HARVARD UNIVERSITY

LIBRARY

OF THE

Museum of Comparative Zoology

ISHtO

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

•>_••*•.•

i

8

I

s

5

g

MUS. CO MP. ZOOL
_ LIBRARY

FEB 1 1 '^(^

HA\NVAi^D

UNIVERSITY

THE COMPARATIVE EXTERNAL

MORPHOLOGY AND SYSTEMATICS

OF THE NEOTROPICAL PARASITIC

FIG WASP GENUS IDARNES

(HYMENOPTERA: TORYMIDAE)

By

GORDON GORDH

>;:

:::

1

1

i

ft

^

:;:;

•.*.

v

<•

.*.

V

•l;

.♦.

.y

V,

• •

■ •

»"#

• •

•M

•.•.

!v

:•>

>:

•K

*!■

•_

■ •

•_a

• •

■ 9^

• •

« •

• m

• •.

• 9

s •

9 «

1 •

. •

• a

• «

I •

• •

« ■

v

■H

• •

t •

• •

• 1

■ •

t •

'A

.V

'.%

«

Vol. 50, No. 9, pp. 389-455

January 30, 1975

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out at
the University of Kansas or by University faculty and students. Since its incep-
tion, volumes of the Bulletin have been variously issued as single bound volumes,
as two or three multi-paper parts or as series of individual papers. Issuance is at
irregular intervals, with each volume prior to volume 50 approximately 1000
pages in length. Effective with volume 50, page size has been enlarged, reducing
the length of each volume to about 750 pages.

The supply of all volumes of the Kansas University Quarterly is now ex-
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased at
$20.00 per volume. Where some of these volumes were issued in parts, individual
parts are priced at the rate of 2 cents per page. Current policy, initiated with
volume 46, is to issue individual paf>ers as published. Such separata may be
purchased individually at the rate of 3 cents per page, with a minimum charge
of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at
the rate of $20.00 per volume. All communications regarding exchanges, sales
and subscriptions should be addressed to the Exchange Librarian, University
OF Kansas Libraries, Lawrence, Kansas 66045.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

The International Standard Serial Number of this publication is US ISSN
0022-8850.

Editor
Charles R. Wyttenbach

Editorial Board

William L. Bloom
Philip W. Hedrick

Rudolf Jander

Harvey Lilly white

Charles D. Michener

Norman A. Slade

Henry D. Stone

George W. Byers, Chairman

i

A
A

-1^ 1/

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. 50, No. 9, pp. 389-455 January 30, 1975

The Comparative External Morphology and Systematics

of the Neotropical Parasitic Fig Wasp Genus Idarnes

(Hymenoptera: Torymidae)

Gordon Gordh

TABLE OF CONTENTS

Abstract 391

Introduction 391

Acknowledgments 391

Materials and Methods 392

The External Morphology of Idarnes 392

I. Female Head 392

II. Female Antenna 397

III. Female Thorax 397

IV. Wings 403

V. Female Legs 405

VI. Female Abdomen 406

VII. Female Genitalia 406

VIII. Male Head 408

IX. Male Antenna 408

X. Male Mesosoma 408

XI. Male Legs 410

XII. Male Gaster 412

XIII. Male Genitalia 412

Systematics, Morphology, and Distribution of Ficus 412

The Parasites in Floral Ecology 414

Host Specificity and Idarnes 415

Redefinition of Idarnes 416

Key to Some Species of Idarnes 417

Systematic Treatment 418

Species Incorrectly Attributed to Idarnes 438

Literature Cited 439

The Comparative External Morphology and Systematics of

The Neotropical Parasitic Fig Wasp Genus Idarnes

(Hymenoptera :Torymidae) ^

Gordon Gordh^

ABSTRACT

The external morphology of the parasitic fig wasp genus Idarnes is examined and compared
to more primitive and advanced hymenopterous forms. Several secondary features, noticed by
earlier workers, have been named in the present paper. These features include setal tracts in
the forewing and mesosomal apodemes of the female.

In spite of its distinctive appearance, Idarnes has never been properly characterized. There-
tore, the genus has been redefined and a neotype designated. Eleven species are descrilied as
new and numerous species erroneously included earlier in Idarnes are discussed.

Although Idarnes has been considered pantropical, the present work develops the hy{x>th-
esis that it is exclusively neotropical. Moreover, evidence is presented which suggests Idarnes
is found exclusively associated with the subgenus Urostigma. Ramirez has demonstrated that
a high degree of specificity exists between species of agaonid poUenator and Ficus host. Data
collected throughout Mexico and Central America indicate a similar pattern exists for Idarnes.
However, it is uncertain whether the specificity exhibited by Idarnes is connected to the pol-
lenator, host fig, or both.

INTRODUCTION

Since its description in 1843 by Francis
Walker, the genus Idarnes has never been
properly characterized, although numer-
ous identifications of "Idarnes carme" have
been made (Wolcott, 1951; Butcher, 1964;
cf. remarks of Burks, 1969). This study is
intended to provide a firm systematic
foundation for Idarnes and to elucidate its
external morphology.

Of the earlier workers, Mayr (1885,
1906) probably came closest to guessing
the true identity of Idarnes. Unfortu-
nately, the original generic description of
Idarnes is so vague that Mayr supposed
his material to be new and erected the
genus Tetragonaspis for it in spite of his
suspicions about Idarnes. Subsequent tax-
onomic studies (Ashmead, 1904; Girault,
1913) only obscured the issue. Aside from

'Contribution 1524 of the Department of Ento-
mology, University of Kansas, Lawrence. Kansas
66045.

'Current address: Systematic Entomology Lab-
oratory, U.S. Department of .Agriculture, % Smith-
sonian Institution, Washington, D.C. 20560.

an occasional remark (Risbec, 1951; Jo-
seph, 1964, 1967; Wiebes, 1966b, 1968, 1970;
Hill, 1967a; Burks, 1969), Idarnes has re-
mained essentially untouched systemat-
ically.

Nonetheless, Idarnes represents the
largest, most conspicious element of the
neotropical fauna of fig wasp parasites.
Although other parasitic genera are found,
they are never collected in the diversity or
abundance of Idarnes.

ACKNOWLEDGMENTS

Numerous individuals were in part re-
sponsible for bringing this work to fru-
ition. Mr. William Ramirez provided
much of the material studied from his
personal collection of Central American
figs and fig wasps; Dr. P. D. Ashlock
(University of Kansas) gave many help-
ful suggestions regarding a cladistic analy-
sis of the group; Dr. R. E. Beer (Univer-
sity of Kansas) aided in collecting material
in Mexico during the summer of 1970;
and Dr. B. D. Burks of the United States

391

392

The University of Kansas Science Bulletin

National Museum lent material from
which a neotype was designated. Dr. J. T.
Wiebes (Rijksuniversiteit, Leiden) had
the onerous task of comparing much of
the material with that of the Mayr collec-
tion and gave several taxonomic sugges-
tions. Professor C. D. Michener (Univer-
sity of Kansas) has offered encouragement,
suggestions, and a critical evaