPC: HY17A). (Arrows point to sensilla in Figs 181, 182.) 



Volume 16, Number 1, 2007 



141 




Figs 183-188. 183-188, Galloromma agapa (holotype 9): 183, habitus (insert: calcar); 184, petiolar segments; 185, 
proximal part of forewings; 186, head, dorsolateral; 187, mesosoma, lateral; 188, forewing posterobasal 
marginal setae. 



142 



Journal of Hymenoptera Research 




Figs 189-196. 189-191, Galloromma sp. (AMNH: B-0107): 189, 9 head and antenna, dorsal; 190, 9 body except 
gaster, lateral; 191, <$ habitus (insert: calcar). 192, Archaeromma minuHssimum, antenna (holotype 9). 193, 
Galloromma sp. 9, mandibles, ventral (GPPC: HY17A). 194, Archaeromma nearcticum 9, forewing posterobasal 
marginal setae (CNC: CAS-598). 195, Archaeromma sp. $, forewing (AMNH: NJ-179). 196, Archaeromma masneri, 
head (dorsal) and antenna (holotype 9). 



Volume 16, Number 1, 2007 



143 




Figs 197-205. 197-202, Archaeromma mandibulatum (holotype 9): 197, habitus; 198, head, dorsolateral; 199, 
antenna; 200, clava; 201, proximal part of forewing stalk; 202, {brewing posterobasal marginal setae. 203-205, 
Archaeromma senonicum (paratype 9): 203, head, ventrolateral; 204, antenna; 205, forewing. 



144 



Journal of Hymenoptera Research 




Figs 206-212. 206, Archaeromma masneri (holotype 9). 207-210, Palaeomymar succini: 207, habitus (neotype c?); 
208, forewing posterobasal marginal setae (neotype $); 209, 9 forewing posterobasal marginal setae (ZMUC: 16- 
1/1961); 210, foreleg (neotype ?). 211 and 212, ? Mymaromella sp. 9 (ZMUC: 17-5/1963): 211, forewings; 
212, calcar. 



Volume 16, Number 1, 2007 



145 




Figs 213-219. 213, Galloromma sp. 9, head, dorsal (GPPC: HY17A). 214-217, Galloromma bezonnaisensis 
(holotype): 214, antenna, dorsal; 215, lateral habitus reproduced from Schliiter (1978); 216, forewing posterobasal 
marginal setae and antennal funicle; 217, head, dorsal. 218 and 219, head and antennae of Serphites sp.: 218, 
lateral (MCZ: 5330); 219, frontal (PIN: 3730/28-30). 



146 



Journal of Hymenoptera Research 




Figs 220-227. 220-222, Serphites sp. (PIN: 3730/31): 220, mesosoma, lateral; 221, pronorum-mesopectus; 222, 
mesosoma, ventrolateral. 223, Serphites sp. $ (MCZ: 5343), apex of gaster. 224 and 225, Serphites sp., gaster: 224, 
ventrolateral (PIN: 3730/28-30); 225, lateral (MCZ: 5330). 226, Serphites s P ., forewing venation (MCZ: 5530). 227, 
forewing of undescribed representative of Scelionidae (PIN: 4703/10). 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 147-166 

Szepligeti's Cyclaulax types Deposited in the Hungarian Natural 
History Museum (Hymenoptera: Braconidae: Braconinae) 

J. Papp 

Department of Zoology, Hungarian Natural History Museum, H-1431 Budapest, Pf. 137, Hungary; 

tel 36-1-267-7101, 267-7007; fax 36-1-3171-669 



Abstract. — The nine Cyclaulax species represented by types and deposited in the Museum 
Budapest as well as the type-species of the genus Cyclaulax grandiceps Cameron, housed in the 
Natural History Museum, London, are re-described and an identification key to the ten species is 
presented. Currently 11 Cyclaulax species are listed from the Neotropical Region. With 94 original 
figures. 



The genus Cyclaulax was created by 
Cameron (1911) on the basis of the new 
species, C. grandiceps Cameron from Gui- 
ana (former British Guiana). The first 
Cyclaulax species, however, were discov- 
ered by Szepligeti, who described a total of 
nine species originally ranged in the genus 
Braccrn Fabricius (Szepligeti 1902, 1904, 
1906). Cyclaulax belongs to the Compsobra- 
con Ashmead group of genera (Quicke 
1997, Leathers et al. 2005), its generic 
characters are listed below. 

Quicke (1991) reviewed the non-Europe- 
an species of the subfamily Braconinae 
deposited in the Hungarian Natural Histo- 
ry Museum ( = Magyar Termeszettudoma- 
nyi Muzeum), Budapest and he was the 
first who recognized the true taxonomic 
position of these nine Bracon species 
(enumerated below). 

In the present paper the nine Cyclau- 
lax species by Szepligeti as well as the 
generic type-species C. grandiceps are re- 
described and an identification key com- 
piled for the ten species to promote their 
reliable recognition. Quicke (1997: 152) 
indicated that Cyclaulax specimens are 
"frequently collected" in the Neotropical 
Region. The number of the Cyclaulax 
species, consequently, will considerably 
increase as a result of future taxonomic 
research. 



Our knowledge of the Braconinae wasps 
of the Neotropical Region is less advanced. 
In this respect it is worthwhile to quote 
Quicke's (1997) assertions: "Comparatively 
few studies have been made on the New 
World fauna since the first two decades of 
this century." and "Undoubtedly, difficul- 
ties in identifying New World braconines 
in general and Neotropical species in 
particular have hampered and even dis- 
couraged work on their biology and hence 
possible use in biological or integrated 
control programs." The present contribu- 
tion is a slight step towards the promotion 
of our improved knowledge of the Neo- 
tropical braconines. 

METHODS 

Abbreviations. — The following abbrevia- 
tions are applied in the re-descriptions of 
the species and in the identification key to 
Cyclaulax species (after van Achterberg 
1993: 4-5): 

Eye: OOL = ocellar-ocular line, i.e. 
shortest distance between hind ocellus 
and compound eye; POL = postocellar 
line, i.e. shortest distance between hind 
two ocelli. 

Fore wing: m—cu = recurrent vein; r = 
first section of the radial vein; 1-M = basal 
vein; 2-CUl = second section of the discal 
vein; 2-SR = first transverse cubital vein; 



148 



Journal of Hymenoptera Research 



3-SR = second section of the radial vein; 
SRI = third section of the radial vein; 1- 
SR+M = first section of the cubital vein. 
Hind wing: cu-a = nervellus. 

TAXONOMY 

Genus Cyclaulax Cameron 

Cyclaulax Cameron, 1911: 7, type species: Cy- 
claulax grandiceps Cameron, 1911. - Quicke 
1989: 119 (taxonomy), 1997: 152 (in key). 

Generic features of Cyclaulax. — (1) Face 
sculptured as in Figs 30-33 (Quicke 1997: 
163), otherwise head and mesosoma po- 
lished. (2) Every tergite polished, second 
tergite antero-medially without an area 
bordered by furrow or not "pinched up"; 
third tergite usually clearly longer than 
second tergite (Figs 7, 23, 49, 75). (3) 
Basitarsus of fore leg laterally compressed 
or flattened (Figs 42, 63, 72). (4) Suture 
between tergites 2-3 trisinuate, median 
sinuation deeply curved (Figs 7, 23, 49, 
75). (5) Venation of forewing as in Fig. 6 
(Quicke 1997: 159), vein 1-SR+M distinctly 
bent. (6) Eye somewhat protruding (Figs 1, 
19, 38, 89). (7) Hypopygium pointed, 
ovipositor sheath long (Fig. 50, 76). 

With the help of Quicke's key (1997) to 
the Nearctic-Neotropic genera of the sub- 
family Braconinae it is fairly easy to 
identify the genus Cyclaulax. The nearest 
genera to Cyclaulax are Gracilibracon 
Quicke, 1995 and Cyclaulacidea Quicke 
and Delobel, 1995. 

Cyclaulax species are distributed in the 
Neotropical Region except Chile. In the 
world catalogue of Braconidae (Shenefelt 
1978: 1682) only one, the type species, is 
listed for this genus. Szepligeti has de- 
scribed nine species originally assigned to 
the genus Bracon (Szepligeti 1902, 1904, 
1906). They were transferred to Cyclaulax 
by Quicke (1991) except B.flexuosus and B. 
linurus with the taxonomic note "Belongs 
to an undescribed genus of the Campsobra- 
conoides (=Compsobracori) group." I do not 
concur with this opinion, i.e. I consider 



these two species as representing also the 
genus Cyclaulax. 

The following nine Cyclaulax species by 
Szepligeti (represented by type material or 
named specimens) are housed in the 
Hungarian Natural History Museum: Cy- 
claulax atriceps (Szepligeti, 1904), C. binota- 
tus (Szepligeti, 1904), C. enotatus (Szepli- 
geti, 1904), C.flexuosus (Szepligeti, 1902), C. 
linurus (Szepligeti, 1906), C. lunatus (Sze- 
pligeti, 1906), C. mesonurus (Szepligeti, 
1906), C. paraguayensis (Szepligeti, 1904) 
and C. sicuaniensis (Szepligeti, 1904). 

By Mrs S. Lewis-Ryder's (London) kind 
assistance I had the opportunity to study 
the type species (C. grandiceps Cameron) of 
the genus Cyclaulax Cameron. In the 
taxonomic part of the present article its 
redescription is given. 

An eleventh species, C. crassitarsis 
(Brues, 1912), is listed in the checklist (see 
below). 

Cyclaulax atriceps (Szepligeti) 
(Figs 1-8) 

Bracon atriceps Szepligeti, 1904: 184 9, type 
locality: "Peru: Marcapata", female holotype 
(designated by Papp in 1969) in Magyar 
Termeszettudomanyi Miizeum, Budapest; 
examined. - Szepligeti 1906: 591 (in key). 
Shenefelt 1978: 1467 (as Bracon atriceps, 
literature up to 1906). Quicke 1991: 171 (as 
Cyclaulax atriceps comb, n., type depository). 

Designation of the female holotype of 
Bracon atriceps. — (first label, printed) "Mar- 
capata / Peru"; second label is the holo- 
type card, third label is with the inventory 
number 1549 (second and third labels were 
attached by me); fourth label is with the 
actual name C. atriceps (Szepligeti) given by 
Quicke in 1989. 

Redescription of the female holotype of 
Bracon atriceps. — Body 8 mm long. Antenna 
somewhat shorter than body and with 41 
antennomeres (left antenna; right flagellum 
deficient, i.e. with 21 flagellomeres). Scape 
in outer-lateral view twice as long dorsally 
as broad apically (cf. Fig. 60). - Head in 



Volume 16, Number 1, 2007 



1 49 




Figs 1-8. Cyclaulax atriceps (Szepligeti), female lectotype: 1 = head in dorsal view, 2 = head in lateral view, 3 = 
hind femur, 4 = tarsomeres 3-4 of fore leg, 5 = claw, 6 = first discal cell of fore wing, 7 = tergites 1-3; female: 8 
= hind femur 



dorsal view (Fig. 1) less transverse, 1.7 
times as broad as long, eye 1.5 times longer 
than temple, temple faintly receded. Eye in 
lateral view clearly 1.5 times as high as 
wide and 1.6 times wider than temple, 
latter evenly beyond eye (Fig. 2). 

Mesosoma in lateral view 1.6 times as 
long as high. Hind femur 2.9 times as long 
as broad distally (Fig. 3). Fourth tarsomere 
of fore leg in lateral view 1.35 times as long 
as high (Fig. 4). Claw of hind leg as in 
Fig. 5. First discal cell high, 1-M 1.35 times 
as long as m-cu, 1-M bent, veins relatively 
thick (Fig. 6). - First tergite 1.4 times as 
long as broad behind, its scutum rather 
narrow; third tergite 1.6 times as long 
medially as second tergite laterally, suture 
between them less strongly trisinuate 
(Fig. 7, see arrows). Ovipositor sheath as 
long as hind tibia + tarsus combined. 

Head and mesosoma black, mesoscutum 
and scutellum anteriorly reddish. Tergites 
1-2 reddish, further tergi blackish to black. 
Legs black. Wings brown famous, veins 
brown, 1-SR+M yellowish. 

Deviating features of two females (from 
the locality "Peru, Pachitea"). Body 
9 mm long. Both flagelli deficient. Head 
in dorsal view 1.6 times as broad as long (1 
9). Hind femur 3.1 times as long as broad 
distally (Fig. 8). 



Deviating features of one female (from 
the locality "Peru, Pachitea"). - Body 
9 mm long. Both flagelli deficient. A light 
coloured (or albanic) specimen: mesoscu- 
tum, scutellum, tegula, upper fourth of 
mesopleuron and lateral corner of prono- 
tum reddish yellow; fore leg: coxa, tro- 
chanters and tarsus yellowish, femur + 
tibia + fifth tarsomere dark brown. 

Male and host unknown. 

Distribution. — Peru. 

Taxonomic position. — C. atriceps is closest 
to C. enotatus, their specific distinction is 
presented in the key-couplets 19(20) - 
20(19). 

Cyclaulax binotatus (Szepligeti) 
(Figs 9-17) 

Bracon binotatus Szepligeti, 1904: 184 (descrip- 
tion) and 186 (in key) 9, type locality: "Peru: 
Marcapata", female lectotype (+ six female 
paralectotypes designated by Papp in 1969, 
see Shenefelt I.e.) in Magyar Termeszettudo- 
manyi Muzeum, Budapest; examined. - 
Szepligeti 1906: 590 (in key). Shenefelt 1978: 
1470 (as Bracon binotatus, literature up to 
1906). Quicke 1991: 171 (as Cyclaulax binota- 
tus comb, n., type depository). 

Type designation of Bracon binotatus — 

Designation of the female lectotype: (first 



150 



Journal of Hymenoptera Research 




Figs 9-17. Cyclaulax binotatus (Szepligeti), female lectotype: 9 = scape in outer-lateral view, 10 = head in 
dorsal view, 11 = head in lateral view, 12 = hind femur, 13 = first discal cell of fore wing, 14 = tergites 1-3; 
female paralectotype: 15 = hind femur, 16 = vein 1-SR-M of fore wing, 17 = tergites 2-3 



label, printed) "Marcapata / Peru"; Second 
label is the lectotype card, third label is 
with the inventory number 1542 (second 
and third labels were attached by me); 
fourth label is with the actual name C. 
binotatus given by Quicke in 1989. - 
Lectotype is in good condition: (1) micro- 
pinned; (2) left flagellum distally deficient, 
i.e. with 22 flagellomeres; (3) fore right 
wing basally torn. 

Designation of the six female paralecto- 
types. — (first label, printed) "Marcapata / 
Peru"; second label is the lectotype card, 
third label is with the inventory numbers 
1543-1548 (second and third labels were 
attached by me); fourth label is with the 
actual name C. binotatus given by Quicke in 
1989. - Paralectotypes are in fairly good 
condition: micropinned, flagelli partly or 
entirely deficient. 

Taxonomic remark. — Quicke (1991: 171) 
correctly indicated that the "specimens 
1545-1548 inclusive appear to belong to 
a different species from specimens 1542- 
1544". - The lectotype (no. 1542) and one 
female paralectotype (no. 1544) are repre- 
senting the nominate form Bracon binotatus 
(albeit the paralectotype is "var. 9" by 
Szepligeti I.e.). - One female paralectotype 



(no. 1543) is near to C. sieuaniensis (Szepli- 
geti) and supposedly representing a new 
Cyclaulax species (specimen in question is 
in poor condition, inappropriate for type 
designation). - Two female paralectotypes 
(nos 1545, 1547) are "var. 2. 9" by 
Szepligeti (I.e.) and received the new name 
C. lunatus (Szepligeti) by me. - One female 
paralectotype (no. 1546) is in the Nationaal 
Natuurhistorisch Museum, Leiden and one 
female paralectotype (no. 1548) is in the 
Zoological Institute, Saint Petersburg as 
exchange material. The female paralecto- 
type in Leiden does not represent the 
species C. binotatus (Szepligeti) and is near 
to C. lunatus (Szepligeti), however, de- 
viating from it by the following features: 
first tergum less broad, head rather sub- 
cubic, fore leg entirely and tarsus of middle 
leg yellow; perhaps it will prove to be 
a new Cyclaulax species (both its flagelli 
missing), further specimens are needed to 
detect its true taxonomic status. 

The true Cyclaulax binotatus (Szepligeti), 
represented by the female lectotype, one 
female paralectotype and one female (with- 
out type status), deviates from all other 
Cyclaulax species of Szepligeti in that its 
scape is not emarginated apically neither in 



Volume 16, Number 1, 2007 151 

its outer- (Fig. 9) nor in its inner-lateral broad somewhat distally (Fig. 15). Vein 1- 

view. This feature combined with the very SR+M of first discal cell broken angularly 

wide scutum of first tergite (Fig. 14) may (Fig. 16). First tergite 1.6 times as long as 

serve in the future either for subgeneric broad behind. Third tergite one-fifth longer 

separation within the genus Cyclaulax or to than second tergite (Fig. 17). Face below 

create a new genus. Again, more material antennal socket entirely black (i.e. without 

is needed to decide this taxonomic prob- yellow macula, female paralectotype) or 

lem. face almost entirely reddish yellow (one 

Redescription of the female lectotype of female). 

Bracon binotatus. — Body 7.2 mm long. An- Male and host unknown, 

tenna about as long as body and with 43 Distribution. — Peru. 

antennomeres (left flagellum deficient). Taxonomic position. — C. binotatus stands 

Scape in outer-lateral view 1.6 times as alone with its clearly transverse head and 

long dorsally as broad apically (Fig. 9). - constricted temple, see key-couplets 7(8) - 

Head in dorsal view (Fig. 10) transverse, 8(7). 
almost 1.9 times as broad as long, temple 

constricted, eye clearly twice as long as Cyclaulax enotatus (Szepligeti) 

temple. Eye in lateral view 1.5 times as ' 1 & S l ° - ^v 

high as wide and clearly two times wider Bracon enotatus Szepligeti, 1904: 184 9, type 

than temple (Fig. 11). locality: "Peru: Marcapata", female Lectotype 

Mesosoma in lateral view 1.5 times as (and three female paralectotypes designated 

long as high. Hind femur 3.6 times as long by Papp in 1969, see Shenefelt I.e.) in Magyar 

as broad distally (Fig. 12). First discal cell Termeszettudomanyi Miizeum, Budapest; 

less high, 1-M 1.3 times as long as m-cu, 1- examined. ■■ Szepligeti 1906: 591 (in key). 

M straight, 1-SR+M broken in its run Shenefelt 1978: 1481 (as Bracon enotatus, 

m- ia\ c- i. * •* 1 cc ..• i literature up to 1906). Quicke 1991: 171 (as 

(rig. 13). - hirst tergite 1.55 times as lone „ , , r , . 

, , , , • , . , , Cyclaulax enotatus comb, n., tvpe deposito- 

as broad behind, its lateral part narrow, \ 

suture between tergites 2-3 weakly trisi- 

nuate; third tergite somewhat more than Type designation of Bracon enotatus. — Des- 
one-fifth longer medially than second ignation of the female lectotype and two 
tergite laterally (Fig. 14, see arrows). Ovi- female paralectotypes (as "var. 9" in 
positor sheath as long as hind tibia + tarsus Szepligeti I.e.): (first label, printed) "Mar- 
combined, capata / Peru"; second label is the lecto- 

Scape black, flagellum blackish. Head type and paralectotype cards, third label is 

yellow, frons, vertex and face medially with the inventory number 1551 (lectotype) 

black, face below antennal socket pale and 1552-1553 (paralectotypes) (second 

yellow. Palpi yellow. Mesosoma reddish and third labels attached by me); fourth 

yellow, propodeum + metapleuron black- label is with the actual name C. enotatus 

ish brown. Tergites black; sternites ochre- given by Quicke in 1989. - Lectotype is in 

ous, medio-longitudinally with a black rather poor condition: (1) pinned by meso- 

streak. Fore leg yellow with faint brownish soma; (2) both flagelli missing; (3) right 

tint, middle and hind legs blackish brown, hind leg (except coxa + trochanters) glued 

Wings brownish fumous, pterostigma and on a separate card, tarsomeres 2-5 of right 

veins dark brown to brown. middle leg missing; (4) right fore wing also 

Deviating features of one female para- glued on a separate card, 

lectotype and one female. - Similar to the Designation of one female paralectotype.— 

female lectotype. Body 6.5-7 mm long, (first label, printed) "Peru / Chanchal- 

Antenna with 46 antennomeres (paralecto- majo"; second label is the paralectotype 

type). Hind femur four times as long as card, third label is with the inventory 



152 



Journal of Hymenoptera Research 




Figs 18-25. Cyclaulax enotatus (Szepligeti), female lectotype: 18 = scape in outer-lateral view, 19 = head in 
dorsal view, 20 = head in lateral view, 21 = hind femur, 22 = first discal cell of fore wing, 23 = tergites 1-3; 
female paralectotype: 24 = temple in dorsal view, 25 = hind femur 



number 1552 (both labels attached by me); 
fourth label is with the actual name C. 
enotatus (Szepligeti) given by me. - The 
three paralectotypes are also in poor 
condition: (1) pinned by mesosoma; (2) 
flagelli either missing or deficient; (3) legs 
partly missing. 

Taxonomic remark. — The two female 
paralectotype ("var. 9" by Szepligeti, from 
Peru: Marcapata) do not represent C. 
enotatus, they belong to C. lunatus (Szepli- 
geti) and I labelled them accordingly. One 
female paralectotype is in Museum Buda- 
pest and one female paralectotype is in 
Museum Leiden (as exchange material). - 
The third female paralectotype, from Peru: 
Chanchalmajo, is a true C. enotatus and is in 
the Museum Budapest. 

Redescription of the female lectotype of 
Bracon enotatus. — Body 8 mm long. Scape 
in outer-lateral view 1.7 times as long 
dorsally as broad apically (Fig. 18). Both 
flagelli missing. - Head in dorsal view 
(Fig. 19) transverse, 1.6 times as broad as 
long, temple narrowing, eye almost 1.5 
times (or clearly one-third) longer than 
temple. Eye in lateral view 1.5 times as 
high as wide and 1.3 times wider than 
temple, temple beyond eye evenly broad 
(Fig. 20). 



Mesosoma in lateral view 1.75 times as 
long as high. Hind femur 3.6 times as long 
as broad medially (Fig. 21). First discal cell 
less high, 1-M nearly 1.6 times as long as 
m-eu, 1-M straight (Fig. 22). - First tergite 
1.2 times as long as broad behind, scutum 
of tergite wide; second tergite less short, 
third tergite nearly 1.4 times as long 
medially as second tergite laterally 
(Fig. 23, see arrows). Suture between ter- 
gites 2-3 trisinuate (Fig. 23). Ovipositor 
sheath nearly as long as hind tibia + tarsus 
combined. 

Scape, pedicel, head, legs and tergi 
black. Palpi dark brown, its ultimate joint 
light brown. Mesosoma testaceous; propo- 
deum, metapleuron and mesosternum 
black. Tegula also testaceous. Fore coxa 
yellowish. Wings brownish fumous, pter- 
ostigma and veins brown. 

Deviating features of one female paralecto- 
type of Bracon enotatus (locality: Peru, 
Chanchalmajo). — Similar to the female lec- 
totype. Body 9 mm long. Both flagelli 
deficient, right flagellum with 28 and left 
flagellum with 7 flagellomeres. First flagel- 
lomere 1.4 times and 28th flagellomere 
cubic, i.e. as long as broad. Temple in 
dorsal view somewhat more narrowing 
(Fig. 24). Hind femur 3.3 times as long as 



Volume 16, Number 1, 2007 



153 




Figs 26-35. Cyclaulax flexuosus (Szepligeti), female lectotype: 26 = scape in outer-lateral view, 27 == head in 
dorsal view, 28 = head in lateral view, 29 = lump of left temple in lateral view, 30 == hind femur, 31 = 
pterostigma and vein r of fore wing, 32 = first discal cell of fore wing, 33 = tergites 1-3; male paralectotype: 34 = 
vein 1-SR-M of fore wing, 35 = first tergite 



broad medially (Fig. 25). Cheek yellow. 
Mesosternum testaceous. 

Male and host unknown. 

Distribution. — Peru. 

Taxonomic position. — C. enotatus is nearest 
to C. ntriceps (Szepligeti), the distinction is 
presented in key-couplets 19(20)-20(19). 

Cycaulax flexuosus (Szepligeti) 
(Figs 26-35) 

Bracon flexuosus Szepligeti, 1902: 42 <$, type 
locality: "Venezuela: Merida", male lectotype 
(and one male paralectotype, designated by 
Papp in 1969) in Magyar Termeszettudoma- 
nyi Muzeum, Budapest; examined. - Szepli- 
geti 1904: 187 (in key), 1906: 591 (in key). 
Shenefelt 1978: 1485 (as Bracon flexuosus, 
literature up to 1906). Quicke 1991: 172 (type 
depository). 

Type designation of Bracon flexuosus — 
Designation of the male lectotype: (first 
label) "Merida" (printed) / "Venezuela" 
(L. Biro's handscript); (second label) "Br. 
flexuosus" (Szepligeti's handscript) / "det. 
Szepligeti" (printed); third label is my 
lectotype card, fourth label is with the 
inventory number 1560 (third and fourth 



labels were attached by me); fifth label is 
with the actual name C. flexuosus given by 
me. - Lectotype is in good condition: (1) 
pinned by mesosoma; (2) left flagellum 
missing, right flagellum deficient. 

Designation of the male paralectotype: 
(first label) "Merida" (printed) / "Vene- 
zuela" (L. Biro's handscript); second label 
is with the paralectotype card, third label is 
with the inventory number 1561 (second 
and third labels were attached by me); 
fourth label is with the actual name C. 
flexuosus given by me. - Paralectotype is in 
fairly good condition: (1) pinned by meso- 
soma; (2) both flagelli distally deficient; (3) 
left pair of wings missing. 

Redescription of the mule lectotype of 
Bracon flexuosus. — Body 7.5 mm long. Right 
flagellum with 33 flagellomeres (left flagel- 
lum missing). Scape in outer-lateral view 
almost 1.6 times as long dorsally as broad 
apically, deeply emargined, ventrally as 
long as dorsally (Fig. 26). First flagellomere 
twice and 33rd flagellomere subcubic, i.e. 
a bit longer than broad. - Head in dorsal 
view (Fig. 27) subcubic, 1.5 times as broad 
as long, eye 1.66 times length of temple, 
temple moderately rounded. Eye in lateral 



154 



Journal of Hymenoptera Research 



view almost 1.5 times as high as wide and 
1.6 times wider than temple, latter ventral- 
ly narrowing (Fig. 28, see arrows). Left 
temple ventrally with a (teratological?) 
small lamp (Fig. 29). 

Mesosoma in lateral view almost twice 
as long as high. Hind femur less broaden- 
ing distally, 3.3 times as long as broad 
medially (Fig. 30). Vein r of fore wing 
clearly longer than half width of pteros- 
tigma (Fig. 31). First discal cell high, 1-M 
weakly bent and 1.7 times as long as in-cu 
(Fig. 32). - First tergite 1.4 times as long as 
broad behind, its scutum more narrowing 
anteriorly; third tergite 1.7 times longer 
medially than second tergite laterally 
(Fig. 33, see arrows), suture between ter- 
gites 2-3 weakly trisinuate (Fig. 33). 

Scape and pedicel rusty brown, flagel- 
lum dark brown. Head, mesosoma and 
legs rusty brown; mesosoma testaceous, 
apically blackish. Palpi light brown to 
brown. Tegula rusty brown. Fore tarsus 
yellow. Wings brown fumous, pterostigma 
brown, vein brown to light brown. 

Redescription of the male paralectotype of 
Bracon flexuosus. — Similar to the male lec- 
totype. Body 6 mm long. Both flagelli 
distally deficient: right flagellum with 17 
and left flagellum with 14 flagellomeres. 
Vein 1-SR-M of first discal cell slightly less 
bent (Fig. 34). First tergite clearly 1.3 times 
as long as broad behind, its scutum 
somewhat narrowing (Fig. 35). 

Female and host unknown. 

Distribution. — Venezuela. 

Taxonomic position. — C. flexuosus is near- 
est to C. paraguayensis (Szepligeti) and to C. 
sicuaninensis (Szepligeti), their distinction is 
presented in the key-couplets 14(15) - 
17(16). 

Cyclaulax grandiceps Cameron 
(Figs 36-50) 

Cyclaulax grandiceps Cameron, 1911: 7 9 (syntype 
series one female), type locality: British 
Guiana, female holotype (="Type") in The 
Natural History Museum, London; exam- 
ined. - Shenefelt 1978: 1682 (as Cyclaualx 
grandiceps, literature up to 1911). 



Designation of the female holotype 
(="Type") of Cyclaulax grandiceps. — (first 
round label with red frame) "Type" 
(printed); (second label) "B. M. Type 
Hym." (printed) "3.C.152" (handscript); 
(third great label with Cameron's hand- 
script) "Cyclaulax / grandiceps / Cam. 
Type / Br. Guyana"; (fourth label, printed) 
"P. Cameron Coll. 1914-110." - The holo- 
type is in fairly poor condition: (1) pinned 
by mesosoma; (2) left flagellum missing, 
right flagellum deficient; (3) left fore wing 
missing, right fore wing glued separately 
on a small card; (4) right fore leg sticked to 
right mesopleuron; (5) missing: tarsi of left 
fore and left hind legs, fifth tarsomere of 
right fore tarsus and tarsomeres 2-5 of 
right middle tarsus. 

Redescription of the female holotype of 
Cyclaulax grandiceps. — Body 12 mm long. 
Scape cylindrical, in outer-lateral view 
twice as long dorsally (Fig. 36, see hori- 
zontal arrow) as broad apically, ventrally 
and dorsally of equal length, its outer side 
deeply emargined (Fig. 37), its inner side 
with an apico-median ledge (Fig. 36, see 
vertical arrow); pedicel short. Right flagel- 
lum deficient, i.e. with 21 flagellomeres. 
First flagellomere 1.5 times, second flagel- 
lomere 1.2 times as long as broad apically, 
further flagellomeres cubic (Fig. 37). - 
Head in dorsal view (Fig. 38) less trans- 
verse, 1.6 times as broad as long, eye 
somewhat protruding and almost 1.4 times 
length of temple, temple moderately 
rounded, occiput excavated. Ocelli near to 
each other, OOL three times as long as 
POL. Eye in lateral view 1.5 times as high 
as wide, one-third wider than temple 
medially, temple ventrally narrowing 
(Fig. 39, see arrows). Face with similar 
sculpture to that of Fig. 32 (cf. Quicke 
1997: 163); otherwise head polished. Fourth 
maxillar palpal joint somewhat thicker and 
shorter than fifth joint (Fig. 40). 

Mesosoma in lateral view almost twice 
as long as high, polished. Notaulix distinct 
weakly. Propodeum polished. - Hind 
femur three times as long as broad distally 



Volume 16, Number 1, 2007 



L55 




Figs 36-47. Cydaulax grandiceps Cameron, holotype: 36 = scape in outer-lateral view, 37 = scape in inner- 
lateral view and flagellomeres 1-2 + 21, 38 = head in dorsal view, 39 = head in lateral view, 40 = maxillar palpal 
joints 3-5, 41 = hind femur, 42 = basitarsus + spur of fore leg, 43 = basitarsus + spur of middle leg, 44 = 
basitarsus, spur and second flagellomere of hind leg, 45 = claw, 46 = first discal cell of fore wing, 47 = subbasal 
cell with vein cu-a of hind wing 



(Fig. 41). Inner spur of middle tibia slightly Fore wing as long as body. Pterostigma 

longer than (Fig. 43) and that of hind tibia (Fig. 48) 3.6 times as long as wide and 

less than half as long as basitarsus, latter issuing r proximally from its middle, r as 

somewhat thick (Fig. 44). Claw curved, long as width of pterostigma. Second 



basally widening (Fig. 45). 



submarginal cell long, 3-SR almost twice 






Figs 48-50. Cydaulax grandiceps Cameron, holotype: 48 = pterostigma and first submarginal cell of tore wing, 
49 = tergites 1-3, 50 = posterior end of metasoma with hypopygium and ovipositor apparatus 



156 



Journal of Hymenoptera Research 




Figs 51-59. Cyclaulax linurus (Szepligeti), female lectotype: 51 = scape in outer-lateral view, 52 = head in 
dorsal view, 53 = head in lateral view, 54 = hind femur, 55 = tarsomeres 2-4 of fore leg, 56 = first discal cell of 
fore wing, 57 = tergites 1-3; female paralectotype: 58 = head in dorsal view, 59 = tergites 2-3 



as long as 2-SR, SRI straight, somewhat 
longer than 3-SR and reaching tip of wing. 
First discal cell less high, 1-M 1.5 times 
length of m-cu, 1-SR+M clearly bent 
(Fig. 46). - Hind wing: cu-n as in Fig. 47 
(see arrow). 

First tergite (Fig. 49) somewhat broader 
behind than long, scutum convex, lateral 
part of tergite fairly wide. Second tergite 
transverse, 3.4 times as broad behind as 
long laterally; third tergite 1.8 times longer 
medially than second tergite laterally 
(Fig. 49, see arrows); suture between them 
deep, smooth, trisinuate, median sinuation 
the deepest (Fig 9). Every tergite polished. 
Hypopygium pointed, ovipositor sheath 
long, as long as hind femur + tibia + 
tarsomeres 1-2 combined (Fig. 50). 

Ground colour of body reddish yellow. 
Head and antenna black, flagellum with 
very weak brownish suffusion. Labrum 
and cheek ferruginous, palpi blackish 
brown to brown. Pronotum and -sternum 
black to blackish. Last two metasomal 
segments black. Legs black, fore coxa + 
trochanters reddish yellow, tarsomeres of 
middle leg apically rusty. Wings dark 
brown fumous, pterostigma blackish, veins 
proximo-distal ly black to brown. 



Male and host unknown. 

Distribution. — Guiana 

Taxonomic position. — C. grandiceps Ca- 
meron differs from all other Cyclaulax 
species by its very broad first tergite; C. 
lunatus (Szepligeti) and C. paraguayensis 
(Szepligeti) appear to be the nearest to C. 
grandiceps with their relatively broad first 
tergites; see also the key-couplets 1(6) - 
5(4). 

Cyclaulax linurus (Szepligeti) comb. n. 
(Figs 51-59) 

Bracon linurus Szepligeti, 1906: 591 (in key) and 
593 (description) 9, type locality: "Bolivia: 
Mapiri", female lectotype (+ one female para- 
lectotype, designated by Papp in 1969) in 
Magyar Termeszettudomanyi Muzeum, Buda- 
pest; examined. - Shenefelt 1978: 1503 (as 
Bracon linurus, literature up to 1906). Quicke 
1991: 172 (taxonomy, type depository). 

Type designation of Bracon linurus. — Des- 
ignation of the female lectotype of: (first 
label, printed) "Bolivia / Mapiri"; second 
label is the lectotype card, third label is 
with the inventory number 993 (second 
and third labels were attached by me); 
fourth label is with the actual name C. 



Volume 16, Number 1, 2007 



157 



linurus (Szepligeti) given by me. - Lecto- 
type is in fairly poor condition: (1) pinned 
by mesosoma; (2) both flagelli deficient 
distally; (3) right fore wing antero-medially 
damaged, distal third part of left fore wing 
glued on a separate small card; (4) left 
ovipositor sheath broken (present its short 
basal part); (5) head broken, glued to 
prosoma. 

Designation of the female paralectotype.— 
(first label, printed) "Peru / Mercapata"; 
second label is the paralectotype card, 
third label is with the inventory number 
994 (second and third labels were attached 
by me); fourth label is with the actual name 
C. linurus (Szepligeti) given by me. - 
Paralectotype is in fair condition: (1) 
micropinned by mesosoma; (2) both flagelli 
deficient distally; (3) right hind wing 
creased. 

Redescription of the female lectotype of 
Bracon linurus. — Body 8 mm long. Scape 
in outer-lateral view 1.7 times as long 
dorsally as broad apically (Fig. 51). Both 
flagelli deficient, right flagellum with 27 
and left flagellum with 18 flagellomeres. 
First flagellomere twice and 27th flagello- 
mere subcubic, i.e. a bit longer than broad. 
- Head in dorsal view (Fig. 52) less trans- 
verse, clearly 1.5 times as broad as long, 
eye 1.4 times longer than temple, temple 
clearly rounded. Eye in lateral view 1.3 
times as high as wide and twice wider than 
temple, latter beyond eye evenly broad 
(Fig. 53, see arrows). 

Mesosoma in lateral view 1.6 times as 
long as high. Hind femur 3.6 times as long 
as broad distally (Fig. 54). Tarsomeres 2-4 
of fore leg in lateral view short, third 
tarsomere 1.6 times as long as high and 
fourth tarsomere cubic, a bit longer than 
high distally (Fig. 55). First discal cell less 
high, 1-M 1.5 times as long as m-cu, 1-M 
just bent (Fig. 56). - First tergite clearly 1 .3 
times as long as broad behind, its scutum 
fairly wide, i.e. lateral margin of tergite less 
wide; third tergite 1.5 times as long 
medially as long second tergite laterally; 
suture between them medially deep 



(Fig. 57, see arrows). Ovipositor sheath 
longer than body. 

Ground colour of body black; legs also 
black, fore tarsus yellow, middle and hind 
tarsi dark brownish black. Scape and 
flagellum black. Mesoscutum, scutellum 
and tergites 1-2 testaceous. Wings brown 
fumous, pterostigma and veins dark brown 
to brown. 

Redescription of the female paralectotype of 
Bracon linurus. — Similar to the female lec- 
totype. Body 7.5 mm long. Head in dorsal 
view 1.5 times as broad as long, eye 1.3 
times longer than temple, temple slightly 
less rounded (Fig. 58). Second tergite al- 
most as long as third tergite (Fig. 59, see 
arrows). Pronotum and propodeum with 
reddish yellow suffusion, mesoscutum and 
scutellum reddish yellow. 

Male and host unknown. 

Distribution. — Bolivia, Peru. 

Taxonomic position. — C. linurus is nearest 
to C. mesouurus (Szepligeti), their specific 
distinction is presented in the key-couplets 
10(11) -11(10). 

Cyclaulax lunatus (Szepligeti) 
(Figs 60-66) 

Bracon lunatus Szepligeti, 1906: 591 (in key) and 
594 (description) 9, type locality: "Peru: 
Pachitea", female lectotype (and three female 
paralectotypes, designated by Papp in 1969, 
see Shenefelt I.e.), in Magyar Termeszettudo- 
manyi Miizeum, Budapest and one female 
paralectotype in Nationaal Natuurhistorisch 
Museum, Leiden as exchange. - Shenefelt 
1978: 1504 (as Bracon lunatus, literature up to 
1906). Quicke 1991: 172 (as Cyclaulax lunatus 
comb, n., type depository). 

Type designation of Bracon lunatus. — Des- 
ignation of the female lectotype: (first label, 
printed) "Peru / Pachitea", second label is 
the lectotype card, third label is with the 
inventory number 1555 (second and third 
labels were attached by me); fourth label is 
with the actual name Cyclaulax lunatus 
(Szepligeti) given by Quicke in 1989. - 
Lectotype is in fairly good condition: (1) 
pinned by mesosoma; (2) right flagellum 



158 



Journal of Hymenoptera Research 





Figs 60-66. Cyclaulax lunatus (Szepligeti): 60 = scape in outer-lateral view, 61 = head in dorsal view, 62 = 
head in lateral view, 63 = basitarsus and spur of middle leg, 64 = pterostigma and vein r of fore wing, 65 = first 
discal cell of fore wing, 66 = tergites 1-3 



missing, left flagellum distally deficient; (3) 
right fore wing glued on a separate card. 

Designation of the four female paralecto- 
types. — (three females in Budapest Muse- 
um, one female in Museum Leiden): (first 
label, printed) "Peru / Pachitea"; second 
label is the paralectotype card, third label is 
with the inventory numbers 1556-1558 (in 
Budapest); fourth (in Budapest) and third 
(in Leiden) labels is with the actual name 
Cyclaulax lunatus (Szepligeti) given by 
Quicke in 1989. - The four paralectotypes 
are in good condition: (1) pinned by 
mesosoma; (2) flagelli partly missing, 
partly deficient. 

Taxonomic rectification. — Two female 
paralectotypes of Bracou binotatus Szepli- 
geti (in Museum Budapest, Nos 1545 and 
1547) and two female paralectotypes of B. 
enotatus var. 9 (one female in Museum 
Budapest, No. 1553, one female in Muse- 
um Leiden) proved to belong to B. lunatus - 
now to the genus Cyclaulax. For further 
comments see "Taxonomic remarks" un- 
der these two species. 

Redescription of the female lectoti/pe of 
Bracon lunatus. — Body 10 mm long. Left 
flagellum broken distally, i.e. with 24 
flagellomeres. Outer-lateral side of scape 
as in Fig. 60. First flagellomere 1.75 times 
and 24th flagellomere cubic, i.e. as long as 



broad. - Head in dorsal view (Fig. 61) less 
transverse, nearly 1.6 times as broad as 
long, eye 1.7 times as long as temple, 
temple rounded. Eye in lateral view 1.6 
times as high as wide and 1.46 times wider 
than temple, latter evenly broad beyond 
eye (Fig. 62). 

Mesosoma in lateral view almost twice 
as long as high. Hind femur 3.1 times as 
long as broad medially. Inner spur of 
middle tibia shorter than half basitarsus 
(Fig. 63). Vein r somewhat longer than 
width of pterostigma and issuing from its 
middle (Fig. 65). First discal eel less high, 
1-M 1.5 times as long as m-cu, 1-M straight 
and clearly not parallel with m-cu, 2-CU1 
straight (Fig. 64). - First tergite as long as 
broad behind, broadest at its two-thirds, its 
scutum and lateral part fairly wide; suture 
between tergites 2-3 deep and trisinuate; 
third tergite 1.65 times as long medially as 
long second tergite laterally (Fig. 66). Ovi- 
positor sheath longer than hind tibia + 
tarsus combined. 

Antenna blackish. Head black, median 
granulose field of face with luniform 
yellow macula, cheek also yellow, palpi 
brown. Ground colour of mesosoma testa- 
ceous, pronotum + presternum and propo- 
deum + metapleuron blackish to black. 
Fore leg rather dark brown, trochanters 



Volume 16, Number 1, 2007 



159 




Figs 67-76. Cyclaulax mesonurus (Szepligeti): 67 = scape in outer-lateral view, 68 = scape in inner-lateral view, 
69 = head in dorsal view, 70 = head in lateral view, 71 = fore femur, 72 = tarsus of first leg, 73 = hind femur, 74 
= first discal cell of fore wing, 75 = tergites 1-3, 76 = hypopygium and ovipositor apparatus 



and tarsus with reddish suffusion. Middle 
and hind legs black to dark brown, middle 
tarsus faintly reddish. Tergites 1-3 reddish 
and laterally blackish to black, further 
tergites black. Wings brown fumous, pter- 
ostigma blackish brown, veins proximo- 
distally blackish to light brown. 

Deviating features of four paralectotypes 
of B. lunatus, (three female in Museum 
Budapest, one female in Museum Leiden, 
from Peru: Pachitea). Two female paralec- 
totypes of B. binotatus var. 9 Szepligeti (in 
Museum Budapest), two female paralecto- 
types of B. enotatus var. 9 Szepligeti (one 
female in Museum Budapest, one female in 
Museum Leiden, from Peru: Marcapata) 
and one female (in Museum Leiden, from 
Peru: Pachytea) (the paralectotypes of B. 
binotatus and B. enotatus are representing B. 
lunatus, present rectification); total nine 
female specimens. - Similar to the female 
lectotype. Body 8-10 mm. Flagelli mainly 
distally deficient, rarely missing. Head in 
dorsal view 1.53-1.6 times as broad as long. 
Hind femur 3.1-3.5 times as long as broad 
medially. First tergite slightly broader 



behind than long (2 9). Tergites 1-3 
blackish with reddish suffusion (3 9). 

Male and host unknown. 

Distribution. — Peru. 

Taxonomic position. — C. lunatus (Szepli- 
geti) is nearest to C. grandiceps Cameron 
and C. paraguayensis (Szepligeti), their 
distinction is presented in the key-couplets 
1(6) - 7(8). 

Cyclaulax mesonurus (Szepligeti) 
(Figs 67-76) 

Bracou mesonurus Szepligeti, 1906: 591 (in key) 
and 593 (description) 9, type locality: "Boli- 
vien: Mapiri", female lectotype (designated by 
Papp in 1969) in Magyar Termeszettudomanyi 
Muzeum, Budapest; examined. - Shenefelt 
1978: 1511 (as Braeon mesonurus, literature up 
to 1906). Quicke 1991: 172 (as Cyclaulax 
mesonurus comb, n., type depository). 

Designation of the female holotype of 
Braeon mesonurus. — (first label, printed) 
"Bolivia / Mapiri"; second label is the 
lectotype card, third label is with the 
inventory number 1541 (second and third 
labels were attached by me), fourth label is 



160 



Journal of Hymenoptera Research 



with the actual name Cyclaulax mesonurus 
(Szepligeti) given by Quicke in 1989. - 
Holotype is in good condition: (1) pinned 
by the mesosoma; (2) both flagelli deficient 
distally; (3) tarsus of left fore leg missing; 
(4) damaged: distal end of costal-subcostal 
vein (proximal from pterostigma) of right 
fore wing, membrane between median (M 
+ CU1 and anal veins (1-1 A) of left fore 
wing longitudinally (i.e. parallel with these 
two veins) splitted. 

Redescription of the female holotype of 
Bracon mesonurus. — Body 9 mm long. Scape 
in outer-lateral view just less than 1.9 times 
as long as broad apically (Fig. 67), in inner- 
lateral view as in Fig. 68. Both flagelli 
deficient, right flagellum with 25 and left 
flagellum with 14 flagellomeres. First fla- 
gellomere almost 1.5 times as long as broad 
and 25th flagellomere cubic, i.e. just broader 
than long. - Head in dorsal view (Fig. 69) 
less transverse, 1.57 times as broad as long, 
eye 1.37 times as long as temple, temple 
moderately rounded. Eye in lateral view 1.5 
times as high as wide and 1.4 times wider 
than temple, latter beyond eye faintly 
narrowing ventrally (Fig. 70). 

Mesosoma in lateral view twice as long 
as high. Fore femur 3.2 times as long as 
broad distally (Fig. 71). Tarsomeres 2-4 of 
fore leg in lateral view short, third tar- 
somere 1.36 times as long as high apically 
and fourth tarsomere cubic (Fig. 72). Hind 
femur 3.1 times as long as broad distally 
(Fig. 73). Inner spur of middle tibia shorter 
than half length of basitarsus (cf. Fig. 42). - 
First discal cell less high, 1-M 1.5 times as 
long as m-cu, 1-M bent, 1-SR+M and 2CU1 
equal in length (Fig. 74). - First tergite 
clearly 1.5 times as long as broad behind, 
scutum narrow, i.e. lateral part of tergite 
wide; third tergite medially nearly twice as 
long as second tergite laterally; suture 
between them less trisinuate and less deep 
(Fig. 75). Hypopygium pointed, ovipositor 
sheath somewhat longer than hind tibia + 
tarsus combined (Fig. 76). 

Scape black, pedicel + flagellum brown- 
ish black. Head black, palpi dark brown. 



Mesosoma testaceous; blackish to black: 
prosoma, propodeum and metapleuron. 
Tegula testaceous. Tergites 1-3 reddish 
yellow, rest of tergites blackish to black. 
Legs black; fore tibia brownish yellow, fore 
tarsus yellow, middle tibia brown, middle 
tarsus yellow, hind tibia + tarsus brownish 
black. Wings brown fumous, pterostigma 
dark brown, veins brown to light brown. 

Male and host unknown. 

Distribution. — Bolivia. 

Taxonomic position. — C. mesonurus (Sze- 
pligeti) is nearest to C. linurus (Szepligeti), 
their distinction is in key-couplets 10(11) - 
11(10). 

Cyclaulax paraguayensis (Szepligeti) 
(Figs 77-87) 

Bracon paraguayensis Szepligeti, 1914: 184 9, type 
locality: "Paraguay", female lectotype (+ one 
female paralectotype designated by Papp in 
1969, see Shenefelt 1978 I.e.) in Magyar 
Termeszettudomanyi Muzeum, Budapest; 
examined. - Szepligeti 1906: 590 (in key). 
Shenefelt 1978: 1523 (as Bracon paraguayen- 
sis, literature up to 1906). Quicke 1991: 172 (as 
Cyclaulax paraguayensis comb, n., type de- 
pository). 

Type designation of Bracon paraguayensis. — 
Designation of the female lectotype and 
one female paralectotype: (first label, my 
handscript) "Paraguay / South America" 
(reverse label of the lectotype) "Paraguay" 
(with Szepligeti's handscript); second label 
is the lectotype and paralectotype cards, 
respectively, third label is with the in- 
ventory numbers 1538 (lectotype) and 
paralectotype (1539) (second and third 
labels were atteched by me); fourth label 
is with the actual name Cyclaulax para- 
guayensis (Szepligeti) given by Quicke in 
1989. - Lectotype is in fairly good condi- 
tion: (1) pinned by the mesosoma; (2) both 
flagelli deficient distally; (3) tarsomeres 3-5 
of left middle leg missing; (4) left pair of 
wings longitudo-distally somewhat 
creased. - Paralectotype is in poor condi- 
tion: (1) pinned by the mesosoma; (2) both 
flagelli deficient distally; (3) metasoma 



Volume 16, Number 1, 2007 



161 




Figs 77-87. Cyclaulax paraguayensis (Szepligeti), female lectotype: 77 = scape in outer-lateral view, 78 = scape 
in inner-lateral view, 79 = head in dorsal view, 80 = head in lateral view, 81 = tarsus of first leg, 82 = hind 
femur, 83 = pterostigma and vein r of fore wing, 84 = first discal cell of fore wing, 85 = tergites 1-3; female 
paralectotype: 86 = head in dorsal view, 87 = pterostigma and vein r of fore wing 



glued on a separate small card; (4) right 
fore wing missing, left hind wing some- 
what creased; (5) left hind leg missing. 

Redescription of the female lectotype of 
Bracon paraguayensis. — Body 9.5 mm long. 
Scape in outer-lateral view 1.5 times as 
long dorsally as broad apically (Fig. 77), in 
inner-lateral view as in Fig. 78. Both flagelli 
deficient, right flagellum with 13 and left 
flagellum with 17 flagellomeres. First fla- 
gellomere almost 1.4 times as long as broad 
and 17th flagellomere cubic. - Head in 
dorsal view (Fig. 79) less transverse, nearly 
1.6 times as broad as long, eye clearly 1.4 
times as long as temple, temple less 
rounded. Eye in lateral view 1.6 times as 
high as wide and just 1.3 times wider than 
temple, latter beyond eye evenly broad 
(Fig. 80). 

Mesosoma in lateral view 1.8 times as 
long as high. Tarsomeres 2-4 of fore leg 
long, in lateral view third tarsomere twice 
and fourth tarsomere nearly 1.6 times as 



long as broad apically (Fig. 81). Hind 
femur 3.1 times as long as broad and 
relatively more broadening distally 
(Fig. 82). Vein r one-fifth shorter than 
width of pterostigma (Fig. 83). First discal 
cell fairly high, 1-M 1.66 times longer than 
m-cu, 1-SR+M twice longer than 1-M, 1-M 
just bent (Fig. 84). - First tergite almost 1.2 
times as long as broad behind, scutum 
wide; third tergite just 1.5 times as long 
medially as long second tergite laterally; 
suture between them trisinuate (Fig. 85). 
Ovipositor sheath as long as hind tibia + 
tarsomeres 1-2 combined. 

Scape black, flagellum brownish black. 
Ground colour oi head and mesosoma 
black with reddish pattern: upper margin 
of eye, run of notaulix + hind field and 
lateral margin of mesoscutum, scutellum 
medially, metanotum and propodeum. 
Tegula reddish. Tergites reddish yellow, 
last two tergites dark brown. Legs trico- 
loured. Fore leg yellow, coxa brown; mid- 



162 



Journal of Hymenoptera Research 




Figs 88-94. Cyclaulax sicuaniensis (Szepligeti): 88 = scape in outer-lateral view, 89 = head in dorsal view, 90 = 
head in lateral view, 91 = hind femur, 92 = pterostigma and vein r of fore wing, 93 = first discal cell of fore 
wing, 94 = tergites 1-3 



die leg: coxa blackish brown, trochanters + 
femur basally + tibia apically brownish, 
otherwise leg yellow; hind leg: coxa and 
trochanters black, femur blackish with very 
weak reddish tint distally, tibia proximo- 
distally light brown to dark brown, tar- 
someres brown to dark brown, basally and 
apically reddish to rusty. Wings light 
brown fumous, pterostigma yellow, veins 
brown to light brown. 

Redescription of the female paralectotype of 
Bracon paraguayensis. — Similar to the fe- 
male lectotype. Body 9 mm long. Both 
flagelli deficient: right flagellum with 42 
and left flagellum with 28 flagellomeres. 
First flagellomere hardly 1.4 times as long 
as broad and 42nd flagellomere tarnsverse, 
i.e. somewhat broader than long. - Head in 
dorsal view (Fig. 86) 1.5 times as broad as 
long, eye 1.4 times as long as temple. 
Corporal colour similar to that of lectotype 
except hind leg: coxa black, trochanters + 
femur + tibia apically dark brown, other- 
wise leg yellow, tarsi apically brownish. 

Deviating features of 13 females (from 
the locality Paraguay: Hohenau; 12 females 
in Museum Budapest, 1 female in Museum 
Leiden). - Similar to the female lectotype 
and paralectotype. Body 6-9 mm long (6: 1 
9, 7: 2 9 9, 7.5: 3 9 9, 8: 4 9 9, 8.5: 2 9 9, 9: 1 
9)- Antenna (4 9 9) about as long as (2 9 9) 
or somewhat longer than body (2 9 9) and 



with 41, 43 and 47 antennomeres. Flagelli 
of further (i.e. 10 9 9) specimens either 
deficient or missing. Penultimate flagello- 
mere subcubic, i.e. a bit longer than broad. 
Head in dorsal view 1.5-1.58 times as 
broad as long, eye 1.4-1.5 times as long as 
temple. Hind femur 3.1-3.3 times as long 
as broad distally. Vein r of fore wing as 
long as width of pterostigma (1 9, Fig. 87) 
or more or less shorter (Fig. 83). First 
tergite as long as broad behind (2 9 9). 

Male and host unknown. 

Distribution. — Paraguay. 

Taxonomic position. — On the one hand, C. 
paraguayensis appears to be nearest to C. 
lunatus (Szepligeti) on the basis of their 
broad first tergites, and to C. flexuosus 
(Szepligeti) + C. sicuaniensis (Szepligeti) in 
view of their subcubic heads, on the other. 
Their separation is presented in key-cou- 
plets 4(5) - 5(4) and 13(18) - 17(16). 

Cyclaulax sicuaniensis (Szepligeti) 
(Figs 88-94) 

Bracon sicuaniensis Szepligeti, 1904: 185 9, type 
locality: "Peru: Sicuani", female holotype 
(designated by Papp in 1969) in Magyar 
Termeszettudomanyi Muzeum, Budapest; 
examined. - Szepligeti 1906: 591 (in key). 
Shenefelt 1978: 1538 (as Bracon sicuaniensis, 
literature up to 1906). Quicke 1991: 173 (as 
Cyclaulax sicuaniensis comb, n., type de- 
pository). 



Volume 16, Number 1, 2007 



163 



Designation of the female holotype of 
Bracon sicuaniensis. — (first label, my hand- 
script) "Peru / Sicuani", (reverse label with 
Szepligeti's handscript) "Sicuani"; second 
label is the holotype card and the third label 
is with the inventory number 1550 (labels 
1-3 were attached by me); fourth label is 
with the actual name Cyclaulax sicuaniensis 
(Szepligeti) given by Quicke in 1989. - 
Holotype is in fairly poor condition: (1) 
micropinned by mesosoma, pin covered 
with copper vitriol crystals; (2) missing: 
right antenna, left flagellum; (3) wings 
more or less creased; (4) legs less visible 
owing to the mounting (specimen on the 
micropin very near to the polyporus stage). 

Redescription of the female holotype of 
Bracon sicuaniensis. — Body 6 mm long. 
Scape in outer-lateral view 1.65 times as 
long dorsally as broad apically (Fig. 88). - 
Head in dorsal view (Fig. 89) less trans- 
verse, 1.5 times as broad as long, eye 1.35 
times as long as tempe, temple clearly 
narrowing. Eye in lateral view 1.5 times as 
high as wide and 1.2 times wider than 
temple, temple beyond eye slightly nar- 
rowing ventrally (Fig. 90, see arrows). 

Mesosoma in lateral view almost twice 
as long as high. Hind femur less broaden- 
ing distally, 2.9 times as long as broad 



distally (Fig. 91). Fore wing: vein r short, 
half as long as width of pterostigma 
(Fig. 92); first discal cell less high, 1-M 1.5 
times as long as m-cu, 1-M weakly bent 
(Fig. 93). - First tergite 1.33 times as long as 
broad behind, its scutum less narrowing 
anteriorly; third tergite nearly 1.6 times 
longer medially than second tergite later- 
ally; suture between them medially less 
sinuate (Fig. 94, see arrows). Ovipositor 
sheath shorter than hind tibia + tarsus 
combined. 

Scape blackish. Head and mesosoma 
black, upper margin of pronotum and 
propodeum entirely with faint rusty tint. 
Metasoma reddish yellow, apically black- 
ish. Legs black to brown, coxa + trochan- 
ters + femur of fore leg and femur of 
middle leg yellow. Wings dark brown 
fumous, pterostigma brown, veins dark to 
light brown. 

Male and host unknown. 

Distribution. — Peru. 

Taxonomic position. — C. sicuaniensis is 
nearest to C. flexuosus (Szepligeti) in view 
of their cubic heads, suture between 
tergites 2-3 medially less deeply sinuate 
and first tergite clearly longer than broad 
behind; for their distinction see key-cou- 
plets 15(14) - 17(16). 



KEY TO THE CYCLAULAX SPECIES DESCRIBED BY SZEPLIGETI AND C. GRAND1CEPS OF 

THE NEOTROPICAL REGION 

(KEY BASED MAINLY ON THE FEMALES) 



1 (6) First tergite broad, i.e. at least as long as broad behind (Figs 66, 85) or broader behind 

than long (Fig. 49). Suture between tergites 2-3 clearly trisinuate (Figs 49, 66, 85). . 

2 (3) First tergite broader behind than long (Fig. 49). Inner spur of middle tibia slightly 

longer than half basitarsus (Fig. 43). Groud colour of body reddish yellow, head 
black. Eye in dorsal view almost 1.4 times as long as temple (Fig. 38). Vein r issuing 
proximally from middle of pterostigma (Fig. 48). Face entirely black. 9: 12 mm. - 
Guiana C. grmuiiceps Cameron, 1911 

3 (2) First tergite at most as long as broad behind (Figs 66, 85). Inner spur of middle tibia at 

most as long as half basitarsus as ususally (Fig. 63). Ground colour of body 
testaceous to reddish, reddish yellow with more or less black pattern 

4 (5) Eye in dorsal view 1.7 times as long as temple (Fig. 61). Vein r somewhat longer than 

width of pterostigma and issuing from its middle (Fig. 65). Pterostigma blackish 
brown. First tergite as long as broad behind (Fig. 66). Scape in lateral view 1 .6 times 
as long ventrally as broad apically (Fig. 60). Head in dorsal view 1.55-1.6 times as 



164 Journal of Hymenoptera Research 

broad as long (Fig. 61). Ground colour of body testaceous to reddish yellow with 
black to blackish pattern on tergi. Head black. Median granulose field of face with 
luniform yellow macula. 9: 8-10 mm. - Peru C. lunatus (Szepligeti, 1906) 

5 (4) Eye in dorsal view 1.5-1.4 times as long as temple (Fig. 79). Vein r shorter than width 

of pterostigma and issuing from its middle (Fig. 83), exceptionally r as long as 
width of pterostigma. Pterostigma yellow. First tergite somewhat longer than broad 
behind (Fig. 85) and less usually as long as broad behind. Further details see 
couplet 14(15) C. paraguayensis (Szepligeti, 1914) 

6 (1) First tergite less broad, i.e. more or less longer than broad behind; suture between 

tergites 2-3 less variably trisinuate (Figs 7, 14, 23, 33, 57, 75, 94) 

7 (8) Head in dorsal view transverse, almost 1.9 times as broad as long, temple constricted, 

eye clearly twice as long as temple (Fig. 10). Hind femur 3.5-3.6 times as long as 
broad distally (Figs 12, 15). First tergite 1.55-1.6 times as long as broad behind, its 
lateral part narrow, third tergite about one-fifth longer than second tergite (Fig. 14). 
Tergites black; labrum + clypeus + cheek yellow or reddish yellow. 9: 6.5-7 mm. - 
Peru C. binotatus (Szepligeti, 1904) 

8 (7) Head in dorsal view less transverse, 1.5-1.7 times as broad as long, temple rounded to 

receded (Figs 27, 52, 69, 79) 

9 (12) Tarsomeres 2-4 of fore leg in lateral view short, third tarsomere nearly 1.4 times as 

long as high and fourth tarsomere cubic (Fig. 55, 72), fore tarsus yellow 

10 (11) Third tergite 1.5-1.4 times as long as second tergite; median sinuation of suture 

between tergites 2-3 deep (Fig. 57). Scutum of first tergite less narrow, i.e. lateral 
part of tergite less wide (Fig. 57). Hind femur 3.5-3.6 times as long as broad 
(Fig. 54). Ovipositor sheath as long as body. Only fore tarsus yellow. 9: 7.5-8 mm. - 
Bolivia, Peru C. linurus (Szepligeti, 1906) 

11 (10) Third tergite nearly twice as long as second tergite; median sinuation of suture 

between tergites 2-3 less deep (Fig. 75). Scutum of first tergite narrow, i.e. lateral 
part of tergite wide (Fig. 75). Hind femur 2.9 times as long as broad (Fig. 73). 
Ovipositor sheath as long as metasoma. Fore and middle tarsi yellow. 9 : 9 mm. - 

Bolivia C. mesonurus (Szepligeti, 1906) 

12(9) Tarsomeres 2-4 of fore leg in lateral view long, third tarsomere twice and fourth 
tarsomere 1.3-1.5 times as long as high (Figs 4, 81), tarsus black 

13 (18) Head in dorsal view subcubic, 1.5-1.55 times as broad as long (Figs 27, 79, 86, 89). 

Scape in lateral view 1.5 times as long dorsally as broad apically (Figs 26, 77-78, 88). 

14 (15) First tergite slightly (1.1-1.2 times) longer than (Fig. 85) or at most (and rather rarely) 

as long as broad behind. Hind femur more broadening distally, 3.1 times as long as 
broad distally (Fig. 82). Suture medially (between tergites 2-3) more sinuate 
(Fig. 85). Vein r as long as (Fig. 87) or, as ususally, slightly shorter (Fig. 83) than 
width of pterostigma. Pterostigma and fore leg (except brownish coxa) yellow. 9 : 
(6-)8-9.5 mm. - Paraguay C. paraguayensis (Szepligeti, 1904) 

15 (14) First tergite clearly (1.3-1.4 times) longer than broad behind (Figs 33, 94). Hind femur 

less broadening distally (Figs 30, 91). Suture medially (between tergites 2-3) less 
sinuate (Figs 33, 94). Vein r shorter than width of pterostigma (Figs 31, 92). 
Pterostigma dark to blackish brown; fore leg either brown or yellow 

16 (17) Temple in dorsal view moderately rounded (Fig. 27). Scutum of first tergite more 

narrowing anteriorly (Fig. 33, see arrows). Vein r clearly longer than half width of 
pterostigma (Fig. 31). Fore coxa + femur and middle femur rusty brown, o*: 6- 
7.5 mm. - Venezuela C. flexuosus (Szepligeti, 1902) 

17(16) Temple in dorsal view clearly narrowing (Fig. 89). Scutum of first tergite less 
narrowing anteriorly (Fig. 94, see arrows). Vein rhalf as long as width of 
pterostigma (Fig. 92). Fore coxa + femur and middle femur yellow. 9: 6 mm. - 
Peru C. sicuaniensis (Szepligeti, 1904) 

18 (13) Head in dorsal view transverse, 1.6-1.7 times as broad as long (Figs 1, 19). Scape in 
lateral view 1.7 times to twice as long dorsally as apically (Figs 18, 60) 



Volume 16, Number 1, 2007 



165 



19 (20) Second tergite less short, third tergite nearly 1.4 times as long as second tergite; first 
tergite 1.2 times as long as broad behind, scutum of tergite wide (Fig. 23, see 
arrows). First discal cell less high, 1-M straight veins relatively thin as usually 
(Fig. 22). Hind femur 3.6-3.3 times as long as broad medially (Figs 21, 25). Fore 
coxa yellowish. 9: 8-9 mm. - Peru C. enotatus (Szepligeti, 1904) 

20(19) Second tergite short, third tergite 1.6 times as long as second tergite; first tergite 1.4 
times as long as broad behind, scutum of tergite less wide (Fig. 7, see arrows). First 
discal cell high, 1-M bent, veins relatively thick (Fig. 6). Hind femur 2.9-3.1 times 
as long as broad medially (Figs 3, 8). Fore coxa blackish brown. 9: 8-9 mm. - 
Peru C. atriceps (Szepligeti, 1904) 



SPECIES OF THE GENUS 
CYCLAULAX CAMERON, 1911 

atriceps (Szepligeti, 1904) (Bracon) - Peru 

binotatus (Szepligeti, 1904) (Bracon) - Peru 

crassitarsis (Brues, 1912) [Bracon) - Brazil 

enotatus (Szepligeti, 1904) (Bracon) - Peru 

flexuosus (Szepligeti, 1904) (Bracon) - Vene- 
zuela 

grandiceps Cameron, 1911 - Guiana 

linurus (Szepligeti, 1906) (Bracon) - Bolivia, 
Peru 

lunatus (Szepligeti, 1906) (Bracon) - Peru 

mesonurus (Szepligeti, 1906) (Bracon) - 
Bolivia, Peru 

paraguayensis (Szepligeti, 1904) (Bracon) - 
Paraguay 

sicuanicnsis (Szepligeti, 1904) (Bracon) - 

Peru 

Remark. — Cyclaulax crassitarsis (Brues) re- 
mains unknown to me, hence not in- 
cluded in the present article. The species 
was assigned to Cyclaulax by Quicke (1989: 
119). 

ACKNOWLEDGEMENTS 

Mrs S. Ryder (The Natural History Museum, 
London) was kind enough to place at my disposal 
for examination the "Type" specimen of Cyclaulax 
grandiceps Cameron. My sincere gratitude should go to 
her cooperation in this respect. The examination of the 
generic type C. grandiceps essentially promoted my 
knowledge of the generic features of Cyclaulax 



Cameron. I express my sincere thank to Dr. D. 1 J, 
Quicke, as referee he considerably contributed to the 
improvements of the present paper. 



REFERENCES 

Achterberg, C. van. 1993. Illustrated key to the 
subfamilies of the Braconidae (Hymenoptera: 
Ichneumonoidea). Zoologische Verhandelingen 283: 
1-189. 

Cameron, P. 1911. On the Hymenoptera of the 
Georgetown Museum, British Guiana. "Timehri", 
the Journal of the Royal Agricultural and Commercial 
Society of British Guiana 1 (3rd series, 8): 308-330. 

Leathers, j. W., D. D. Judd,, and A. V. Z. Brower. 2005. 
A review of the species of the new world 
braconid genus Cyclaulacidea (Hymenoptera) with 
key and descriptions of nine new species, journal 
of Hymenoptera Research 14: 151-176. 

Quicke, D. L. J. 1989. Reclassification of some New 
World species of Braconidae (Hym., Braconidae). 
Entomologist's Monthly Magazine 125: 119-121. 

. 1991. The non-European Braconinae tvpes of 

Szepligeti housed in Budapest (Hymenoptera, 
Braconidae). Annates historico-naturales Musei na- 
tionals hungarici 83: 169-186. 

. 1995. Two new Neotropical genera of Braco- 



ninae (Hym., Braconidae). Entomologist's Monthly 
Magazine 131: 223-228. 

— . 1997. Subfamily Braconinae. Pp. 14^-174 in: 
Wharton, R. A., P. M. Marsh, <md M. J. Sharkey, 
eds. Manual of the New World Genera oj the Family 
Braconidae (Hymenoptera). Special Publication of 
the International Society of Hymenopterists 
No. 1. 439 pp. 

— , and A. Delobel. 1495. A new Neotropical 
braconine (Hvm., Braconidae) parasitic on Bru- 
chidae (Col.). Entomologist's Monthly h\agazine 
131: 215-221. 

— , and M. J. Sharkey. WSM. \ key to and notes on 
the genera of Braconinae (Hymenoptera: Braco- 
nidae) from America North of Mexico with 
descriptions of two new genera and three new 
species. Canadian Entomologist 121: 337-361. 



166 



Journal of Hymenoptera Research 



Shenefelt, R. D. 1978. Braconidae 10: Braconinae, 
Gnathobraconinae, Mesostoinae, Pseudodicro- 
geniinae, Telengainae, Ypsistocerinae, plus 
Braconidae in general, major groups, un- 
placed genera and species. Pp. 1425-1872 in: 
van der Vecht, J., and R. D. Shenefelt, eds. 
Hymenopterorum Catalogus. (nova editio) pars 
15. 



Szepligeti, Gy. 1902. Tropische Cenocoelionidae unci 
Braconiden aus der Sammlung des Ungarischen 
National-Museums. Termeszetrajzi Fiizetek 25: 39-84. 

. 1904. Sudamerikanische Braconiden. Annates 

Musei Nationalis Hungarici 2: 173-197. 

. 1906. Braconiden aus der Sammlung des 

Ungarischen National-Museums. Annates Musei 
Nationalis Hungarici 4: 547-618. 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 167-177 

Three New Species of Cenocoeliinae (Hymenoptera: Braconidae) with 
Novel Morphological Characteristics and Habitat Records 

Kevin M. Pitz and Michael J. Sharkey 

S-225 Agricultural Science Center North, Department of Entomology, University of Kentucky, 

Lexington, Kentucky 40546, USA; email: braconid@gmail.com, msharkey@uky.edu 

"Author for correspondence: Kevin M. Pitz, Department of Zoology - Insects, Field Museum of 

Natural History, Chicago, IL 60605, USA; tel 312-665-7710; fax 312-665-7754; braconid@gmail.com 



Abstract. — The purpose of this paper is to introduce three new Neotropical species of 
cenocoeliine wasps that expand the morphological and biological boundaries of the subfamily. 
These three species share a complex of morphological characters that distinguish them from other 
known species of Capitonius, and because of this we propose these species form a distinctive and 
monophyletic species group. Unlike other members of the subfamily, the new species are 
dorsoventrally flattened and represent some of the smallest recorded cenocoeliines. The following 
species are described; Capitonius subcrusta Pitz and Sharkey n. sp. from Mexico, C. vegrandis Pitz 
and Sharkey n. sp. from Costa Rica, and C. tenuiflagellum Pitz and Sharkey n. sp. from Colombia. 



The subfamily Cenocoeliinae is relatively 
small with approximately 70 described 
species (Achterberg 1997, Braet and van 
Achterberg 2001, Yu et al. 2005). The few 
cenocoeliines with known biologies are 
koinobiont endoparasitoids of mostly 
wood-feeding coleopteran larvae, mainly 
in the families Cerambycidae and Curcu- 
lionidae (Scolytinae) but with some host 
records of Buprestidae and non-scolytine 
Curculionidae (Saffer 1982, Shaw and 
Huddleston 1991). Cenocoeliines are 
known to parasitize hosts that utilize 
a variety of woody substrates, ranging 
from tree trunks (personal observation) to 
smaller branches and twigs (Shaw 1999), 
with scattered records from a variety of 
other plant material such as herbaceous 
stems, fruits, and nuts (Saffer 1982). Al- 
though cenocoeliines are cosmopolitan 
(van Achterberg 1997, Pitz and Sharkey 
2005), they are most diverse in the neo- 
tropics. The subfamily has been largely 
overlooked in the past, leaving many new 
species to be described (Achterberg 1997, 
Ent and Shaw 1998). 



Here, we describe three new species of 
Cenocoeliinae collected in Mexico, Costa 
Rica, and Colombia. These new species 
differ from other cenocoeliines in several 
ways. The mesosoma is dorsoventrally 
flattened (Fig. 1). The metasoma is inserted 
only slightly above the hind coxae (Fig. 1). 
The hind coxal grooves, which are ovipos- 
itor guides, are located along the anterior 
margin of the medial face of the hind coxae 
(Fig. 2d). The antennal scrobe is truncated 
and shallow, leaving the median ocellus 
outside the antennal scrobe and level with 
the lateral ocelli (Fig. 3a-c). The scape is 
swollen in the apical half (Fig. 4a). Females 
have short, thick antennae; the flagello- 
meres are about as wide as long, most 
having only a single row of longitudinal 
placodes that are nearly as long as a flagel- 
lomere (Fig 4b). Males (Fig. 5b) have nor- 
mal antennae with flagellomeres that are 
about twice as long as wide, each with 
multiple rows of placodes that range from 
one third to one half the length of 
a flagellomere. Finally, these wasps are 
some of the smallest members of the 



168 



Journal of Hymenoptera Research 




Fig. 1. Capitonius vegrandis, holotype female. 



subfamily Cenocoeliinae, with a maximum 
length of 3.28 mm. All of these characters 
are putative synapomorphies of the three 
newly described species. One of these 
remarkable species has lost the r-m cross- 
vein of the forewing (Fig. 3g), the first 
report of this characteristic for the sub- 
family. 

DISCUSSION 

Collection labels of the Mexican speci- 
mens state that they were found under the 
bark of a tree and we suggest that the 
dorsoventrally flattened body is an adap- 
tation for this habitat. Unrelated species in 
the genus Chartobracon (Braconinae) have 
members that are also dorsoventrally flat- 
tened (van Achterberg 1983). The type 
species, Chartobracon huggerti Achterberg, 
was reared from cocoons collected from 
cerambycid tunnels under the bark of 



spruce trees (van Achterberg 1983), and 
other adult specimens were collected un- 
der the bark of trees (Quicke and Sharkey 
1989). Species of Chartobracon range be- 
tween 2.7-2.9 mm in body length, (van 
Achterberg 1983, Quicke and Sharkey 
1989). These observations are consistent 
with the hypothesis presented here, i.e., the 
small flattened bodies of braconids attack- 
ing xylophagous beetles are adaptations to 
facilitate searching for hosts under bark. 

Both the newly described species of 
Capitonius and species of Chartobracon have 
short ovipositors, ranging between 1.96- 
2.24 mm for the new species of Capitonius, 
and 1.2 mm for Chartobracon (Quicke and 
Sharkey 1989), such that they would not be 
able to penetrate far into any substrate. 
This suggests these species may oviposit 
directly into the host larvae. This is in 
contrast to other dorsoventrally flattened 



Volume 16, Number 1, 2007 



169 




Fig. 2. Hind coxal grooves of female cenocoeliines: a, Foenomorpha ftlicornis; b, undescribed species ol 
Cenocoelius from Arizona, USA; c, Capitonius chontalensis; d, Capitonius subcrusta. 



species, such as those in the genus Atain/- 
colus that are known to use their oviposi- 
tors to penetrate the host's substrate from 
the outside; these wasps are larger (over 
5 mm) with longer ovipositors that allow 
them to penetrate into a substrate far 
enough to reach host larvae. 

Some cenocoeliines, such as Foenomorpha 
filicornis (Cameron), have a groove on the 
medial surface of the hind coxa that is 
situated along the longitudinal axis. This 
directs the ovipositor posteroventrally dur- 



ing oviposition (Fig. 2a). Other species, 
such as Capitonius chontalensis (Cameron), 
have a groove that is situated along the 
vertical axis of the coxa, between the 
midline and anterior margin that directs 
the ovipositor ventrally during oviposition 
(Fig. 2c). There are also species of Ceno- 
coeliinae, such as an undescribed species of 
Cenocoelius from Arizona, which have ^n 
intermediate position of the hind coxal 
grooves angled somewhere between the 
previously discussed orientations (Fig. 2b). 



170 



Journal of Hymenoptera Research 

C 





0o oooO 








Fig. 3. Illustrations of new species of Capitonius: a, C. subcrusta dorsal head; b, C. tenuiflagellum dorsal head; c, 
C. vegrandis dorsal head; d, C. tenuiflagellum lateral mesosoma illustrating sternaulus; e, C. subcrusta lateral 
mesosoma illustrating sternaulus; f, C. subcrusta wings; g, C. vegrandis wings. 

The ovipositor guides on the hind coxae of In most cenocoeliines, the ratio of the 

the newly proposed species are located ovipositor to body length is proportional to 

along the anterior margin of the hind coxae the angle of the ovipositor guide, such that 

and indicate that the ovipositor is directed the longest ovipositors are found in species 

anteroventrally during oviposition (Fig. with guides that are almost parallel to the 

2d). long axis of the coxae, and shorter ovipo- 



Volume 16, Number 1, 2007 



171 




Fig. 4. Capitonius subcrusta: a, scape, paratype female; b, flagellomeres, paratype female. 



sitors are found in species with grooves 
that direct the ovipositor ventrally. Exam- 
ining a randomly chosen representative 
from the species discussed above, we 
found F. filicomis had an ovipositor: body 
length ratio of 1.57, the undescribed spe- 
cies of Cenocoelius from Arizona had a ratio 
of 1.43, and C. chontalensis had a ratio of 
1.29, demonstrating that species with 
grooves that run closer to parallel to the 
midlength of the hind coxae tend to have 
longer ovipositors in relation to their body 
length. The newly described species have 
some of the shortest ovipositor: body ratios 
found in Cenocoeliinae, with ratios of 0.66, 
0.73, and 0.67 for C. tenuiflagellum, C. 
subcrusta, and C. vegrandis respectively. 
They also have the most anteriorly located 
ovipositor guides of any known member of 
the subfamily. 

Based on the microhabitat in which the 
Mexican specimens were collected, the 
orientation of the ovipositor guides, the 
small body size, and the short ovipositor 
length, we hypothesize members of the 
new species-group do not drill or probe 
into the woody substrate from the outside 



of the tree, as is typical of other members of 
the subfamily, but crawl under the bark 
and directly parasitize hosts. Morphologi- 
cal features of the antenna 1 scrobe also 
corroborate this hypothesis. The three 
species described here have truncated and 
shallow antennal scrobes, whereas all other 
members of the Cenocoeliinae have deep 
antennal scrobes that extend to the lateral 
ocelli. The antennal scrobe is a modification 
found in many hymenopterans that 
emerge from tunnels in wood. While 
emerging from the host substrate, the 
antennae are folded back and fit into the 
scrobe, thereby protecting them as the 
wasps emerge. All three new species have 
very shallow and truncated antennal 
scrobes suggesting that they have no need 
to protect the antennae. 

Generic limits are not yet well estab- 
lished in the Cenocoeliinae, making 
placement of these species problematic. 
In the most recent treatment of the group, 
van Achterberg's (1994) generic concepts 
are based on suites of characters, and 
lack a phylogenetic framework. His keys 
and diagnoses must be used to under- 



172 



Journal of Hymenoptera Research 




Fig. 5. Capitonius tenuiflagellum: a, lateral habitus , holotype female; b, lateral habitus, paratype male. 



Volume 16, Number 1, 2007 



173 



stand generic limits. Recent molecular 
analyses (Pitz in prep) provide a clearer 
understanding of relationships within 
Cenocoeliinae, though molecular data 
were not collected from these three 
species. 

MATERIALS AND METHODS 

Generic placement of the three new 
species is based on the results of phyloge- 
netic analyses of cenocoeliine genera (Pitz 
in prep.). We tentatively assign these 
species to the genus Capitonius primarily 
based on the ratios of the hind wing veins 
M+CU to 1M; no molecular data are avail- 
able for the newly described species, but 
their hind wing vein ratios are within the 
range found for other species of Capitonius. 
The specimens were compared to original 
descriptions and determined specimens of 
Capitonius to establish that they represent 
new species. 

Morphological terminology used follows 
that of Sharkey and Wharton (1997). All 
photographs were taken using a JVC KY- 
¥75 3CCD digital camera attached to 
a Leica MZ-16 stereoscope and were pre- 
pared using and measurements made with 
software in an Auto-Montage® imaging 
system. Scanning Electron Micrographs 
were produced using a Hitachi S-800 Field 
Emission Scanning Electron Microscope. 

Capitofiius subcrusta Species Group 

Description. — Female. Length: 2.68- 
3.40 mm. Head: Antenna shorter than 
forewing, with 22 flagellomeres [on intact 
specimens], scape swollen apically 
(Fig. 4a), flagellomeres with single row of 
longitudinal placodes which are almost the 
length of the flagellomere (Fig. 4b); anten- 
nal scrobe greatly reduced, ending imme- 
diately anterior to median ocellus, carina 
along antennal scrobe ending before lateral 
ocellus (Fig. 3a-c); frons slightly convex 
laterally, medial lamella of antennal scrobe 
short; vertex smooth, with sparse, weak 
punctures and sparse setae laterally; area 



posterior to lateral ocellus flat to slightly 
convex; face and clypeus smooth with 
sparse punctures and setae; clypeus 
with medioventral tooth, occipital carina 
smoothly rounded dorsomedially, without 
a sharp angle. Mesosoma: Dorsoventrally 
compressed (Figs. 1,5,6); pronotum not 
distinctly protruding anterodorsally, with 
large triangular pronope, and large tri- 
angular sub-pronope; propleuron without 
transverse carina; mesopleuron smooth 
with sparse setae; mesoscutum smooth 
with sparse setae, notauli narrow, com- 
posed of moderately sized rectangular 
fovea, nearly united immediately anterior 
to transscutal articulation; transscutal ar- 
ticulation not impressed as a groove, pres- 
ent as a line near midline only; scutellum 
without medioposterior depression; meta- 
notum moderately crenulate; metapleuron 
and propodeum irregularly areolate; hind 
coxal groove sharply defined, situated 
along anterior edge of hind coxa (Fig. 2d); 
forefemur normal, not flanged; rectangular 
space between hind coxae and metasomal 
insertion (metasomal pseudosternite) de- 
lineated by four strong carinae; transverse 
groove of metapleuron at level of epister- 
nal scrobe; tarsal claws simple. Wing: Vein 
1-M slightly curved (Fig. 3f-g); crossvein 
lr-m present or absent. Metasoma: Second 
median tergite smooth; third median ter- 
gite smooth and without acute lateral 
margin. 

Male (Fig. 5b) - as in female except for 
primary sexual characters and male anten- 
na longer than forewing, with 24 flagello- 
meres, scape not swollen apically, flagello- 
meres with two to three rows of longitudi- 
nal placodes. 

Diagnosis. — These species are the only 
known members of Capitonius that have the 
following suite of characters: body dorso- 
ventrally flattened, antennal scrobe re- 
duced, female with short and thick flagel- 
lomeres, female with hind coxal groove 
situated at extreme anterior margin of 
coax. This species group has a larger suite 
of hypothesized synapomorphies than G?- 



174 



Journal of Hymenoptera Research 




Fig. 6. Lateral habitus, holotype female, Capitonius subcrusta. 



pitonius itself, providing a high level of Distribution. — Southern Nearctic to 

support for the proposed monophyly of northern Neotropical (Mexico to Colom- 

C. subcrusta, C. tenuiflagellum, and C. bia). Containing only three known 

vegrandis. species. 



KEY TO SPECIES OF THE SUBCRUSTA GROUP 



1. Second cubital cell present (Fig. 3f) 2 

Second cubital cell absent (Fig. 3g) Capitonius vegrandis Pitz and Sharkey n. sp. 

2. Sternaulus complete, composed mostly of large, subovoid fovea (fig. 3d); carina 

bordering antennal scrobe nearly complete, ending three quarters of the way to 

lateral ocellus from antennal insertion (Fig. 3b) 

Capitonius tenuiflagellum Pitz and Sharkey n. sp. 

Sternaulus present only over posterior three-quarters of metapleuron, composed of 
small to moderate sized ovoid fovea (Fig. 3e); carina bordering antennal scrobe 
truncated, ending approximately halfway between antennal insertion and lateral 
ocellus (Fig. 3a) Capitonius subcrusta Pitz and Sharkey n. sp. 



Volume 16, Number 1, 2007 



175 



Capitonius subcrusta Pitz and Sharkey 
n. sp. 

Etymology. — Latin for under bark; a ref- 
erence to the microhabitat in which all 
known specimens were collected 

Description. — Holotype Female (Fig. 6). 
Length: 2.68 mm. Color: Body mostly 
melanic except testaceous (yellowish 
brown) as follows: ventral margin of malar 
space and clypeus; mandible basally; fore- 
and midlegs, hind tibia basally. Wings 
clear with stigma melanic; ovipositor red- 
dish brown. Head: Antenna with 22 fla- 
gellomeres; lateral carina bordering anten- 
nal scrobe reduced, ending far anterior to 
lateral ocellus (Fig. 3a); median lamella of 
antennal scrobe wide and slightly flat- 
tented anteriorly becoming acute poster- 
iorly. Mesosoma: Pronotum with three 
large fovea extending from sub-pronope 
along posterior margin, lateral margin 
weakly rugose, otherwise pronotum 
smooth; scutellar sulcus with four fovea; 
propleuron smooth with moderately dense 
setae; sternaulus incomplete, only occupy- 
ing posterior three-quarters of meso- 
pleuron, composed of single row of fovea 
(Fig. 3e); space between hind coxae and 
metasomal insertion weakly rugose be- 
tween carinae. Wings: Second submarginal 
cell present (Fig. 3f). Metasoma: First me- 
dian tergite smooth; length 1.03 times its 
apical width; length of ovipositonlength of 
forewing ratio 0.71. 

Biology. — Unknown. Specimens found 
under bark. 

Male. — As in female except for pri- 
mary sexual characters and male antenna 
longer than forewing, with 25 flagello- 
meres, scape not swollen apically, flagello- 
meres with two to three rows of longitudi- 
nal placodes. 

Material Examined. — Holotype female: MEX- 
ICO: Tam[auli]p[a]s., Mun.[icipio] Gomez Far- 
ias, Altas Cimas, 1000 m., 8-9.III.1986, P. Ko- 
varik, K. Haack, under bark. Deposited in Texas 
A&M University Collection. 



Paratype male: MEXICO: MEXICO: Chipas, 
Lagunas de Montebello National] Park, 
11. VIII. 1990 elfevation] 5000' [1524m], J.B. 
Woolley. Deposited in Texas A&M University 
Collection. 

Capitonius tenuiflagellum Pitz and 
Sharkey n. sp. 

Etymology. — Latin for small whip; a ref- 
erence to the short antennae of the female. 

Description. — Holotype Female (Fig. 5a). 
Length: 3.40 mm. Color: Body mostly 
melanic except testaceous (yellowish 
brown) as follows: head below level of 
antennal insertion; mandible basally; all 
legs except dorsal face of all tibiae, fore and 
mid basitarsi, hind tarsus, hind tibia 
basally. Wings clear with stigma melanic; 
ovipositor reddish brown. Head: [Antenna 
broken, 16 flagellomeres remaining on 
right, 11 on left]; lateral carina bordering 
antennal scrobe nearly complete, ending 
immediately anterior to lateral ocellus 
(Fig. 3b); median lamella of antennal 
scrobe acute over entire length. Mesosoma: 
Pronotum with four large fovea extending 
from sub-pronope along posterior margin, 
lateral margins weakly rugose, otherwise 
pronotum smooth; scutellar sulcus with 
four fovea; propleuron smooth with mod- 
erately dense setae; sternaulus complete, 
composed of single row of fovea (Fig. 3d); 
space between hind coxae and metasomal 
insertion with area between carinae rugo- 
sofoveate. Wings: Second submarginal cell 
present. Metasoma: First median tergite 
with two strong carinae basally, 1.40 times 
its apical width; length of ovipositor: 
length of forewing ratio 0.77. 
Biology. — Un kno w n . 

Male. — (Fig. 5b) - As in female except for 
primary sexual characters and male anten- 
na longer than forewing, with 24 flagello- 
meres, scape not swollen apically, flagello- 
meres with two to three rows of longitudi- 
nal placodes. 

Material Examined. — Holotvpe female: CO- 
LOMBIA, Amazonas, PNN Amacayacu, Mata- 
mata, 3 23'S 70 06'W, 150m, Apr' [il] 02-11/ 



176 



Journal of Hymenoptera Research 



2000, M.101, A. Parente Leg., [Amazonian rain- 
forest]. Deposited in the Instituto de Investiga- 
tion de Recursos Biologicos Alexander von 
Humboldt, Villa de Leyva, Colombia. 

Paratype male: COLOMBIA: Amazonas, 
PNN Amacayacu, Lorena, 3'02.86'S 69 59.7' W. 
200 m., red [sweep net], 28.vii.2001, M. Sharkey 
& D. Campos [Amazonian rainforest]. Deposit- 
ed in the Instituto de Investigation de Recursos 
Biologicos Alexander von Humboldt, Villa de 
Leyva, Colombia. 

Paratype male: Amazonas, PNN Amacayacu, 
Matamata, 3 23'S 70 06' W, 150m., Mar[ch] 12- 
19.2000, A. Parente Leg., [Amazonian rainfor- 
est]. Deposited in the Hymenoptera Institute 
Collection, University of Kentucky, Lexington. 

Capitonius vegrandis Pitz and Sharkey 
n. sp. 

Etymology. — Latin for tiny or diminutive, 
in reference to the short body length. 

Description. — Holotype Female (Fig. 1). 
Length: 2.95 mm. Color: Body mostly 
melanic except testaceous (yellowish 
brown) as follows: malar space; mandible 
basally, foreleg except apical tarsomere; 
midleg except apical and basal tarsomeres; 
wings clear with stigma melanic; oviposi- 
tor reddish brown. Head: Antenna with 22 
flagellomeres, lateral carina bordering an- 
tennal scrobe distinctly reduced, ending far 
anterior to lateral ocellus (Fig 3c); median 
lamella of antennal scrobe wide and blunt 
anteriorly, becoming narrow and acute 
posteriorly. Mesosoma: Pronotum with 
seven large fovea extending from sub- 
pronope to posteroventral corner, lateral 
margin smooth, otherwise pronotum 
smooth; scutellar sulcus with five fovea; 
propleuron smooth and bare; sternaulus 
complete, composed of single row of fovea 
(Fig 1); space between hind coxae and 
metasomal insertion (propodeal pseudo- 
sternite) weakly rugose between carinae. 
Wings: Second submarginal cell absent 
(Fig. 3g). Metasoma: First median tergite 
smooth; 1.26 times its apical width; length 
of ovipositor / length of forewing ratio = 
0.61. 



B iology. — unknown 
Male. — unknown. 

Material Examined. — Holotype female: COS- 
TA RICA, Prov.[incia] Punta.[renas], La Casona, 
R.B., Monteverde, [10 17' 54.3376" N 84° 47' 
33.1719"W], 1520m. Abr [April] 1994. N. 
Obando, L, N 253250_449700 #2820. Second 
label: COSTA RICA INBIO CRI001781333 [with 
bar code]. Deposited in Instituto National de 
Biodiversidad, Costa Rica. 

ACKNOWLEDGEMENTS 

Dr. Robert Wharton of Texas A&M University, 
Diana Arias of the Humboldt Institute in Colombia, 
and Carolina Godoy of INBio in Costa Rica loaned us 
specimens. We thank P. Southgate and R. Denomme 
for their help with SEM images. This research was 
funded by NSF grants DEB-0205982, DEB-0334945 and 
DEB-0072702 and supported by the University of 
Kentucky, Department of Entomology. Thanks to Dr. 
D.A. Potter, B.J. Sharanowski, C.A. Boring, and four 
anonymous reviewers for their critical remarks on 
previous versions of this manuscript. This is paper 
number 06-08-052 of the Kentucky Agriculture Exper- 
iment Station. 

LITERATURE CITED 

Achterberg, C. van. 1994. Generic revision of the 
subfamily Cenocoeliinae Szepligeti (Hymenop- 
tera: Braconidae). Zoologische Verhandelingen 292: 
1-52. 

— . 1997. Subfamily Cenocoeliinae. Pp. 185-192 in: 
Wharton, R. A., P. M. Marsh, and M. J. Sharkey, 
eds. Manual of the New World genera of the Family 
Braconidae (Hymenoptera). Special publication of 
the International Society of Hymeonpterists. 
No.l. 439 pp. 

Braet, Y. and C. van Achterberg. 2001. New species of 
the genera Foenomorpha Szepligeti (Cenocoeliinae) 
and Chelonus Panzer (Cheloninae) (Hymenoptera: 
Braconidae), from French Guiana, Suriname, and 
Brazil. Zoologische Mededelingen 75: 103-118. 

Ent, L. J. van der and S. R. Shaw. 1998. Species 
richness of Costa Rican Cenocoeliinae (Hyme- 
noptera: Braconidae): a latitudinal and altitudinal 
search for anomalous diversity, journal of Hyme- 
noptera Research 7: 15-24. 

Pitz, K. M. and M. J. Sharkey. 2005. Cenocoelius 
huggerti, the first record of the subfamily Cen- 
ocoeliinae (Braconidae: Hymenoptera) from 
Africa, journal of Hymenoptera Research 14: 92-95. 

Quicke, D. L. J. and M. J. Sharkey. 1989. A key to and 
notes on the genera of Braconinae (Hymenoptera: 
Braconidae) from America north of Mexico with 



Volume 16, Number 1, 2007 



177 



descriptions of two new genera and three new 
species. The Canadian Entomologist 121: 337-361. 

Saffer, B. 1982. A systematic revision of the genus 
Cenocoelius (Hymenoptera, Braconidae) in North 
America including Mexico. Polskie Pismo Entomo- 
logiczne. 52: 73-167. 

Sharkey, M. and R. Wharton. 1997. Morphology and 
Terminology. Pp. 19-38 in: Wharton, R. A., P. M. 
Marsh, and M. J. Sharkey, eds. Manual of the New 
World genera of the Family Braconidae (Hymenop- 
tera). Special publication of the International 
Society of Hymeonpterists. No.l. 439 pp. 



Shaw, M. R. 1999. Rearing records of two species of 
Cenocoelius Haliday from Britain. Entomologist's 
Gazette. 50: 283-286. 

and T. Huddleston. 1991. Classification and 

biology of braconid wasps (Hymenoptera: Braco- 
nidae). Handbooks for the Identification of British 
Insects.. 7, part 11: 1-126. 

Yu, D. S., C. van Achterberg, and K. Horstmann. 2005. 
Biological and taxonomic information of world Ich- 
neumonoidea, 2004. Electronic Compact Disk, 
Taxapad, Vancouver, Canada. http://www. 
Taxapad.com. 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 178-191 

The Male of Megachile nivalis Friese, with an Updated Key to Members 
of the Subgenus Megachile s. str. (Hymenoptera: Megachilidae) in 

North America 

Cory S. Sheffield and Sue M. Westby 

(CSS) Department of Biology, York University, 4700 Keele Street, Toronto, ON Canada M3J 1P3 

email: corysheffield@yahoo.ca 
(SMW) Agriculture and Agri-Food Canada, 32 Main Street, Kentville, NS, Canada B4N 1J5 



Abstract. — The previously unknown male of Megachile nivalis Friese is described. Males of this 
species are very similar to those of M. relativa Cresson and because of geographic overlap of the two 
species, many male specimens presently identified as M. relativa within collections may actually be 
M. nivalis. An identification key and illustrations of mandibles are provided for females and males 
of the subgenus Megachile s. str. of North America. Images of genitalia, selected sterna, the lower 
genal area, clypeus, and forewings for males of both M. relativa and M. nivalis are also provided for 
comparison and to facilitate differentiation of the two species. A tabular summary is also provided 
for species of Megachile in North America that are known from only one sex to encourage the search 
for possible additional synonyms or hitherto unknown sexes. 

Key words. — Apoidea, Megachilidae, Megachile (Megachile), Megachile nivalis, male description, 
North America 



Leafcutter and mason bees of the genus 
Megachile Latreille (Hymenoptera: Mega- 
chilidae) are a common and diverse group 
(Mitchell 1980, O'Toole and Raw 1991, 
Michener et al. 1994, Baker and Engel 
2006) whose members display many mor- 
phological and behavioral adaptations 
(Michener 2000). Currently, 55 extant sub- 
genera are recognized (Baker and Engel 
2006), 30 of which are known from the 
western Hemisphere. In North America, 
thirteen subgenera are indigenous, but an 
additional three have been introduced 
(Michener 2000, Cane 2003). Hurd (in 
Krombein et al. 1979) listed 134 species of 
Megachile in North America north of 
Mexico (including the genus Oialicodoma); 
Michener et al. (1994) indicate 139 species. 
Since the publication of Krombein et al. 
(1979), at least 111 of the cataloged North 
American species have undergone changes 
in subgeneric allocation (Raw 2002), and 



Raw (2004) indicates that the 519 described 
species of Megachile in the Western Hemi- 
sphere are now allocated to their proper 
subgenus. 

Much is known about the biology of 
many leafcutter bees due to their impor- 
tance in crop pollination (Pengelly 1955, 
Osgood 1974, Ivanochko 1979, Peterson et 
al. 1992, Free 1993, Richards 1993, Dela- 
plane and Mayer 2000, Raw 2002) and the 
fact that many species accept trap-nests 
(Medler 1959, 1964, Fye 1965, Krombein 
1967, Frolich and Parker 1983, O'Toole and 
Raw 1991). Most members of the genus 
nest above ground in pre-existing cavities 
or excavate into pithy stems or decompos- 
ing wood (Stephen 1956, Ivanochko 1979). 
Trap-nesting of bees allows detailed study 
of life-history, nest building, provisioning 
and egg laying behaviors (Medler 1959, 
1964, Klostermeyer and Gerber 1969, Fro- 
lich and Parker 1983, Kim 1992), and 



Volume 16, Number 1, 2007 



179 



incidences of cleptoparasitism (Scott et al. 
2000); it also allows one to associate males 
and females of the same species. However, 
several North American species within the 
subgenera Argyropile, Litomegachile, Mega- 
clniloid.es (including Derotropis and Xerome- 
gachile), and Xanthosarus (including Delo- 
megachile and Phaenosarus) are ground- 
nesters (Eickwort et al. 1981, Williams et 
al. 1986, Krombein and Norden 1995), 
some exclusively so. Eickwort et al. (1981) 
indicate that nesting in pre-existing cavities 
is probably derived within the Megachili- 
dae. 

Difficulties in associating males and 
females in several groups of bees some- 
times arise due to sexual dimorphism, and 
the comparatively ephemeral nature of 
males (Michener 2000). Unless specimens 
are collected and reared from nests (i.e., 
trap-nesting) or are caught during copula- 
tion, matching conspecifics of a given 
species is usually based on higher taxo- 
nomic classification, morphological simi- 
larities, geographic overlap, and specula- 
tion. Despite numerous studies, many 
North American bees are known from only 
one sex (Mitchell 1960), including 37% of 
described Megachile species (Table 1), and 
as a result the number of valid species for 
a given region may be significantly lower 
than suggested by catalogs (i.e., Krombein 
et al. 1979, Raw 2004). Because of their 
importance as pollinators, many bee collec- 
tions are based on surveys from floral hosts 
which provide no knowledge of conspeci- 
fics, and other commonly used methods of 
mass collecting bees, such as Malaise traps 
and pan trapping, are equally problematic 
for similar reasons; although males do get 
collected, pairing them with their respec- 
tive mates is not always possible. This 
issue is even more problematic for ground- 
nesting species (and their respective clep- 
toparasites); the nests of only a small pro- 
portion of these bees have been found or 
studied. 

Molecular methods are commonly em- 
ployed for analysis of bee phytogeny 



(Pedersen 1996, Danforth 1999, Danforth 
et al. 2006 a and b), revealing cryptic 
species (Carman and Packer 1996, Packer 
and Taylor 1997, Hebert et al. 2004, 
Simmons and Scheffer 2004) and more 
recently have been advocated for accurate 
identification of organisms to species level 
(Hebert et al. 2003, Savolainen et al. 2005), 
including insects (Pinto et al. 2003). As 
such, molecular methods offer much hope 
for associating male and female conspeci- 
fics of sexually dimorphic organisms (Pil- 
grim and Pitts 2006). 

Megachile Subgenus Megachile Latreille 

s. str 

The subgenus Megachile s. str. is a holarc- 
tic group found mostly in cool climates 
(Michener 2000), and five species are 
currently recognized in the western hemi- 
sphere (Mitchell 1935, 1962). Megachile s. 
str. are common members of temperate, 
boreal and subarctic North America, rang- 
ing from Nova Scotia (Sheffield et al. 2003) 
and Newfoundland through to Alaska and 
as far south as Mexico (Mitchell 1962). The 
North American species are M. centuncu- 
laris Linnaeus, M. inermis Provancher, M. 
immtivaga Cresson, M. nivalis Friese and M. 
relativa Cresson; M. centuncularis has a hol- 
arctic distribution (Mitchell 1935, Mich- 
ener, 2000) and is occasionally bivoltine in 
parts of its North American range (C.S. 
Sheffield, personal observations in Nova 
Scotia, Canada). Three species, Al. centun- 
cularis, M. inermis and M. relativa are 
collected commonly in trap-nest surveys 
within Canada and the northern United 
States (Stephen 1956, Medler 1959, Fye 
1965, Krunic and Salt L971, [vanochko 
1979, Sheffield 2006). In recent trap-nest 
surveys in Nova Scotia (Sheffield 2006), 
these three species accounted for 3.6%, 
13.8% and 21.4% of all bees collected, 
respectively, surpassed only by Osnmi 
tersula Cockerell (Osmiinae). Megachile iu- 
ermis also has been recorded nesting in 
decaying wood (Mitchell 1935, Stephen 
1956). Megachile montivaga differs from 



180 



Journal of Hymenoptera Research 



Table 1. North American species of Megachile (Hymenoptera: Megachilidae) known from only one sex, and 
suggested conspecific or synonomy. 





Species, author 


Described 


sex 


Subgenus 


Female 


Male 


Argyropile 


asterae Mitchell 


9 






rossi Mitchell 


9 






sabinensis Mitchell 


9 






tulariana Mitchell 


9 




Chelostomoides 


cartagenensis Mitchell 




6* 




subspinotulata Mitchell 


9 




Megachile 


nivalis Friese 


9 




Megachiloides 


alamosana Mitchell 




6* 




anograe Cockerell 


9 






boharti Mitchell 


9 






bradleyi Mitchell 


9 






bruneri Mitchell 




cJ 




coloradensis Mitchell 


9 






dulciana Mitchell 


9 






hilata Mitchell 


9 






hooker i Cockerell 


9 






impartita Mitchell 


9 






instita Mitchell 




cT 




inyoensis Mitchell 


9 






laguniana Mitchell 


9 






latita Mitchell 


9 






laurita Mitchell 


9 






macneilli Mitchell 




cT 




maurata Mitchell 




cT 




melanderi Mitchell 


9 






micheneri Mitchell 




6* 




mojavensis Mitchell 


9 






nelsoni Mitchell 


9 






oslari Mitchell 


9 






pagosiana Mitchell 


9 






parksi Mitchell 


9 






pseudolegalis Mitchell 




6" 




pseudonigra Mitchell 


9 






seducta Mitchell 


9 






semilaurita Mitchell 


9 






stoddardensis Mitchell 




cT 




subanograe Mitchell 


9 






sublaurila Mitchell 


9 






toscata Mitchell 


9 






victoriana Mitchell 




c? 




wyomingensis Mitchell 




6" 




yumensis Mitchell 


9 




Neochelynia 


aegra Mitchell 




6 




chichimeca Cresson 


9 




Pseudocentron 


morio Smith 


9 






pruina nigropinguis Mitchell 


9 




Sayapis 


dentipes Vachal 




6 




helianthi Cockerell 


9 






pugnata pomonae Cockerell 


9 




Xanthosarus 


gemula cressonii Dalla Torre 


9 






giliae Cockerell 




3 



Suggested conspecific or synonym 



o of M. hilata Mitchell 



9 of M. bruneri Mitchell 

Possible syn. of M. nevadensis Cresson 



o* of M. chichimeca Cresson 

9 of M. aegra Mitchell 

Possible form of M. pruina Smith 



9 known, but not described 



Volume 16, Number 1, 2007 



181 



other members of the subgenus in that it 
lacks the mandibular cutting edges (Fig. 1) 
and uses flower petals for nest cell con- 
struction (Mitchell 1935, Michener 2000). 
Unlike the preceding three species, M. 
montivaga does not appear to accept trap- 
nests and instead nests in pithy plant stems 
as well as in soil (Ivanochko 1979). Mitchell 
(1935) indicates variability in nesting site 
choice/substrate for these four species. 

In contrast to the preceding species 
which range as far south as Texas and 
Mexico, M. nivalis appears to have a more 
northern distribution (Mitchell 1935, 1962, 
Ivanochko 1979, Krombein et al. 1979). 
Little is known about its nesting biology 
but apparently females have been excavat- 
ed from a river bank in the Yukon Territory 
in Canada along with other Megachile 
species: M. giliae Cockerell (^'s only), M. 
montivaga, M. relativa, and M. frigida Smith 
(Ivanochko 1979). Males of M. nivalis have 
never been described (Mitchell 1935, 1962, 
Ivanochko 1979). Mitchell (1935, 1942, 
1962) and Ivanochko (1979) indicate the 
similarity of female M. nivalis to M. relativa 
(and to a lesser extent, M. centuncularis), 
the main distinguishing characters being 
differences in the color and length of the 
pubescence on T6, and the color of the 
scopal hairs on S6 (Mitchell 1935, 1962, 
Ivanochko 1979); M. nivalis females are also 
generally larger than those of M. relativa 
(Mitchell 1942). Mitchell (1942) suggested 
that M. nivalis may represent a race of M. 
relativa, and speculated that the male 
would be very similar to that of M. relativa 
(Mitchell 1935). He subsequently (Mitchell 
1942, 1962) examined specimens of male 
Megachile (not collected in copula) and 
suggested they may be related to female 
M. nivalis. However, he could not distin- 
guish these males from those of M. relativa 
(Mitchell 1962). 

Megachile giliae, with no described female 
(but with a northern distribution that 
overlaps with that of M. nivalis) has been 
indicated as the possible conspecific of M. 
nivalis (Mitchell 1935, Ivanochko 1979), but 



this association seems unlikely since mor- 
phologically, it classifies within the sub- 
genus Xanthosarus (Mitchell 1935, Krom- 
bein et al. 1979 as subgenus Delomegachile). 
However, the female of M. giliae has been 
collected and identified (see McGuire 1993, 
Bishop and Armbruster 1999), although no 
published descriptions presently exist. 

In 2005, a trap-nest survey conducted in 
Yellowknife, Northwest Territories yielded 
many female M. nivalis, ten specimens of 
a male Megachile, and a male and female of 
the cleptoparasite, Coelioxys funeraria Smith 
(Megachilidae). Examination of genitalia 
(Figs 2, 3) and S5, S6 and S8 (Fig. 4) of M. 
relativa and the newly collected male speci- 
mens revealed similarities, but consistent 
and distinct differences were noted (see 
below). Additional differences between the 
Yellowknife males and M. relativa were 
observed on the lower genal area (Fig. 5), 
the clypeal margin (Fig. 6), and in wing 
venation (Fig. 7) - further details are pro- 
vided below. These new specimens con- 
firm the association of male M. nivalis with 
the female, and an updated key to the 
North American members of the subgenus 
Megachile s. str. is provided. The specimens 
are currently held in the senior author's 
collection, but material will also be placed 
in the Packer Bee collection, York Univer- 
sity, Toronto, ON and the Canadian Na- 
tional Collection, Ottawa, ON upon com- 
pletion of this study. 

Diagnosis of Megachile s. str. — The body 
length of the subgenus varies considerably, 
from 7-20 mm (Michener 2000); M. inermis 
being the largest North American species. 
Females of Megachile s. str. can be separat- 
ed from other North American subgenera 
by the five-dentate mandibles, with the 
fourth tooth separated from the inner tooth 
by a broad and shallow interspace which 
lacks a cutting edge (Fig. 1), including M. 
montivaga, although the indentation be- 
tween the two inner teeth is obscure 
(Fig. la). Megachile s. str females also have 
a single incomplete cutting edge in the 
second interspace (Fig. 1) which is some- 



182 



Journal of Hymenoptera Research 













Fig. 1. Mandibles of female (left column) and male (right column) Megachile s. str.: M. montivaga (A and F); M. 
inermis (B and G); M. centuncularis (C and H); M. relaiiva (D and I); M. nivalis (E and J). 



Volume 16, Number 1, 2007 



183 








1 




^fe ' ^v 




■ I ■ 
^^B ill 






■ 


^^_ PP 


1 




Fig. 2. Dorsal and lateral views of genital capsules of Megachile relativa (A and C) an J A I nivali* (B and D). 



what reduced in M. relativa (Michener profile, not straight. The scopal hairs are 

2000) and absent in M. montivaga (Fig. la), uniformly colored, ranging from fulvous to 

Female M. montivaga also differs from the ochraceous; in M. nivalis the scopal hairs of 

other four species as T6 is concave in S6 (and often S5) are black, not concolorous 



184 



Journal of Hymenoptera Research 




Fig. 3. Close up of lateral views of genitalia for Megachile relativa (A) and M. nivalis (B). Horizontal lines and 
double-ended arrows show relative length of dorsal lobe of gonocoxite to base of gonoforceps. Scale bars 
represent 100 urn. 



Volume 16, Number 1, 2007 



185 



® 






Fig. 4. Sterna V (bottom), VI (middle) and VIII (top) of male Megachile relativa (A) and M. nivalis (B). 



with those of the preceding sterna which 
are ochraceous. 

Males have three-dentate mandibles, 
and the teeth are equally spaced except in 
M. inermis (Fig. lg). The mandibles also 
possess a narrow, distinct, basal or sub- 
basal, inferior process (Fig. 5a), the shape 
of the apical margin of this process varies 
slightly among species. The front coxa of 
male Megachile s. str. are hairy, with no 
spine and no patch of rufescent bristles, the 
exception being males of M. montivaga in 



which the front coxal spines are present, 
represented by dentiform tubercles which 
are often difficult to see. Further descrip- 
tions of this subgenus can be found in 
Mitchell (1935), Ivanochko (1979), and 
Michener (2000). Descriptions of the North 
American species are found in Mitchell 
(1935, 1962) and Ivanochko (1979). 

The subgeneric assignment of M. monti- 
vaga has come into question (Mitchell 1980, 
Michener 2000) due to the exceptions listed 
above. In addition, this species collects rose 




Fig. 5. The lower genal area of male Megachile relativa (A) and Al. nivalis (B). 



186 



Journal of Hymenoptera Research 





Fig. 6. Clypeal margins of male Megachile relativa (A) and M. nivalis (B). 



petals instead of leaves for nest construction 
(Mitchell 1935, Michener 2000), and unlike 
the remaining four species, M. montivaga 
does not appear to accept trap-nests. Rob- 
ertson (1903) proposed the generic name 
Cyphopyga just for M. montivaga. However, 
Mitchell (1935) indicated that the morpho- 
logical differences were at the species level 
and had no real generic or even subgeneric 



value, especially when the genitalia and 
hidden sterna of the males were considered. 
Despite this, he later (Mitchell 1980) recog- 
nized Cyphopyga as a four-toothed sub- 
genus of Megachile, with M. montivaga as 
the only species, but Michener (2000) 
considered it unnecessary to recognize 
a unique supraspecific taxon for this species 
alone. 



KEY TO NORTH AMERICAN BEES OF THE SUBGENUS MEGACHILE LATREILLE S. STR. 

The key provided is based on those provided elsewhere (Mitchell 1962, Ivanochko 1979) and 
through examination of specimens collected throughout Canada and currently held in the senior 
author's collection, and the Packer Bee collection, both at York University. The description of male 
Megachile nivalis is based on examination and dissection of the ten specimens collected in 
Yellowknife, NT in 2005. In mention of genitalia, gonoforceps refers to the distal (free) part of the 
gonocoxite plus the apical gonostylus. 





Fig. 7. Forewings of male Megachile relativa (A) and M. nivalis (B). Double ended arrows show the relative 
lengths of vein r (first submarginal cell) to Rs (second submarginal cell), r being much shorter in M. nivalis. 



Volume 16, Number 1, 2007 187 

FEMALES 

1 a. Scopal hairs on S6 (and often S5) black nivalis Friese 

b. Scopal hairs entirely pale fulvous or ochraceous 2 

2 a. T6 with a mixture of very short, suberect or appressed pubescence and abundant, 

long and erect pubescence which is visible in profile 3 

b. T6 with few erect hairs, showing mostly very short, sub-erect or appressed 

pubescence in profile 4 

3 a. Pubescence of T6 entirely dark centuncularis Linnaeus 

b. T6 with erect and appressed golden tomentum relativa Cresson 

4 a. T6 concave in profile; mandibles entirely lacking a cutting edge (Fig. la); punctures 

of clypeus and supraclypeal area coarse and close, interspaces much less than their 
diameter up to the edges of the distinct median impunctate line; interocellar 
distance subequal to distance of ocelli from edge of vertex; smaller species 

(<11 mm) montivaga Cresson 

b. T6 straight in profile; mandibles with a distinct cutting edge in the second 
interspace (Figs lb-e), surface of clypeus and supraclypeal area highly polished in 
central third, especially on apical ends, surface impunctate and/or with interspaces 
much greater than puncture diameter; interocellar distance much less than the 

distance from ocelli to edge of vertex; larger species (13 mm or more) 

inermis Provancher 

MALES 

For users of Mitchell's (1962) "Bees of the Eastern United States", substitute couplets 2 and 3 
below into his couplet 17 - page 113) 

1. a. Distance from the apex of the middle tooth to the apex of the inner tooth nearly 

twice as great as the distance from the apex of the middle tooth to the apex of the 
outer tooth (Fig. lg); interocellar distance much less than the distance from ocelli to 

edge of vertex; larger species (13 mm or more) inermis Provancher 

b. Distance from the apex of the middle tooth to the apices of either the inner or outer 
teeth subequal (Figs If, h-j); interocellar distance subequal to distant from ocelli to 
edge of vertex; smaller species (<12 mm) 2 

2. a. Clypeal margin with a distinct median tubercle (Fig. 6); surface of T6 polished 

above carina, the central punctures separated by their diameter 3 

b. Clypeal margin not tuberculate, but possibly narrowly produced medially (a few 
minute crenulations may also be visible medially in M. montivaga); surface of T6 
either more closely punctate (the interspaces less than one puncture diameter) or 
surface tuberculate 4 

3. a. Hypostomal tubercle short (Fig. 5a); hypostomal concavity shallow and not well 

defined (Fig. 5a); hypostomal carina distinct for most of its length (pile must be 
removed to see these features) (Fig. 5a); clypeal margin sinuous on either side of 
the prominent, shining, median tubercle (Fig. 6a); dorsal lobe of gonocoxite short, 
not attaining the base of gonoforceps (Fig. 3a), its length subequal to the width of 
gonobase (Figs 2a and c); vein r of the first submarginal cell normally subequal to 

vein Rs of the second submarginal cell (Fig. 7a) relativa Cresson 

b. Hypostomal tubercle more prominent and wider at base (Fig. 5b); hypostomal 
concavity deeper and well defined; hypostomal carina interrupted by the 
hypostomal tubercle (Fig. 5b); clypeal margin slightly curved to nearly straight 
on either side of the less prominent, median tubercle (Fig. 6b); dorsal lobe of 
gonocoxite long, fully attaining the base of gonoforceps (Fig. 3b) and longer than 
the width of gonobase (Figs 2b and d); vein r of the first submarginal cell shorter 
than vein Rs of the second submarginal cell (Fig. 7b) nivalis Friese 



188 



Journal of Hymenoptera Research 



a. Coxal spines represented by dentiform tubercles; carina of T6 with a definite 
median emargination, apical margin of the segment with conspicuous inner teeth 
and spine-like lateral teeth, the surface above the carina dull, minutely rugoso- 
punctate montivaga Cresson 

b. Coxal spines entirely lacking; carina of T6 with an obscure median emargination, 
the apical margin of the segment with broad inner teeth and obscure lateral teeth, 
the surface above the carina with numerous small tubercles, the punctures very 
obscure centuncularis Linnaeus 



Megachile nivalis Friese 

Megachile nivalis Friese, 1903. Ztschr. System. 
Hym. Dipt. 3: 246. 9 

Megachile (Anthemois) santiamensis Mitchell, 
1934. Trans. Am. Ent. Soc. 59: 311. 9 

Megachile (Anthemois) nivalis Mitchell, 1935. 
Trans. Am. Ent. Soc. 61: 174. 9 

Megachile (Anthemois) nivalis Mitchell, 1942. Pan- 
Pacific Ent. 38: 15-16. 9 (J misdet.) 

Megachile (Megachile) nivalis Mitchell, 1962. 
North Carolina Ag. Exp. Stat. Tech. Bull. 
152: 129. 9 

Description of male presented here 
follows format used by Mitchell (1962). 

Male. — Length 9-12 mm; entirely black 
except as follows: tegula testaceous along 
margins, basal tarsal segment black to 
somewhat reddened, following segments 
reddish testaceous; eyes slightly conver- 
gent below; clypeal margin nearly straight 
on either side of a distinct but small 
median tubercle (Fig. 6b); mandible three- 
dentate, with a rather narrow, sub-basal, 
inferior tooth which is subtruncate apically 
(Fig. 5a - Megachile relativa, but similar in 
structure); apical segment of flagellum 
slender and elongate; distance of lateral 
ocellus from margin of vertex and from 
margin of eye subequal; cheek somewhat 
broader than compound eye; punctures 
fine, slightly separated across vertex pos- 
teriorly, sparse between ocelli and eye, 
becoming close on cheek above and dense- 
ly crowded or rugose below; face below 
ocelli rather coarsely rugosopunctate, be- 
coming finely so below antennae and on 
clypeus; hypostomal depression well de- 
fined (Fig. 5b), hypostomal tubercle long 
and relatively prominent, broadly inter- 



rupting hypostomal carina (Fig. 5b); pu- 
bescence golden, becoming paler on lower 
part of cheek, quite long and copious 
around antenna and lower part of face, 
on cheek below and on thorax laterally and 
posteriorly; vertex with an admixture of 
pale and black pubescence; mesoscutum 
and scutellum with more or less inter- 
mixed light and dark hairs which are quite 
long and erect but thin; mesoscutum dull, 
punctures close, shallow, not very coarse, 
slightly separated only in center of disc; 
punctures of scutellum slightly separated 
along mid-line, but otherwise quite uni- 
formly close, those on axilla much finer 
and densely crowded; pleura dull, punc- 
tures shallow, quite close and poorly- 
defined; propodeum relatively smooth 
and shining; basitarsi quite short and 
slender; mid tibial spur short but well 
developed; tegula shining, rather uniform- 
ly, minutely and rather closely punctate; 
wings subhyaline, veins brownish, vein r 
of the first submarginal cell shorter than 
vein Rs of the second submarginal cell 
(Fig. 7b); T2-T4 shallowly grooved or de- 
pressed across base, basal margin of 
grooves not distinctly carinate, apical mar- 
gins of terga depressed only toward sides, 
depressed medially only on T4 and T5, 
pale apical fasciae evident at extreme sides 
of the more basal terga, more or less 
complete on T4 and T5, discal pubescence 
rather thin, largely black but with pale 
hairs evident toward sides, length of discal 
pubescence exceeding apical margin of all 
terga when viewed laterally, basal tergum 
covered with copious, elongate, whitish 
pubescence; punctures very fine, surface 



Volume 16, Number 1, 2007 



ISM 



shining, close on T2 barely evident on Tl, 
quite sparse on T3 and T4, becoming 
somewhat coarser laterally, but still well 
separated, T5 with somewhat closer and 
coarser punctures throughout; T6 shining, 
carina very low, broadly and shallowly in- 
curved medially, punctures fine and close 
above carina, separated by their diameter, 
becoming somewhat more coarse and 
sparse laterally, inner teeth of apical 
margin broadly carinate, widely separated, 
relatively near the short, acute, lateral 
teeth; 17 quite prominent, broad and short, 
with a deep excavation on dorsal surface; 
S1-S4 exposed, closely but rather obscurely 
punctate, apical margins of S2-S4 broadly 
yellowish-hyaline and with thin, apical 
fringes of pale hairs; setose area of S5 
restricted, finely setose (Fig. 4); S6 sparsely 
setose on each side, apical lobe barely 
evident (Fig. 4); gonoforceps slender with 
acute apex, gonocoxite basally with a dis- 
tinct dorsal lobe which fully attains the 
base of gonoforceps (Figs 2 and 3). 

Distribution. — The type locality for Mega- 
chile nivalis is Pikes Peak, Colorado (Mitch- 
ell 1935). This species is most common in 
northwestern areas of North America, 
having been reported from Alaska, Yukon 
Territory, Northwest Territories, British 
Columbia, Alberta, Saskatchewan, Mani- 
toba, Ontario and Quebec. It is less 
common in the southern limits of its range 
which include Washington, Oregon, Idaho, 
Montana, Wyoming, Minnesota, and Colo- 
rado. It has also been reported from Maine 
(Mitchell 1962). 

ACKNOWLEDGMENTS 

We thank Anne Gunn, Department of Resources, 
Wildlife and Economic Development, NT, for her 
interest in bee studies in the Northwest Territories, 
and for setting up trap-nests in Yellowknife in 2005; 
Anne and Suzanne Carriere, also from the Depart- 
ment of Resources, Wildlife and Economic Develop- 
ment provided the opportunity for one of us (C. 
Sheffield) to come to Yellowknife to collect and 
discuss northern bees in 2005. Thanks also to Laurence 
Packer, York University, Toronto, ON for helpful 
comments during the preparation of this manuscript, 



Anthony Raw (Departmento de Ciencias Biologicas, 
Universidade Estadual Santa Cruz, Ilheus, Bahia, 
Brazil) for providing information of western hemi- 
sphere Megachile, and Debra L. Moreau, Atlantic Food 
and Horticulture Research Centre, Agriculture and 
Agri-Food Canada, Kentville, NS for producing the 
SEM images. All illustrations were prepared by Sue 
M. Westby. 

Final preparation of this manuscript was completed 
by the senior author at York University, Toronto, ON, 
Canada during a Post Doctoral Fellowship with 
Laurence Packer supported through funding to the 
Canadian Barcode of Life Network from Genome 
Canada (through the Ontario Genomics Institute), 
NSERC (Natural Sciences and Engineering Research 
Council of Canada) and other sponsors listed at 
www.BOLNET.ca 

LITERATURE CITED 

Baker, D. B. and M. S. Engel. 2006. A new subgenus of 
Megachile from Borneo with arolia (Hymenoptera: 
Megachilidae). American Museum Novitates 3505: 
1-12. 

Bishop, J. A. and W. S. Armbruster. 1999. Thermoreg- 
ulatory abilities of Alaskan bees: effects of size, 
phylogeny and ecology. Functional Ecology 13: 
711-724. 

Cane, J. H. 2003. Exotic non-social bees (Hymenoptera: 
Apiformes) in North America: ecological implica- 
tions. Pp. 113-126. in: Strickler, K., and J. H. Cane 
eds. For nonnative crops, whence pollinators of the 
future? Thomas Say Publications in Entomologv: 
Proceedings. Entomological Society of America, 
Lanham, MD. 

Carman, G. M. and L. Packer. 1996. A cryptic species 
allied to Halictus ligatus Say (Hymenoptera: 
Halictidae) detected by allozyme electrophoresis. 
Journal of the Kansas Entomological Society 69: 
168-176. 

Danforth, B. N. 1999. Phylogeny of the bee genus 
Lasioglossum (Hymenoptera: Halictidae) based on 
mitochondria COI sequence data. Systematic 
Entomology 24: 377-393. 

, J. Fang, and S. Sipes. 2006a. Analysis of 

family-level relationships in bees (Hymenoptera: 
Apiformes) using 28S and two previously un- 
explored nuclear genes: CAD and RNA poly- 
merase II. Molecular Phylogenetics and Evolution 3 C >: 
358-372. 
-, S. Sipes, J. Fang, and S. G. Brady. 200bb. The 



history of early bee diversification based on five 
genes plus morphology. Proceedings of the National 
Academy of Sciences 103: 15118-15123. 

Delaplane, K. S. and D. F. Mayer. 2000. Crop Pollination 
by Bees. CABI Publishing, Mew York, NY. 

Eickwort, G. C, R. W. Matthews, and J. Carpenter. 
l l 'Sl. Observations on the nesting behaviour of 
Megachile rubi and AI. texana with a discussion of 



190 



Journal of Hymenoptera Research 



the significance of soil nesting in the evolution of 
megachild bees (Hymenoptera: Megachilidae). 
Journal of the Kansas Entomological Society 54: 
557-570. 

Free, J. B. 1993. Insect Pollination of Crops. Second 
Edition. Academic Press, Inc. San Diego, CA. 

Fye, R. E. 1965. Biology of Apoidea taken in trap nests 
in northwestern Ontario. Canadian Entomologist 
97: 863-877. 

Frohlich, D. R. and F. D. Parker. 1983. Nest building 
behavior and development of the sunflower 
leafcutter bee: Eumegachile [Sayapis) pugnata 
(Say) (Hymenoptera: Megachilidae). Psyche 90: 
193-209. 

Hebert, P. D. N., A. Cywinska, S. L. Ball, and J. R. 
deWaard. 2003. Biological identifications through 
DNA barcodes. Proceedings of the Royal Society of 
London B 270: 313-321. 

, E. H. Penton, J. M. Burns, D. H. Janzen, and 

W. Hallwachs. 2004. Ten species in one: DNA 
barcoding reveals cryptic species in the neotrop- 
ical skipper butterfly Astraptes fulgerator. Proceed- 
ings of the National Academy of Sciences 101: 
14812-14817. 

Ivanochko, M. 1979. Taxonomy, biology and alfalfa 
pollinating potential of Canadian leaf-cutter bees - 
genus Megachile Latreille (Hymenoptera: Megachili- 
dae). M.Sc. Thesis, McGill University, Montreal, 
QC. 378 pp. 

Kim, J.-K. 1992. Nest dimensions of two leaf-cutter 
beees (Hymenoptera: Megachilidae). Annals of the 
Entomological Society of America 85: 85-90. 

Klostermeyer, E. C. and H. S. Gerber. 1969. Nesting 
behavior of Megachile rolundata (Hymenoptera: 
Megachilidae) monitored with an event recorder. 
Annals of the Entomological Society of America 62: 
1321-1325. 

Krombein, K. V. 1967. Trap-nesting Wasps and Bees: 
Life Histories, Nests, and Associates. Smithsonian 
Press, Washington, DC. 

and B. B. Norden. 1995. Notes on the behavior 

and taxonomy of Megachile (Xeromegaclule) brim- 
leyi Mitchell and its probable cleptoparasite, 
Coelioxys (Xerocoelioxys) galactiae Mitchell (Hyme- 
noptera: Megachilidea). Proceedings of the Entomo- 
logical Society of Washington 97: 86-89. 
, P. D. Hurd, Jr., D. R. Smith, and B. D. Burks. 



1979. Catalog of Hymenoptera in America North of 
Mexico. Volume 2. Apocrita (Aculeata). Smithsonian 
Institution Press, Washington. 

Krunic, M. D. and R. W. Salt. 1971. Seasonal changes 
in glycerol content and supercooling points of 
Megachile rotuudata (F.) and M. relativa Cress. 
Canadian journal of Zoology 49: 663-666. 

McGuire, A. D. 1993. Interactions for pollination 
between two synchronously blooming Hedysar- 
ium species (Fabaceae) in Alaska. American journal 
of Botany 80: 147-152. 



Medler, J. T. 1959. A note on Megachile centunularis 
(Linn.) in Wisconsin (Hymenoptera: Megachili- 
dea). Canadian Entomologist 91: 113-115. 

. 1964, A note on Megachile (Sayapis) pugnata 

pugnata Say in trap-nests in Wisconsin (Hyme- 
noptera: Megachilidae). Canadian Entomologist 96: 
918-921. 

Michener, C. D. 2000. The Bees of the World. Johns 
Hopkins University Press, Baltimore, MD. 

, R. J. McGinley, and B. N. Danforth. 1994. The 

Bee Genera of North and Central America. Smithso- 
nian Institution Press, Washington, DC. 

Mitchell, T. B. 1935. A revision of the genus Megachile 
in the Nearctic region. Part III. Taxonomy of 
subgenera Anthemois and Delomegachile (Hyme- 
noptera: Megachilidae). Transactions of the Amer- 
ican Entomological Society 61: 155-208. 

. 1942, Notes and descriptions of Nearctic 

Megachile (Hymenoptera, Megachilidae). Pan-Pa- 
cific Entomologist 18: 115-118. 

. 1960, Bees of the Eastern United States. 

Volume I. North Carolina Agricultural Experi- 
ment Station Technical Bulletin No. 141, Raleigh, 
NC. 

. 1962, Bees of the Eastern United States. 

Volume II. North Carolina Agricultural Experi- 
ment Station Technical Bulletin No. 152, Raleigh, 
NC. 

1980, A generic revision of the megachiline 



bees of the Western Hemisphere. North Carolina 
State University Department of Entomology. 
95 pp. 

Osgood, C. E. 1974. Relocation of nesting populations 
of Megachile rolundata, an important pollinator of 
alfalfa. Journal of Apicultural Research 13: 67-73. 

OToole, C. and A. Raw. 1991. Bees of the World. 
Bland ford, London, UK. 

Packer, L. and J. S. Taylor. 1997. How many hidden 
species are there? An application of the phyloge- 
netic species concept to genetic data for some 
comparatively well known bee "species". Cana- 
dian Entomologist 129: 587-594. 

Pedersen, B. V. 1996. A phylogenetic analysis of cuckoo 
bumblebees (Psithyrus, Lepeletier) and Bumble- 
bees (Bombus, Latreille) inferred from sequences 
of the mitochondrial gene cytochrome oxidase I. 
Molecular Phylogenetics and Evolution 5: 289-297. 

Pengelly, D. H. 1955. The biology of bees of the genus 
Megachile with special reference to their importance 
in alfalfa seed production in southern Ontario. Ph.D. 
Thesis, Cornell University, Ithaca, NY. 269 pp. 

Peterson, S. S., C. R. Baird, and R. M. Bitner. 1992. 
Current status of the alfalfa leafcutter bee, 
Megachile rotuudata, as a pollinator of alfalfa seed. 
Bee Science 2: 135-142. 

Pilgrim, E. M. and J. P. Pitts. 2006. A molecular 
method for associating the dimorphic sexes of 
velvet ants (Hymenoptera: Mutillidae). Journal of 
the Kansas Entomological Society 79: 222-230. 



Volume 16, Number 1, 2007 



191 



Pinto, M. A., J. S. Johnston, W. L. Rubink, R. N. 
Coulson, J. C. Patton, and W. S. Sheppard. 2003. 
Identification of Africanized honey bee (Hyme- 
noptera: Apidae) mitochondrial DNA: validation 
of a rapid polymerase chain reaction-based assay. 
Annals of the Entomological Society of America 96: 
679-684.' 

Raw, A. 2002. New combinations and synonymies of 
leafcutter and mason bees of the Americas 
(Megachile, Hymenoptera, Megachilidae). Zootaxa 
71: 1-43. 

. 2004, Nomenclatural changes in leafcutter 

bees of the Americas: Megachile Latreille 1802 
(Hymenoptera; Megachilidae). Zootaxa 766: 1—4. 

Richards, K. W. 1993. Non-/4r>/s bees as crop pollina- 
tors. Revue Suisse de Zoologie 100: 807-822. 

Robertson, C. 1903. Synopsis of Megachilidae and 
Bombinae. Transactions of the American Entomolog- 
ical Society 29: 163-178. 

Savolainen, V., R. S. Cowan, A. P. Vogler, G. K. 
Roderick, and R. Lane. 2005. Towards writing the 
encyclopaedia of life: an introduction to DNA 
barcoding. Philosophical Transactions of the Royal 
Society B 360: 1805-1811. 

Scott, V. L., S. T. Kelley, and K. Strickler. 2000. 
Reproductive biology of two Coelioxys cleptopar- 
asites in relation to their Megachile hosts (Hyme- 



noptera: Megachilidae). Annals of the Entomologi- 
cal Society of America 93: 941-948. 

Sheffield, C. S. 2006. Diversity and management of bees 
for the pollination of apple in the Annapolis Valley of 
Nova Scotia. Ph.D. Thesis, University of Guelph, 
Guelph, ON. 301 pp. 

, P. G. Kevan, R. F. Smith, S. M. Rigby, and R. E. 

L. Rogers. 2003. Bee species of Nova Scoita, 
Canada, with new records and notes on bi- 
onomics and floral relations (Hymenoptera: 
Apoidea). Journal of the Knn^a> Entomological 
Society 76: 357-384. 

Simmons, R. B. and S. J. Scheffer. 2004. Evidence of 
cryptic species within the pest Copitarsia decolora 
(Guenee) (Lepidoptera: NIoctuidae). Annals of the 
Entomological Society of America 97: 675-680. 

Stephen, W. P. 1956. Notes on the biologies of 
Megachile frigida Smith and M. incrmis Provancher 
(Hymenoptera: Megachlidae). Pan-Pacific Ento- 
mologist 32: 95-101. 

Williams, H. ]., M. R. Strand, G. W. Elzen, S. B. Vinson, 
and S. J. Merritt. 1986. Nesting behavior, nest 
architecture, and use of Dufour's gland lipids in 
nest provisioning by Megachile Integra and M. 
mendica mendica (Hymenoptera: Megachilidae). 
Journal of the Kansas Entomological Society 59: 
588-597. 



J. HYM. RES. 
Vol. 16(1), 2007, pp. 192-205 

A Comparison of Pyrethrum Fogging and Screen-sweep Netting of 
Micro-Hymenoptera in Southern California Chaparral 

Alex R. Van Dam, Matthew H. Van Dam, and John M. Heraty 

Department of Entomology, University of California, Riverside CA 92521, USA; email: 
(AVD) vandama01@yahoo.com; (MVD) mvandam01@hotmail.com; (JMH) john.heraty@ucr.edu 



Abstract. — Three chaparral plant species, Adenostoma fasciculatum Hook, and Arn. (Rosaceae), 
Ceanothus megacarpus Nutt. (Rhamnaceae) and Quercus bcrbcridi folia Liebm. (Fagaceae), were 
sampled for micro-Hymenoptera in the Santa Rosa Plateau Nature Reserve in Southern California. 
Two sampling methods of the shrub's canopy are contrasted: screen-sweep netting and pyrethrum 
fogging. Using both sampling methods and across all of the plant hosts, 242 species of 
Hymenoptera were collected. A total of 558 individuals and 173 species were collected by fogging, 
and 287 individuals and 115 species by screen sweeping. Although fogging captured more 
individuals and species, results were significant only for the number of individuals collected on 
Quercus and number of species on Adenostoma. On the three different plants, fogging sampled 
a similar or greater number of species than did screen sweeping. In terms of estimating species 
richness, fogging had an equivalent or greater efficiency than sweeping for collecting individuals 
and species. When combined with the labor efficiency involved in processing field samples, fogging 
is superior to screen sweeping. However, given the sample sizes within this study, both techniques 
are necessary, with the fogging technique sampling only 71.5% of the total number of species of 
Hymenoptera. 



Hymenoptera are one of the most di- 
verse groups of insects, with approximate- 
ly 115,000 described species and 300,000 to 
2.5 million undescribed species (LaSalle 
and Gauld 1992, 1993, Gauld and Gaston 
1995, Stork 1988, Grissell 1999). Based on 
conservative estimates, more than 10 per- 
cent of all insect species are parasitoids, 
and approximately 75% of these are Hy- 
menoptera (Eggleton & Belshaw 1992). 
Recent estimates for Chalcidoidea alone 
estimate 357,000-400,000 species, of which 
only about 22,000 have been described 
(Noyes 1978, Noyes 2000, Heraty and 
Gates 2003). Parasitic Hymenoptera are 
valuable to agriculture as biological control 
agents (Van Driesche and Bellows 1996) 
and to conservation as a means of measur- 
ing biodiversity and a potential indicator of 
the diversity of lower trophic levels (Kre- 
men et al. 1993, Heraty and Gates 2003). 



Many of these parasitoids are small, 
usually ranging in size from 1-5 mm, and 
difficult to collect except with specialized 
methods (Noyes 1982, Noyes 1989). Vari- 
ous authors have attempted to evaluate the 
best methods to sample these micro-Hy- 
menoptera with an emphasis on both 
numbers of individuals and species (Mas- 
ner and Goulet 1981, Darling and Packer 
1988, Noyes 1982, Noyes 1989, Buffington 
and Redak 1998). 

Many studies have compared collecting 
techniques for Hymenoptera in tropical 
and temperate ecological regions (Hender- 
son and Whitaker 1977, Noyes 1989, 
Gadagkar et al. 1990, Erwin 1995, Hill 
and Cermak 1997, Longino and Colwell 
1997, Stork and Hammond 1997, Hoback et 
al. 1999, Yanoviak 2003), but little has been 
written about sampling in the short, dense, 
scrub vegetation that is typical of Mediter- 



Volume 16, Number 1, 2007 193 

ranean climate zones. Use of an "Allen and a limited amount of debris, but only 
Vac" in a coastal sage scrub plant commu- the latter can be used to selectively sample 
nity in Southern California produced more specific plant hosts, as explained herein, 
individuals and a higher diversity of Canopy fogging techniques were re- 
Hymenoptera than did sweep netting, viewed by Erwin (1989), who noted that 
and was considered more effective because such techniques allow for sampling of 
insects were sampled from deeper within tropical and temperate forest canopies 
the shrub canopy than possible for a sweep more effectively than with other methods 
net (Buffington and Redak 1998). The such as sweep netting. Problems with 
primary disadvantage of vacuum sampling fogging in a tree canopy include collecting 
is the damage caused to small, fragile insects outside of the sampling area, drift 
parasitic Hymenoptera, which can make of specimens from within the sampling 
identifications difficult. Another important area, and the need to collect at dawn, when 
aspect for choosing a particular method is there is no breeze, but perhaps less insect 
the amount of time spent sorting through activity (Erwin 1989, Stork and Hammond 
the accumulated debris to find specimens 1997). Importantly, insects are collected 
(Southwood 1978). The efficiency of vacu- somewhat randomly and can be sampled 
um sampling is counterbalanced by the in replicated samples for a specific area 
labor necessary to sort specimens from the (Stork and Hammond 1997). Insecticide 
accumulated debris collected along with fogging can also be applied to collecting 
the specimens. Unfortunately, sweep net- insects on rough or inaccessible surfaces 
ting and the direct aspiration of minute such as tree trunks (C. Burwell, pers. 
specimens is probably the least efficient comm.) or vertical rock faces (S.B. Peck, 
method of sampling - many specimens pers. comm). In a chaparral vegetation 
may simply escape during collection, avoid community, the issue is not whether 
detection in the accumulated plant debris, sweeping can reach the upper canopy, 
or may not be sampled if they are not but whether sweeping can efficiently and 
readily accessible by the net. Adding thoroughly sample insects from within the 
a metal screen to the net opening to interior of the 'canopy' of dense, often 
exclude debris (Noyes 1982) can increase thorny, bushes. Insecticide fogging of this 
the efficiency of finding specimens, but this miniature tree canopy has a potential for 
can result in greater damage to the speci- sampling a different array of insects in 
mens, and the efficiency of processing will both numbers and species than would be 
depend on whether specimens are aspirat- sampled by beating the exterior of the 
ed (maximizing loss of specimens) or if the shrub with an insect net. Canopy fogging is 
entire sample of specimens and plant also an easily quantifiable method since 
debris is collected into alcohol and later a known surface area of catch basins can be 
sorted in the laboratory (maximizing pro- put underneath the canopy being fogged, 
cessing time). To improve the quality of Canopy fogging in chaparral ecosystems 
specimens and reduce the time spent might also produce samples that are free of 
sorting, new methods for rapid assessment debris or damage unlike screen-sweep 
of Hymenoptera populations, and espe- netting. Another attribute of fogging is that 
daily micro-Hymenoptera, in dense cano- it allows for the collection of specimens 
py situations are needed. Possible solutions from individual plant species like screen- 
include passive collecting techniques such sweep netting. In this paper we hope to test 
as Malaise trapping (indirect method), pan the efficacy of pyrethrum fogging compared 
trapping (indirect), or insecticide fogging to screen-sweep netting in a chaparral 
techniques (direct method). These tech- ecosystem in Southern California for col- 
niques yield fewer damaged specimens lecting parasitic micro-Hymenoptera. 



194 



Journal of Hymenoptera Research 







Fig. 1. Fogging method used to collect Hymenoptera from a Quercus berberidifolia bush in Southern California. 



MATERIALS AND METHODS 

Location and date. — Sampling took place 
on the two dates July 11, 2001 and July 18, 
2001 at 3 adjacent sites in the Santa Rosa 
Nature Preserve, in Riverside County, 
California, at 33 31'N 117'14'W and 
590 m elevation. We chose 3 similar stands 
of dense chaparral over a 5-acre area: one 
stand adjacent to a field of endemic bunch 
grass, the second adjacent to a road bor- 
dered by invasive grass, and the third in 
the heart of a dense stand of chaparral. We 
used both fogging and screen sweeping to 
collect from 3 individual bushes from 3 
dominant plant species on two dates: 
Adenostoma fasciculatum Hook, and Arn. 
(Rosaceae), Ceanothus megacarpus Nutt. 
(Rhamnaceae), and Quercus berberidifolia 
Liebm. (Fagaceae). 

Screen-sweeping. — We used a triangular 
net hoop with 38-cm sides and a recessed 
covering of 6.4-mm hardware cloth to 
exclude large debris. The sweep net was 
a fine-meshed net bag from Bioquip (Gar- 
dena, CA), with the apex of the net bag 



open and held closed by a twist tie that 
could be removed to empty the contents 
into a 1 -quart plastic Ziploc® bag contain- 
ing 80% EtOH. The contents of the Ziploc® 
bag were rinsed with additional 80% EtOH 
to kill and preserve the insects. Each bush 
was swept over all its of the surfaces by 
a single collector (John Pinto, UCR) to keep 
the sampling as uniform as possible. 
Sampling of all 3 sites took about 45 min 
on each date. 

Fogging. — The insecticide fogging of 
shrubs required several steps on the two 
dates. First, 36 yellow Dixie® bowls (total 
area = 1 m 2 ) were placed underneath the 
canopy of each bush to be sampled. Each 
bush was sprayed with Raid Yard Guard® 
for 1 min from a distance of about 1.5 m, 
enveloping the bush in a fine fog with no 
visible droplets on the leaves (Fig. 1). 
Approximately one spray can (473 ml; 
16 fl oz) was used for 3 bushes. After 
5 minutes, pans were emptied and rinsed 
with 80% EtOH into one gallon plastic 
Ziploc® bags. Fogging took approximately 



Volume 16, Number 1, 2007 



195 



2 hours to complete from setup to finish 
for all 3 sites. The air movement was 
minimal during sampling periods and 
was not considered to have impacted 
specimen drift. 

Processing samples. — Two strainers with 
square mesh openings of 3.2 and 1.6 mm 
were used to separate the screen-sweep 
samples into course, medium and fine 
debris samples. Because of the lack of 
debris, fogging samples were directly 
sorted without screening. Each sample 
was sorted with use of an llxll-cm Rose 
Entomology® sorting tray with parallel 
sorting lanes separated by raised ridges 
13-mm apart. To ensure that all specimens 
were discovered in the samples, each tray 
was sorted twice, and in some screen- 
sweep samples three times. Specimens 
were transferred to small glass vials and 
then dried for mounting by use of the 
Hexamethyldisilizane (HMDS) technique 
(Heraty and Hawks 1998) and then card 
mounted (Noyes 1982). All mounted speci- 
mens were individually labeled with col- 
lection information and a unique specimen 
identifier number. Data were input into 
a Filemaker® database for the UCR Ento- 
mology Research Museum, where all ma- 
terial was deposited. All Hymenoptera 
were identified to family, genus, and 
morphological species groups using avail- 
able identification keys. Certain groups 
were identified, or our identifications 
verified, by other local specialists at UCR: 
Mymaridae identified by Serguei Triapit- 
syn, Trichogrammatidae by John Pinto, 
Pteromalidae and Eulophidae by Roger 
Burks, Signiphoridae by James Munro, 
Figitidae by Matthew Buffington, and 
Aphelinidae by Jung-Wook Kim. 

Data analysis. — ANOVA analysis re- 
vealed no significant difference in speci- 
mens collected between sampling dates 
(p<0.05), so we pooled the data for the 
remaining analysis. A 2-tailed Student's t- 
test was used to compare the number of 
individuals collected by fogging and 
sweep netting for each family (MendenhaU 



et al. 2003), the number of individuals in 
the two higher taxonomic groups (Chalci- 
doidea and non-Chalcidoidea) and families 
of Hymenoptera (Tables 1-3). 

The ecological modeling program Esti- 
mates version 7.0 (Colwell 2004) was used 
to compare the two methods for species 
richness (Figs 1-3) and similarity of shared 
species (Table 4) for Chalcidoidea and non- 
Chalcidoid micro-Hymenoptera by plant 
species. The diversity settings for Esti- 
mates 7.0 were set to sample with re- 
placement and the number of replications 
set to 1,000 to calculate the Chaol richness 
estimator (Chao 1984), Sobs estimator 
(Colwell 2004) and singletons estimator 
(Chazdon et al. 1998, Colwell and Cod- 
dington 1994) (Figs 2-7). The advantage of 
estimating the diversity by selection of 
samples with replacement is that estimator 
variance remains meaningful at the right 
hand end of the accumulation curve, and 
thus can be used to compare data sets 
(Colwell 2004). Scatter plots of these 
estimator values of species were plotted 
against the estimated number of individu- 
als observed in pooled quadrat samples to 
construct models (Figs 2-7). The Chaol 
species estimator is used for the sampling 
history of species represented by at least 
two individuals (Magurran 2004). The Sobs 
estimator estimates sampling of the mean 
number of new species collected among 
the samples (Colwell 2004). The Singletons 
estimator estimates sampling of the mean 
number of new species represented only by 
one individual (Colwell and Coddington 
1994), and thus is a rough estimate of the 
number of rare species. Accumulation 
curves from each of these estimators >:^n 
be used to compare the relative efficiency 
of fogging and screen sweeping in captur- 
ing species diversity (Figs 2-7). 

A Morisita-Horn species similarity index 
was calculated with use of standard de- 
fault settings of Estimates to compare 
the number of shared species collected on 
their respective plant species with fogging 
and screen sweeping after correcting for 



196 



Journal of Hymenoptera Research 



U3 



(J 
01 

bo 
C 



01 

e 

ai 

OJ 

in 

o 

60 

g 
'3d 

o 

X 

a 

u 
_oj 

'o 
o 
en 
a; 

'G 

0) 

en 

o 



-a 
c 



01 

Oh 

o 
C 

01 

i 

x 

o 



T3 

0> 

u 

_OI 

"o 

d 

c 

-a 
c 

re 
O) 
60 

f3 



w 




en 




— - 




j- 




o 




i-H 




H 




a; 




T3 

>h 


td 




E 


C 


o 


ro 


4H 




' ~* 


<n 


ia 


+ 1 


U 


OJ 


cu 


X 




e 


0) 
en 


2 


01 


C 


c4 


C 


ns 


« 


en 


01 


O 


s 


Pi 




n 






T— < 


C 
3 


0) 


en 






X 


01 


ia 


£ 


H 





- 
-r 



r-l is. 

•V NO 



- co m 



ts 

I 

o 



I 

(N 



LO CN 
IS. 00 



1U 

»H 


o 


O 


o 


<u 


+1 


+ 1 


+ 1 


~E_ 


, — 1 


IS. 


o 


o 


no 


no 


o 


r^ 


CO 


CO 


NO 



o> 

B 

>^ 



o 
H 



, — . i— l NO 

N ro in 

<N s .. 

.00 H 

m ^h i— i 

i ! i 

o o o 



^ oo in 

f) ON \q 

o o o 

+ 1 



NO On 

o o o 



7 



+1 

00 Is 
Is. 



00 NO 

is. in 



o 



0) 

-a 
'o 

"2 

"0 

X 

U 



I 



(N 00 

<-i CN 

I I 

o a 



in m 

IT, is 



-r 
o 



m in is. 

in ■<* 

00 ON CN 

I I X 

r+ o o 



n a. in 
in in in 



03 
01 

12 

"S 



u 

c 

o 
c 



00 


ON 


NO 


.*-» 


co 


^ 


CN 


3 


O 


o 


O 


X 


+ 1 


+ 1 


+ 1 


"O 


1— I 


IS. 


NO 


01 


o 


NO 


in 


"5 



^ CO 

i— i co 



NO 

cn 
f f 

l-H O 



ND 00 LTj 

co cn ts 



O O CN 

+1 +1 +1 

00 ON ON 

CN CO CO 

no in o 



o 

+1 

IS. 



+1 



o 
+1 



+1 



CN 00 
ts CN 

CO IS 



co oo in 

°°. ^ ro 

CM <— i 



^ ^ CO 
CN CN CN 

00 00 ^-i 

tN 00 N 



I I 

o o 



CO 



^ ON M 
CO NO ON 



IS. CN 

CN i—i 

I I 

o o 



ts 

I 
o 



ON CO ^ 

CO i— 1 i— 1 

v . CO 

o o - 

l-H i — I NO 

I I ± 

o o o 



ON 

l-H 

o 

+1 



O nO 
cn on 

o 

+1 



on in 

o • 



+1 

CN ON 

IS 00 

-r no 



>N 

X 



o 
H 



CO 
01 

-d 

• r-i 

o 

T3 



re 
X 



m 

o i-i 

+ 1 +1 

ON CO 1—1 

CO 00 i-h 

cn co in 



ts 

o 

+1 



oo in 
in ^h 



o 
+ 1 



ca 

01 



o 

+ 1 

IS. NO CN 

in ts 



o 
+ ! 



73 CN O 

'3 
-a 



C3 

x: 
U 



no ts m 

NO i— i ts. 
v— ^h CN 


u 


CN O 

o — o 

i— O CN 

. 00 „ 


C 
o 
c 


^ ts IS. 

NO CO 


oo" oo no" 

rH i-H ■** 

1 1 1 

CO O O 




O .NO 

i—l ON CO 
1 1 1 




. . in 

T T 7 




1 1 
^H O O 




^h O O 



00 

o 
+1 



•-f 

CO 

+1 



no oo in 

NO oo NO 



+ 1 

NO oo 

h in n 

On NO in 



o 

+1 



CN 

+ 1 



^h m o 

in co o 

o ^-< 

+1 +1 



o 

+1 



ts -r 

no -r 



NO NO IS 

in o i—i 



CO CN ** 



C 

C3 
U 



&0 



O o ^J 

5 S 51 



o 
R 



Uj 



o 

Oh 

Oi 
■*-f 

-a 

CN 

01 



c 

01 
3i 



01 
Xj 









« en 

° o 

60 _c 








+-> 4_, 








r 3 c 








^ <3 








ro a 








■S c 








X 0) 








.ti ii 
















4-* Q> 


Q 






C X 


^ 


tn 


r ^j 


"3 i, 

.y c 


Ui 

^ 
^ 


5 
U 


S3 


' Signif 
differe 



c 



& 



O 
C 

OI 
u 
01 



0) 

o 

60 
OJ 

u 



ns 
en 

OI 
X 



en 

01 
j3 

> 

Cn in 

o o 

d d 

V V 

O- Oh 

tn <! 

a > 

c< 

Oi kri 



tn 



U 



Volume 16, Number 1, 2007 



197 



Table 2. Number of species collected by each sampling method (pooled) and average number of individuals 
(x ± SE) of Chalcidoidea sampled at the Santa Rosa Reserve. 



Number of species 



\tt\m no. of indi\ idu.iU 



Fogging Sweeping Both 



logging 



Sweeping 



Aphelinidae 

Adenostoma 

Ceanothns 

Quercus 

Chalcididae 

Adenostoma 

Ceanothus 

Quercus 

Encyrtidae 

Adenostoma 

Ceanothus 

Quercus 

Eulophidae 

Adenostoma 

Ceanothus 

Quercus 

Eupelmidae 

Adenostoma 

Ceanothus 

Quercus 

Eurytomidae 

Adenostoma 

Ceanothus 

Quercus 

Mymaridae 

Adenostoma 

Ceanothus 

Quercus 

Ormyridae 

Adenostoma 

Ceanothus 

Quercus 

Pteromalidae 

Adenostoma 

Ceanothus 

Quercus 

Signiphoridae 

Adenostoma 

Ceanothus 

Quercus 

Torymidae 

Adenostoma 

Ceanothus 

Quercus 

Trichogrammmatidae 

Adenostoma 

Ceanothus 

Quercus 

Chalcidoidea 

Adenostoma 

Ceanothus 

Quercus 



13 



23 



29 



14 



92 



25 



56 



11 



38 



0.90 ± 0.20* (16) 
0.61 ± 0.23* (11) 
1.28 ± 0.40* (23) 


0.33 ±0.18* (6) 
0.28 ± 0.14* (5) 
0.50 ± 0.17* (9) 




0.06 ± 0.06 (1) 









0.67 ± 0.21 (12) 
0.61 ± 0.22 (11) 
0.50 ± 0.33 (9) 


0.28 ± 0.18 (5) 
0.33 ± 0.14 (6) 
0.17 ± 0.09 (3) 


1.24 ± 0.29 (30) 
0.67 ± 0.23 (12) 
1.50 ± 0.50 (27) 


0.78 ± 0.25 (14) 
0.50 ± 0.23 (9) 
1.33 ± 0.31 (24) 



0.11 ± 0.08 (2) 
0.06 ± 0.06 (1) 



0.06 ± 0.06 ( 1 ) 
0.22 ± 0.02 (4) 


0.06 ± 0.06 (1) 


0.06 ± 0.06 (1) 


0.06 ± 0.06 (1) 


0.06 ± 0.06 ( 1 ) 


0.61 ± 0.18 (11) 
0.22 ±0.10 (4) 
1.17 ± 0.40 (21) 


0.39 ± 0.14 (7) 
0.33 ± 0.14 (6) 
1.00 ± 0.29 (18) 






0.11 ± 0.08 (2) 





11.11 ± 0.08 (2) 


0.39 ± 0.14* (7) 
0.28 ± 0.14* (5) 
3.61 ± 1.19* (65) 


0.11 ± 0.08* (2) 

0* 
0.67 ± 0.21* (12) 


0.11 ± 0.08 (2) 
0.28 ± 0.11 (5) 
0.89 ± 0.77 (34) 



0.44 ± 0.15 (8) 
0.78 ± 0.29 (14) 






0.06 ± 0.06 ( 1 ) 


0.11 ± 0.08 (2) 
0.06 ± 0.06 (1) 
0.06 ± 0.06 (1) 


1.28 ± 0.39* (23) 
1.61 ± 0.45 (29) 
0.89 ± 0.30* (16) 


0.33 ± 0.18* (6) 
1.83 ± 0.64 (33) 
0.22 ± 0.10* (4) 


5.67 ± 0.66* (102) 

4.44 ± 0.88 (80) 

11.11 ± 2.65* (200) 


2.39 ± 0.51* (43) 
3.83 ± 1.09 (69) 
5.11 ± 0.75* (92) 



198 



Journal of Hymenoptera Research 



Table 3. Number of species collected by each sampling method (pooled) and average number of individuals 
(x + SE) of non-chalcidoid Hymenoptera sampled at the Santa Rosa Reserve. 



Number of species 



Mean no. of individuals 



Fogging Sweeping 



Both 



Fogging 



Sweeping 



Bethylidae 

Adenostoma 

Ceanothus 

Quercus 

Braconidae 

Adenostoma 

Ceanothus 

Quercus 

Ceraphronidae 

Adenostoma 

Ceanothus 

Quercus 

Crabronidae 

Adenostoma 

Ceanothus 

Quercus 

Figitidae 

Adenostoma 

Ceanothus 

Quercus 

Diapriidae 

Adenostoma 

Ceanothus 

Quercus 

Dryinidae 

Adenostoma 

Ceanothus 

Quercus 

Formicidae 

Adenostoma 

Ceanothus 

Quercus 

Ichneumonidae 

Adenostoma 

Ceanothus 

Quercus 

Platygastridae 

Adenostoma 

Ceanothus 

Quercus 

Scelionidae 

Adenostoma 

Ceanothus 

Quercus 

Sphecidae 

Adenostoma 

Ceanothus 

Quercus 

non-chalcidoid Hymenoptera 

Adenostoma 

Ceanothus 

Quercus 











35 















11 



















13 




0.06 ± 0.06 (1) 
0.17 ± 0.09* (3) 

0.33 ± 0.19* (6) 
0.50 ± 0.22 (9) 
0.33 ± 0.14* (6) 

0.44 ± 0.18 (8) 
0.06 ± 0.06 (1) 
0.17 ± 0.12 (3) 





0.06 ± 0.06 ( 1 ) 










0.06 ± 0.06 (1) 

0.11 ± 0.08 (2) 



0.56 ± 0.17* (10) 
0.44 ± 0.15* (8) 
1.28 ± 0.39 (23) 





0.06 ± 0.06 (1) 

0.06 ± 0.06 (1) 


0.11 ± 0.08 (2) 

2.00 ± 0.46 (36) 
1.00 ± 0.30* (18) 
1.77 ± 0.55* (32) 

0.06 ± 0.06 (1) 



3.56 ± 0.51 (64) 
2.06 ± 0.35* (37) 
4.17 ± 1.00* (75) 





0* 

0* 

0.06 ± 0.06 (1) 

0* 

0.39 ± 0.16 (7) 
0.06 ± 0.06 (1) 
0.06 ± 0.06 (1) 










0.17 ± 0.09 (3) 0.06 ± 0.06 (1) 













0* 

0* 

0.78 ± 0.13 (14) 







0.28 ± 0.18 (5) 
0.06 ± 0.06 (1) 
0.33 ± 0.20 (6) 

1.50 ± 0.41 (27) 
0.27 ±0.11* (5) 
0.50 ± 0.12* (9) 



0.11 ± 0.08 (2) 



2.17 ± 0.58 (39) 
0.56 ± 0.15* (10) 
1.72 ± 0.34* (31) 



Volume 16, Number 1, 2007 



199 



Chalcidoidea 



non-Chalcidoidea 



S>5 

■8 



70 



60 



50 
(A 

<D 40 

O 
0) 

Q. 30 

CO 



20 



10 - 




{{'fcfrffii 



DO 



an 



,n 



oooooooooo 



20 40 60 80 

Individuals 



100 



30 



25 -I 



20 
W 
Q) 

O 15 

Q. 
0) 

10 




mum 

-L _ n -O 



jOO-OO-OOO-OO-Oo 



10 20 30 40 50 60 



Individuals 



05 

S 

3 



70 



60 



50 



w 

0) 40 

u 
a) 

Q. 30 

20 



JpSHffi 




DO 



^OOOOOOOOOO 



10 20 30 40 50 60 70 80 90 

Individuals 




°o°oooooooo 



10 20 30 40 

Individuals 



80 n 

70 

60 





tn 50 




0) 


3 


8 40 


O 


Q. 


Jk 


CO 30 


^ 




a 




01 


20 




Ififi^Jf 



000000000 



30 60 90 120 150 180 210 

Individuals 



35 

^0 



V> 25 
0) 

a 
co 15 



:ggflffi 



f 



{■ 



D 




boooooooooo 



20 40 60 

Individuals 



80 



Figs 2-7. Species accumulation curves generated by use of Estimates 7.0 for the pooled togging (open 
symbols) and sweep-netting (closed symbols) data. Triangles represent the Chaol estimator, squares the Sobs 
estimator, and circles the singletons estimator. 



200 



Journal of Hymenoptera Research 



Table 4. Morisita-Horn indices for shared species 
of micro-Hymenoptera collected at the Santa Rosa 
Plateau Reserve by canopy fogging and screen netting 
by plant host. 





Ch 


alcidoidea 


Non 


-Chalcidoidea 


Adenostoma 




0.69 




0.73 


Ceanothus 




0.84 




0.59 


Quercus 




0.75 




0.63 


Total 




0.84 




0.77 



sample size (Table 4). The Morisita-Horn 
index has been shown to be robust and 
more reliable than other shared species 
indices because it is not strongly influ- 
enced by species richness and sample size 
(Wolda 1981, Magurran 2004). 

RESULTS 

Fogging of the plant canopy consistently 
collected more individuals and species 
than screen sweeping across the 3 types 
of plants sampled (Table 1). A total of 558 
micro-Hymenoptera specimens were col- 
lected by fogging as compared to 287 by 
screen sweeping. Samples were significant- 
ly different only for the number of indi- 
viduals of all micro-Hymenoptera from 
Quercus and the number of species from 
Adenostoma (Table 1). More individuals of 
Chalcidoidea (Table 2) and non-Chalcidoid 
micro-Hymenoptera (Table 3) were collect- 
ed with fogging across all plants sampled. 
Use of fogging produced significantly 
higher numbers of Chalcidoidea on Ade- 
nostoma and Quercus (Table 2) and non- 
Chalcidoid micro-Hymenoptera on Ceano- 
thus and Quercus (Table 3) than sweeping. 
In virtually all of the family level compar- 
isons, use of fogging produced a greater 
number of individuals than did sweeping 
(Tables 2, 3). However, the differences 
were significant for only some Chalcidoi- 
dea (Aphelinidae, Pteromalidae and 
Trichogrammatidae) and other Hymenop- 
tera (Bethylidae, Braconidae, Formicidae, 
Scelionidae). Except for Pteromalidae 
on Quercus, Aphelinidae and Tricho- 
grammatidae were the most commonly 



sampled wasps. Individuals of these fam- 
ilies are minute and are likely most 
common in the interior canopy of the 
bushes where they attack sessile Hemi- 
ptera and eggs of various insects. 

A total of 242 species of Hymenoptera 
were collected across all of the samples, 
with Chalcidoidea represented by 186 
species (76.9%). For all species (micro- 
Hymenoptera, Chalcidoidea and non-Chal- 
cidoidea), fogging consistently produced 
more species, on average, than screen- 
sweeping (Table 1) (173 versus 115 total 
species collected by each technique) and 
significantly more species on Adenostoma 
(Table 1). Significantly more species of 
Chalcidoidea were collected on Adenostoma 
and Quercus, whereas more species of non- 
Chalcidoidea were collected only on Quer- 
cus (Table 1). Each method did not always 
collect the same species. Only 38 species of 
Chalcidoidea and 13 species of non-Chal- 
cidoid micro-Hymenoptera were sampled 
by both methods (Tables 2 and 3). Sweep 
netting collected an additional 56 species of 
Chalcidoidea (30.1%) and 8 species of non- 
Chalcidoid micro-Hymenoptera (14.3%) 
(Tables 2 and 3). As expected, because of 
the greater number of specimens collected 
fogging collected a larger proportion of 
additional species that were not collected 
by screen sweeping (92 species or 49.5% of 
Chalcidoidea; 35 species or 62.5% of non- 
Chalcidoid micro-Hymenoptera). When 
considering the unique species collected 
by a particular method, along with the 
species collected by both methods, fogging 
sampled 69.9% of the species of Chalcidoi- 
dea whereas screen sweeping sampled 
50.5% of the species, and respectively 
85.7% and 37.5% of the non-Chalcidoid 
micro-Hymenoptera. Because of different 
sample sizes obtained from each method 
(significantly more for fogging), an un- 
biased Morisita-Horn analysis estimated 
that the two sampling methods produced 
samples with 69-84% shared species of 
Chalcidoidea and 59-77% shared species of 
non-Chalcidoid micro-Hymenoptera (Ta- 



Volume 16, Number 1, 2007 



201 



ble 4). However, especially for Chalcidoi- 
dea, screen sweeping collected a large 
number of unique specimens (56) despite 
the low sample size. 

Of the three estimators, Chaol provides 
an indication of the ability to sample the 
species thoroughly (more than two indi- 
viduals of each species sampled), Sobs 
focuses on the accumulation of new spe- 
cies, and the singletons estimator is the 
accumulation of species based only on 
a single specimen. Only the singletons 
estimator is expected to decline as a habitat 
is more thoroughly sampled and species 
are shifted to the Chaol category. The 
Chaol and Sobs estimates should both 
plateau as the number of species becomes 
thoroughly sampled. In all cases, estimates 
for fogging were consistently based on 
a sample with greater number of individ- 
uals (Figs 2-7; Table 1). Results for the 
Chaol estimator species accumulation 
curve had the number of 'common' species 
both accumulate and also reach a plateau 
at a significantly faster rate using the 
fogging technique for most of the data 
partitions (Figs 2-4, 7), whereas screen 
sweeping accumulated common species at 
a faster rate for non-Chalcidoidea on 
Ceanothus (Fig. 5) and Chalcidoids on 
Quercus (Fig. 6). In these latter two cases, 
fogging still sampled more species overall 
on Quercus (71 versus 56), whereas the 
same number of species of non-Chalcidoi- 
dea (11) were sampled on Ceanothus and in 
neither case did the number of species 
appear to plateau (Figs 5, 6; Table 1). Thus 
fogging will generally sample the highest 
and best represented diversity of common 
species with the least effort, as based on the 
number of specimens collected. The mean 
number of new species accumulated (Sobs 
estimate) was virtually the same for Chal- 
cidoidea using both methods (Figs 2, 4, 6), 
and for the non-Chalcidoid micro-Hyme- 
noptera, slightly higher on Ceanothus 
(Fig. 5) or lower on Adenostema and Quer- 
cus (Figs 3, 7). The number of species 
represented by a single specimen (single- 



ton) accumulated at a slightly faster rate in 
most of the fogging samples (Figs 2-A, 7), 
but were roughly the same for the Ceano- 
thus non-Chalcidoidea (Fig. 5) and Quercus 
Chalcidoidea (Fig. 6). Only the non-Chal- 
cidoid micro-Hymenoptera on Adenostema 
(Fig. 5) demonstrated a decline in the 
number of singletons, suggesting overall 
that the maximum number of species had 
been sampled even though the species 
accumulation curves (Chaol and Sobs) 
had not yet reached a plateau. 

DISCUSSION 

Insecticide fogging of tree canopies has 
been experimented with since the late 
1960's (Martin 1966, Gange and Martin 
1968, Roberts 1973, Erwin and Scott 1980, 
Erwin 1983, Adis et al. 1984, Stork and 
Hammond 1997). Typical canopy fogging 
in the tropics is used to access the forest 
canopy 30-60 m above the ground (Erwin 
1983, Stork and Hammond 1997). Here we 
suggest that the canopy of dense thorny 
shrubs in chaparral habitat can present 
some of the same problems of sampling, 
but on a much smaller scale. The fogging 
strategy employed in this paper has 
a number of advantages in: 1) relying 
upon compact and inexpensive equipment 
that can be carried easily to the field, 2) 
the sampling area can be defined by the 
collecting surface under the plant, 3) 
a specific bush or species of plant can be 
targeted, 4) debris is minimized and the 
specimens can be quickly and efficiently 
processed, 5) there is very minimal, if any, 
damage to specimens, and 6) there is no 
damage to the plants being sampled, 
which may be a factor in some conserva- 
tion studies. Our method draws many 
parallels with the typical tropical forest 
canopy fogging as in Stork and Hammond 
(1997), and faces similar issues of speci- 
men drift within and outside of the 
sampling area, but on a less dramatic 
scale. Climatic conditions (i.e. wind) re- 
mains an important factor, but can be 
monitored and controlled throughout the 



202 



Journal of Hymenoptera Research 



sampling period, and sampling can be 
done during presumed periods of peak 
insect activity. Typical chaparral shrubs 
stand waist high and thus access to the 
canopy is not a problem, and fogging of 
chaparral or similar shrub canopies may 
allow access to this seldom collected 
niche. 

Noyes (1989) demonstrated variable re- 
sults when comparing sweep netting to 
canopy fogging of trees in the tropical 
forests of Sulawesi, but did not speculate as 
to which was more effective at collecting 
parasitic Hymenoptera. Noyes (1989) ar- 
gues that each method of collecting will 
have its own advantages over another, but 
this may relate to sampling different 
ecological niches, more than the overall 
efficiency of collecting the same niche. We 
observed this within our study, in which 
fogging sampled only 71.5% of the Hyme- 
noptera and screen sweeping sampling 
only 47.5%. A large number of species 
were represented by only one or two 
specimens, and the differential sampling 
may be due to a different distribution of 
species on the individual host plants being 
sampled. The only way to account for this 
would be to increase the number of plant 
hosts being sampled in order to decrease 
the variance in species being sampled; 
however, this would dramatically increase 
the effort for sampling with the screen 
sweep method. 

In this study, insecticide fogging sam- 
pled a greater or equivalent number of 
individuals and species of micro-Hyme- 
noptera as compared to sweep netting in 
a chaparral ecosystem (Tables 1-3, Figs 2- 
7). Similar to vacuum sampling, the 
difference was likely because of greater 
access to wasps within the interior shrub 
canopy (Dietrick et al. 1960, Buffington 
and Redak 1998). Sweep netting generally 
samples insects from the tops and sides 
of the shrub canopy (Southwood 1978, 
Buffington and Redak 1998). Differences 
in the shrub architecture may have led to 
some of the variability in the effectiveness 



of fogging versus screen-sweeping (Ta- 
ble 1, Figs 2-7). Both Quercus berberidi folia 
and Adenostoma fasciculatum have dense 
overhanging canopies that the screen- 
sweep net could not penetrate. However, 
Ceanothus megacarpus has a sparse wil- 
lowy canopy architecture and the screen- 
sweep net could be used to sample most 
of the canopy. Thus, when sampling 
dense chaparral shrubs, canopy fogging 
would have an advantage over screen 
sweeping at capturing a greater diversity 
of micro-Hymenoptera. When sampling 
open shrubs, no difference in the wasps 
being sampled by either method is 
expected. 

Insecticide fogging, coupled with the 
collection of specimens into pans of a de- 
fined size, allowed for better quantification 
of the capture of wasps in a defined area, 
with 1 m 2 being the combined area of the 
pan traps placed under each shrub. This is 
somewhat similar to the multiple 1 m 2 
funnel sampling method employed in 
canopy fogging of tree canopies (Stork 
and Hammond 1997), although we did 
not treat each pan as a separate sampling 
unit because of the expected low sample 
size. Sampling by screen sweeping is more 
arbitrary, being based on the number of 
sweeps using an undefined arc, velocity, 
and the area sampled (Southwood 1978). It 
is possible to define the area sampled 
through screen sweeping by the size of 
the shrubs being sampled, which in this 
study certainly had a surface area greater 
than 1 m 2 , but each shrub varied sub- 
stantially in size. Other factors that miti- 
gate against screen-sweeping are collector 
bias in sweeping efficiency and potential 
damage to the host plants by intensive 
sweep netting. 

The efficiency of processing samples is 
an important factor. More time was spent 
in the field setting the pans under each 
shrub, fogging the canopy, and collecting 
specimens from the pans. However, the 
fogging technique produced samples al- 
most entirely free of debris, which allowed 



Volume 16, Number 1, 2007 



203 



for specimens to be easily located and 
processed. Fogging could theoretically al- 
low for more samples to be taken, which 
overall is the best way reduce the variance 
in samples from natural habitat (South- 
wood 1978). 

It is difficult to compare trapping meth- 
ods directly for numbers of individuals 
and species when, because of the method, 
they are not comparable for a similar 
investment of effort. Modeling of trap 
catches through various resampling meth- 
ods allows for an estimate of whether the 
diversity and quota of specimens can reach 
the same asymptote, the relative efficiency 
of reaching that value, and whether a par- 
ticular method has already reached that 
estimated value. In almost all cases, fog- 
ging was estimated to collect more species 
and at a faster rate than sweep netting 
(Figs 2-3, 5-7). Only on Ceanothus was the 
diversity of non-Chalcidoidea estimated to 
be equal and the number of species 
accumulated at a faster rate with screen 
sweeping (Fig. 5). The upright growth and 
open canopy of Ceanothus may allow for an 
equal number of individuals and species to 
be sampled by both methods. The Mor- 
isita-Horn shared species index (Table 4) 
indicates that the use of both fogging and 
screen sweeping sampled similar species of 
Hymenoptera (59%-84% similarity), with 
no bias in groups. Thus, when corrected for 
sample size either method would sample 
approximately the same groups of species 
in a chaparral ecosystem. 

The goals of sampling parasitic Hyme- 
noptera in different habitats are endless. 
Here we were interested in sampling 
numbers of individuals and species from 
isolated plants in a dense shrub canopy in 
chaparral habitat at a single point in time. 
This is a diverse ecosystem, with 242 
species collected on only two sample dates. 
The same or more individuals were sam- 
pled from each plant using fogging as 
compared to screen sweeping. In terms of 
specimen quality, efficiency and quantifi- 
cation, insecticide fogging, with collection 



of specimens into circular pans placed 
under the shrub canopy, is a superior 
technique over both screen sweeping and 
vacuum sampling. 

ACKNOWLEDGMENTS 

We thank Matt Buffington, Albert Owen, Jeremiah 
George, James Munro, John Pinto (UCR), and Jung 
Wook Kim (North Carolina State University) for their 
assistance with setup and running of the field 
experiments. Doug Yanega helped establish the 
specimen database. Serguei Triapitsyn, Roger Burks, 
Matt Buffington and John Pinto all helped to identifj 
and verify specimen identifications across a variety of 
groups. We also thank David Hawks (UCR) helped 
with specimen preparation and instruction. Matt 
Buffington, Mark Shaw, Gavin Broad and an anony- 
mous reviewer provided valuable comments on the 
manuscript. We thank Robert K. Colwell (Universitj 
of Connecticut) for answering questions about species 
similarity indices. This research was supported by an 
NSF PEET grant BSR-9978150 to JMH and John Pinto. 

LITERATURE CITED 

Adis, J., Y. D. Lubin, and G. G. Montgomery. L984. 
Arthropods from the canopy of inundated and 
terra firme forests near Manaus, Brazil, with 
critical considerations on the pyrethrum-fogging 
technique. Studies on the Neotropical Fauna and 
Environment 19: 223-236. 

Buffington, M. L. and R. A. Redak. 1998. A compar- 
ison of vacuum sampling versus sweep-netting 
for arthropod biodiversity measurements in 
California costal sage scrub, journal of Insei I 
Conservation 2: 99-106. 

Chao, A. 1984. Non-parametric estimation of the 
number of classes in a population. Scandinavian 
journal of Statistics 11: 265-270. 

Chazdon, R. L, R. K. Colwell, J. S. Denslow, and M R. 
Guariguata. 1998. Statistical methods for estimat- 
ing species richness of woody regeneration in 
primary and secondary rain forests of NE Costa 
Rica. Pp. 285-309 in: Dallmeier, \., and J. A. 
Comiskey, eds. Forest biodiversity research, moni- 
toring and modeling: Conceptual background and ( )ld 
World case studies. Parthenon Publishing Pans 
671 pp. 

Colwell, R. K. 2004. l.stimatcS: Statistical estimation of 
species richness and shared species from samples. 
Version 7. User's Guide and application published 
at: http://purl.ock.org/estimates. 

and J. A. Coddington. 1994. Estimating 

terrestrial biodiversity through extrapolation. 
Philosophical Transactions of the Royal Society 
(Series B) 345: 101-118. 



204 



Journal of Hymenoptera Research 



Darling, D. C. and L. Packer. 1988. Effectiveness of 
malaise traps in collecting Hymenoptera the 
influence of trap design mesh size and location. 
Canadian Entomologist 120: 787-796. 

Dietrick, E. J., E. I. Schlingei, and M. J. Garber. 1960. 
Vaccuum cleaner principle applied in sampling 
insect populations in alfalfa fields by new 
machine method. California Agriculture 14: 9-11. 

Eggleton, P. and R. Belshaw. 1992. Insect parasitoids: 
and evolutionary overview. Philosophical Transac- 
tions: Biological Sciences 337: 1-20. 

Erwin, T. L. and J. C. Scott. 1980. Seasonal and size 
patterns, trophic structure, and richness of Co- 
leoptera in the tropical arboreal ecosystem: the 
fauna of the tree Luehea seemannii Triana and 
Planch in the Canal Zone of Panama. Coleopterists 
Bulletin 34: 305-322. 

. 1983. Tropical forest canopies: the last biotic 

frontier. Bulletin of the Entomological Society of 
America 29: 14-19. 

. 1989. Canopy arthropod biodiversity: a chro- 
nology of sampling techniques and results. 
Revista Peruana de Entomologia 32: 71-77. 

1995. Measuring arthropod biodiversity in the 



tropical forest canopy. Pp. 109-127 in: Lowman, 
M. D., and N. M. Nadkarni, eds. Forest Canopies. 
California: Academic Press, San Diego. 624 pp. 

Gadagkar, R., K. Chandrashekara, and P. Nair. 1990. 
Insect species diversity in the tropics: sampling 
methods and a case study. Journal of the Bombay 
Natural History Society 87: 337-353. 

Gange, W. C. and J. L. Martin. 1968. The insect ecology 
of red pine plantations in Central Ontario. V. The 
Coccinellidae (Coleoptera). Canadian Entomologist 
100: 835-846. 

Gauld, I. D. and K. J. Gaston. 1995. The Costa Rican 
Hymenoptera fauna. Pp. 40-55 in: Hanson, P., and 
I. D. Gauld eds. The Hymenoptera of Costa Rica. 
Oxford, United Kingdom: Oxford University Press. 
893 pp. 

Grissell, E. E. 1999. Hymenoptera biodiversity: Some 
alien notions. American Entomologist 45: 235-244. 

Henderson, I. F. and T. M. Whitaker. 1977. The 
efficiency of an insect suction sampler in grass- 
land. Ecological Entomologi/ 2: 57-60. 

Heraty, J. M. and M. Gates. 2004. Diversity of 
Chalcidoidea (Hymenoptera) at El Eden Ecological 
Reserve, Mexico. Pp. 227-292 in: Gomez-Pompa, 
A., M. F. Allen, S. L. Fedick, and ]. J. Jimenez- 
Osornio eds. The Lowland Maya Area, Three Millen- 
nia at the Human-Wildland Interface. The Haworth 
Press Inc, Binghamton, NY, USA. 659 pp. 

and D. Hawks. 1998. Hexamethyldisilazane: A 

chemical alternative for drying insects. Entomo- 
logical News 109: 369-374. 

Hill, C. J. and M. Cermak. 1997. A new design and 
some preliminary results for a flight intercept 
trap to sample forest canopy arthropods. Austra- 
lian Journal of Entomology 36: 51-55. 



Hoback, W. W., T. M. Svatos, S. M. Spomer, and L. G. 
Higley. 1999. Trap color and placement affects 
estimates of insect family-level abundance and 
diversity in a Nebraska salt marsh. Entomologia 
Experimentalis et Applicata 91: 393^02. 

Kremen, C, R. K. Colwell, T. L. Erwin, D. D. Murphy, 
R. F. Noss, and M. A. Sanjayan. 1993. Terrestrial 
arthropod assemblages: their use in conservation 
planning. Conservation Biology 7: 796-808. 

LaSalle, J. and I. D. Gauld. 1992. Parasitic Hymenop- 
tera and the biodiversity crisis. Redia 74: 315-334. 

and I. D. Gauld. 1993. Hymenoptera: Their 

diversity, and their impact on the diversity of 
other organisms. Pp. 1-26 in: LaSalle, J., and I. D. 
Gauld eds. Hymenoptera and Biodiversity. Wall- 
ingford, United Kingdom. 348 pp. 

Longino, J. T. and R. K. Colwell. 1997. Biodiversity 
assessment using structured inventory: capturing 
the ant fauna of a tropical rain forest. Ecological 
Applications 7: 1263-1277. 

Magurran, A. E. 2004. Measuring Biological Diversity. 
Maiden, MA: Blackwell Publishing. 256 pp. 

Martin, J. L. 1966. The insect ecology of red pine 
plantations in Central Ontario. IV. The crown 
fauna. Canadian Entomologist 98: 10-27. 

Masner, L. and H. Goulet. 1981. A new model of 
flight-interception trap for some Hymenopterous 
insects. Entomological Neivs 92: 199-202. 

Mendenhall, VV„ R. J. Beaver, and B. M. Beaver. 2003. 
Introduction to Probability and Statistics 11 th edn. 
California: Brooks /Cole-Thomson Learning. 720 pp. 

Noyes, J. S. 1978. On the numbers and species of 
Chalcidoidea (Hymenoptera) in the world. En- 
tomologist's Gazette 29: 163-164. 

. 1982. Collecting and preserving chalcid wasps 

(Hymenoptera: Chalcidoidea). Journal of Natural 
History 16: 315-334. 

. 1989. A study of five methods of sampling 



Hymenoptera (Insecta) in a tropical rainforest, 
with special reference to the Parasitica. Journal of 
Natural History 23: 285-289. 

2000. Encyrtidae of Costa Rica (Hymenoptera: 



Chalcidoidea), 1. The subfamily Tetracneminae, 
parasitoids of mealybugs (Homoptera: Pseudo- 
coccidae). Memoirs of the American Entomological 
Institute 62: 1-355. 

Roberts, H. R. 1973. Arboreal Orthoptera in the rain 
forest of Costa Rica collected with insecticide: 
a report on the grasshoppers (Acrididae), in- 
cluding new species. Proceedings of the Academy of 
Natural Sciences of Philadelphia 125: 46-66. 

Southwood, T. R. E. 1978. Ecological Methods, 2nd edn. 
Chapman and Hall, New York. 524 pp. 

Stork, N. E. 1988. Insect diversity: Facts, fiction and 
speculation. Biological Journal of the Linnean Society 
35: 321-337. 

and P. M. Hammond. 1997. Sampling arthro- 
pods from tree-crowns by fogging with knock- 
down insecticides: lessons from studies of oak 



Volume 16, Number 1, 2007 



205 



tree beetle assemblages in Richmond Park (UK). 
Pp. 3-26 in: Stork, N. E., J. Adis, and R. K. 
Didham, eds. Canopy Arthropods. Chapman and 
Hall, London. 567 pp. 
Van Dreische, R. G. and T. S. Bellows. 1996. Biological 
Control. New York, NY, USA: Chapman and Hall. 
567 pp. 



Wolda, H. 1981. Similarity indices, sample size and 
diversity. Oecologia 50: 296-302. 

Yanoviak, S. P., N. M. Nadkarni, and J. C. Gering. 
2003. Arthropods in epiphytes: a diversity com- 
ponent that is not effectively sampled by 
canopy fogging. Biodiversity and Conservation 12: 
731-741. 



J. HYM. RES. 

Vol. 16(1), 2007, pp. 206-209 

Marjorie Chapman Townes 
28 March 1909-8 October 2006 

Jean Townes and John Morse* 

*Author for correspondence: Department of Entomology, Soils and Plant Sciences, Clemson 
University, Long Hall, Box 340315, Clemson, SC 29634-0315, USA; email: jmorse@clemson.edu 



Marjorie Chapman Townes, 97, of Tim- 
berhill Place, Corvallis, Oregon, died 8 
October 2006. 

She was the widow of Henry Keith 
Townes, Jr., who died in 1990. 

Born in Pawcatuck, Connecticut, on 28 
March 1909, she was the daughter of the 
late William Robinson Chapman and Win- 
ifred Naomi Brown Chapman. 

Marjorie attended Westerly, Rhode Is- 
land public schools and was Valedictorian 
of her class at Westerly High School in 
1927. She then attended Mt. Holyoke 
College, from which she graduated "Cum 
Laude" with a BA in Botany in 1931, 
a member of Phi Beta Kappa honor society. 
Marjorie completed her MA in Botany at 
Cornell in 1932 with a thesis on the 
anatomy of Witches' Broom and her PhD 
in 1933, with a dissertation on the floral 
anatomy of Berberidaceae. She then taught 
Biology at Mt. Holyoke College during 
1935-1937. 

On 7 October 1937, she married ento- 
mologist Dr. Henry Townes from Green- 
ville, South Carolina, whom she had met at 
Cornell. The young couple lived in Syr- 
acuse, New York, for a year, where her 
husband taught at Syracuse University and 
where their son David was born, then two 
years at Ithaca, New York, while her 
husband taught at Cornell University. 
They next lived a year in Philadelphia, 
Pennsylvania, where their daughter Jean 
was born, and as Henry conducted re- 
search at the Academy of Natural Sciences, 
then in Takoma Park, Maryland (1941- 
1948), and McLean, Virginia (1948-1949), 



while he was employed by the U.S. De- 
partment of Agriculture in Washington, 
D.C. as a research taxonomist. From 1949 
to 1952, they lived in Raleigh, North 
Carolina, while Henry taught at North 
Carolina State University and investigated 
tobacco pests. During 1952-1954, they lived 
in Manila, The Philippines, where Henry 
served as an Advisor to the Philippine 
Department of Agriculture concerning 
pests of rice and corn. They then moved 
to Ann Arbor, Michigan (1956-1985), 
where they worked together on various 
research projects concerning Hymenop- 
tera, supported by grant funds, and Henry 
occasionally taught at the University of 
Michigan and Michigan State University. 
While in Ann Arbor, they established the 
American Entomological Institute as a not- 
for-profit organization to manage the huge, 
world-class Hymenoptera collection and 
library that the two of them had by then 
amassed. In 1985, they moved the Institute 
to Gainesville, Florida, operating it as 
a private organization supported by grants 
and endowment income. After Henry died 
in 1990, Marjorie moved to Eugene, Ore- 
gon, to be near her daughter Jean, later 
moving to Corvallis in 2003, to be even 
closer. 

Marjorie and Henry were a remarkable 
research team. She published 14 mono- 
graphs with him mostly concerning para- 
sitic wasps (see below). Together they 
undertook numerous entomological expe- 
ditions to many countries around the 
world and prepared the resulting speci- 
mens for further scientific study. The 



Volume 16, Number 1, 2007 



207 



Institute began to publish books and re- 
search articles in the 1960s, for which 
Marjorie undertook responsibility for sub- 
scription and sales. Her work on the 
labelling and arrangement of the Institute's 
collection was indispensable, and contrib- 
uted to its consisting of 900,000 specimens 
by 1990. Together, the Townes were ap- 
preciated internationally as a team who 
contributed significantly to our under- 
standing of the taxonomy of Hymenoptera, 
especially of the family Ichneumonidae. 

Marjorie is survived by her son David 
Townes of New York City, New York, and 
her daughter Jean Townes of Corvallis, 
Oregon, and by her grandchildren Andrew 
(28), Edward (20), Alice (19), and Catherine 
(16). 

PUBLICATIONS OF M.C. TOWNES 

Townes, H. and M. Townes. 1950. A revision of the 
genera and of the American species of Tryphonini 
(Hymenoptera: Ichneumonidae). Annals of the 
Entomological Society of America 42: 321-447. 

and M. Townes. 1951. Family Ichneumonidae. 

In C. F. W. Muesebeck, K. V. Krombein, and H. K. 
Townes. Hymenoptera of America North of 
Mexico - synoptic catalog. United States Depart- 
ment of Agriculture Agricultural Monograph 2: 
1-1420. 

and M. Townes. 1952. A revision of the genera 

and of the Nearctic species of Grypocentrini 
(Hymenoptera, Ichneumonidae). Proceedings of 
the Entomological Society of Washington 53: 301-313. 

and M. Townes. 1959. Ichneumon-flies of 

America north of Mexico: 1. Subfamily Metopii- 
nae. United States National Museum Bulletin 216 
(1): 1-318. 

and M. Townes. 1960. Ichneumon-flies of 

America north of Mexico: 2. Subfamilies Ephialti- 
nae, Xoridinae,and Acaenitinae. United States 
National Museum Bulletin 216 (2): 1-676. 

, M. Townes, and V. K. Gupta. 1961. A 

catalogue and reclassification of the Indo-Austra- 
lian Ichneumonidae. Memoirs of the American 
Entomological Institute 1: 1-522. 

and M. Townes. 1962. Ichneumon-flies of 

America north of Mexico: 3. Subfamily Gelinae, 
tribe Mesostenini. United States National Museum 
Bulletin 216 (3): 1-602. 

, S. Momoi, and M. Townes. 1965. A catalogue 

and reclassification of the eastern Palearctic 
Ichneumonidae. Memoirs of the American Entomo- 
logical Institute 5: 1-671. 



and M. Townes. 1966. A catalogue and 

reclassification of the Neotropic Ichneumonidae. 
Memoirs of the American Entomological Institute 8: 
1-367. 

and M. Townes. 1973. A catalogue and 

reclassification of the Ethiopian Ichneumonidae. 
Memoirs of the American Entomological Institute 19: 
1-416. 

— and M. Townes. 1978. Ichneumon-flies ot 
America north of Mexico: 7. Subfamily Banchi- 
nae, tribes Lissonorini and Banchini. Memoir- oj 
the American Entomological Institute 26: 1-614. 

— and M. Townes. 1981. A revision of the 
Serphidae (Hymenoptera). Memoirs of the Ameri- 
can Entomological Institute 32: 1-541. 

— and M. Townes. 1982. A description of the 
Townes collection of Hymenoptera. Contributions 
of the American Entomological Institute 20: 15-30. 

-, V. K. Gupta, and M. Townes. 1992. Ichneu- 



mon-flies of America North of Mexico: 1 1 . Sub- 
family Tryphoninae, Tribes Oedemopsini, Try- 
phonini, and Idiogrammatini. Memoirs of the 
American Entomological Institute 51: 1-292. 

Townes, M. and E. Linna. 1963. The location of some 
obscure entomological collecting localities in the 
United States and Canada. Proceedings of the 
Entomological Society of Washington 65: 233-246. 

. 1972. A.A. Girault and his privately printed 

papers. The Great Lakes Entomologist 5: 129-132. 

Jean Townes, Covallis, OR 
John Morse, Pendleton, SC 

Several scientists, mostly those who knew her 
and who are now members of the Board of 
Directors of the AEl, have shared some of their 
reflections about Marjorie: 

"Several nights ago I was working late, 
making a list of Neotropical ichneumonids 
as part of setting up a research project lor 
a beginning graduate student. As I ex- 
tracted information from several of the 
catalogs published by the AEI in the 
Sixties, I was struck by the realization - 
and not for the first time - of how much of 
this was the result of Marjorie's Labor. She 
was not a simple amanuensis. Although 
the scientific content was Henry's, an 
enormous amount of the arrangement, 
editing, and the like was due to Marjorie. 
She accompanied Henry to museums and 
went out into the field for extended trips, 
both foreign and domestic. The assemblage 
of the AEI collection would have been 



208 



Journal of Hymenoptera Research 



severely curtailed if not for her careful and 
assiduous labeling, and assistance with 
myriad curatorial tasks. In short, the 
modern ichneumonid enterprise would be 
very different if Marjorie had taken anoth- 
er branch of life's road. 

"While 'the modern ichneumonid enter- 
prise' is a fairly obscure branch of science, 
those of us involved with it believe in its 
importance. This group of wasps is enor- 
mous in both numbers of individuals and 
species, and we think that its study is 
necessary: both from a practical aspect 
(pesticide-free insect control and the like) 
and for a better appreciation of fellow 
travelers on this planet. Henry and Mar- 
jorie thought so too, and devoted their lives 
to building up an unique resource and 
study center. 

"Marjorie's labors are used every day in 
the form of examined specimens and 
consulted publications. Her life touches 
each of us constantly and will continue to 
do so as long as the entomological endeav- 
or continues." - Dr. David Wnhl, American 
Entomological Institute, Gainesville, Florida. 

"My first visit to the Townes' house was 
sometime in the early 1980s in Ann Arbor. 
Although I enjoyed getting direction and 
advice from Henry I can't remember any 
details of my interactions with him on that 
visit. What I do remember the most is 
Marjorie, and three things come to mind: 
the wonderful cookies that she made for 
our coffee breaks, her unique call, sort of 
like a whistle but not quite, that was used 
to call us for lunch and breaks, and her 
feeding of the birds outside of the window 
where I was working. These things may 
seem trivial but obviously they are not. 
Marjorie brought warmth and charm and 
made my visit delightful. I had other visits 
to Ann Arbor and many more to Gaines- 
ville; Marjorie was always gracious and 
full of energy. In later years, when Henry 
was ailing, Marjorie's bright personality 
obscured the difficulties that they were 
facing. I think of Marjorie often and I'm 
glad that the genus of braconid wasps, 



Marjoriella that I named after her is so 
difficult to place in the tree of life. It is 
beautiful, rare, and enigmatic, much like its 
namesake. She will be missed by the many, 
many scientists who had the great fortune 
to be touched by her." - Dr. Michael 
Sharkey, University of Kentucky. 
"My recollections of Marjorie Townes: 

"First, as a scientist, she ably assisted 
Henry in the publication of his descriptive 
work. I had an opportunity to observe this 
directly on some of Henry's later publica- 
tions, where Marjorie read the descriptions 
out to Henry as he mentally proofed them. 
As one who has proofed hundreds of 
articles, I can attest to the tremendous 
efficiency of this method. Proofing your 
own articles is a difficult task. Marjorie's 
expert assistance saved many hours and 
also improved the quality of the end- 
product as she would catch small typos 
and other errors that are easily overlooked 
when reading your own work. She also 
managed, as far as I could determine, the 
AEI publications, which was a pretty 
amazing achievement when you consider 
that the AEI publications routinely made 
more money than most periodicals. 

"Earlier in Henry's career, Marjorie co- 
authored several publications and justifi- 
ably so. She took an active part in all phases 
of the research, giving up her own area of 
training and expertise to provide full 
support to Henry's endeavors. Additional- 
ly, she assisted Henry in mounting newly 
acquired material and took over most of the 
duties labeling these accessions. In essence, 
Marjorie served as combination technician 
and fellow researcher. Without her very 
real contributions to AEI, Henry would not 
have been able to accomplish nearly as 
much as he did. 

"For a period of about a decade (or 
maybe a little more), I regularly visited the 
AEI as part of my taxonomic research, 
often staying one or two weeks at a time, 
and on several occasions accompanied by 
my wife and young daughters. While 
Henry was a wealth of information and 



Volume 16, Number 1, 2007 



209 



graciously allowed me full access to the 
collections, it was Marjorie who made us 
feel most welcome as visitors, opening up 
their home as a place for us to stay and, 
perhaps most importantly, assuring a re- 
turn visit on our part. From my perspec- 
tive, the extensive use of the AEI resources 
by scientists during that period was in 
large part due to the welcoming atmo- 
sphere that Marjorie was responsible for." 



- Dr. Bob Wharton, Texas A&M University, 
College Station, Texas. 

"I once received a shipment of speci- 
mens from the AEI that were packed in an 
assortment of colored foam peanuts. Marj- 
orie had included a little note: 'just like 
a party, isn't it!' She was a very nice and 
kind person and will be greatly missed." - 
Dr. John Heraty, University of California, 
Riverside. 



INSTRUCTIONS FOR AUTHORS 

General Policy. The Journal of Hymenoptera Research invites papers of high scientific quality reporting 
comprehensive research on all aspects of Hymenoptera, including biology, behavior, ecology, systematica, 
taxonomy, genetics, and morphology. Taxonomic papers describing single species are acceptable if the species has 
economic importance or provides new data on the biology or evolution of the genus or higher taxon. Manuscript 
length generally should not exceed 50 typed pages; however, no upper limit on length has been set for papers of 
exceptional quality and importance, including taxonomic monographs at generic or higher level. All papers will 
be reviewed by at least two referees. The referees will be chosen by the appropriate subject editor. However, it 
would be helpful if authors would submit the names of two persons who are competent to review the manuscript. 
The language of publication is English. Summaries in other languages are acceptable. 

The deadline for receipt of manuscripts is 1 September (for the April issue) and 1 March (for the October 
issue). 

Format and Preparation. Authors are strongly encouraged to submit manuscripts electronically to the editor 
at the email address below, and in the format specified below. On the upper left of the title page give name, 
address, telephone and fax numbers, and email address of the author to whom all correspondent is to be sent. 
The paper should have a concise and informative title, followed by the names and addresses of all authors. The 
sequence of material should be: title, author(s), abstract, text, acknowledgments, literature cited, appendix, figure 
legends, figure copies (each numbered and identified), tables (each numbered and with heading). Each of the 
following should start a new page: (1) title page, (2) abstract, (3) text, (4) literature cited, (5) figure legends, (6) 
footnotes. 

Upon final acceptance of a manuscript, the author should provide the editor with an emailed IBM formatted 
electronic version. CD-ROMs or 3.5 inch floppy disks are acceptable. Because symbols and tables are not always 
correctly translated it is best to also send a printed copy of the manuscript. Preferred word processing programs 
are Microsoft Word and WordPerfect. If possible, all words that must be italicized should be done so, not 
underscored. Tables may be formatted in a spread sheet program such as MS Works or MS Excel. Text should be 
double-spaced typing, with 25 mm left and right margins. Tables should be put in a separate file. CDs and 
Diskettes should be accompanied by the name of the software program used (e.g., WordPerfect, Microsoft Word). 
Authors should keep backup copies of all material sent to the Editor. The Society cannot be responsible for 
diskettes or text mislaid or destroyed in transit or during editing. 

Illustrations should be planned for reduction to the dimension of the printed page (14 x 20.5 cm, column 
width 6.7 mm) and allow room for legends at the top and bottom. Do not make plates larger than 14 x 18 in. (35.5 
x 46 cm). Individual figures should be mounted on a suitable drawing board or similar heavy stock. Photographs 
should be trimmed, grouped together and abutted when mounted. Figure numbers should be on the plate. 
Include title, author(s) and address(es), and illustration numbers on back of each plate. Original figures need not 
be sent until requested by the editor, usually after the manuscript has been accepted. Reference to figures/ tables 
in the text should be in the style "(Fig.l)" "(Table 1)". Measurements should be in the metric system. 

Electronic plates may be submitted on disc, via email or uploaded to an ftp site (instructions will be given). 
They must be fully composited, labeled, and sized to fit the proportions of the journal page. Line art should be 
scanned at 1200 dpi (minimum input resolution is 600 dpi). Color or grayscale (halftone) images should have a 
dpi of 300-350. Color files should be in CMYK and not RGB. Graphics should be submitted as TIFF, Adobe 
Illustrator or EPS files. No PowerPoint or Word /Word Perfect files with images embedded in them are 
acceptable. 

All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or 
animal name should include the full scientific name including the authority. Genus names should not be 
abbreviated at the beginning of a sentence. In taxonomic papers type specimens must be clearly designated, type 
depositories must be clearly indicated, and new taxa must be clearly differentiated from existing taxa by means 
of keys or differential diagnoses. Authors are required to deposit all type material in 

Recognized institutions (not private collections). Voucher specimens should be designated for specimens 
used in behavioral or autecological studies, and they should be deposited similarly. DNA sequences must be 
deposited in GenBank/EMBL/DNA Databank of Japan. 

Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will 
be expected to specify the phylogenetic program used, including discussion of program options used. A data 
matrix should be provided for morphological characters. Cladograms must be hung with characters and these 
should include descriptors (not numbers alone) when feasible. The number of parsimonious cladograms 
generated should be stated and reasons given for the one adopted. Lengths and consistency indices should be 
provided. Adequate discussions should be given for characters, plesiomorphic conditions, and distributions ol 
characters among outgroups when problematical. 

References in the text should be (Smith 1999), without a comma, or Smith (1999). Two articles by a single 
author should be (Smith 1999a, 1999b) or Smith (1999a, 1999b). For multiple authors, use the word -\\nd;' not the 
symbol "&" (Smith and Jones 1999). For papers in press, use "in press," not the expected publication date. The 
Literature Cited section should include all papers referred to in the paper. Journal names should be spelled out 
completely and in italics. 

Charges. Publication charges are $10.00 per printed page. At least one author of the paper must be a member 
of the International Society of Hymenopterists. Reprints are charged to the author and must be ordered when 
returning the proofs; there are no free reprints. Author's corrections and changes in proof are also charged to the 
author. Color plates will be billed at full cost to the author. 

All manuscripts and correspondence should be addressed to: 

Dr Gavin Broad 

Dept. of Entomology 

The Natural History Museum 

Cromwell Road 

London SW7 5BD, UK 

Phone: +44(0)207 9425938; Fax: +44(0)207 9425229; Email: editor@hymenopterists.org 




3 9088 01327 2612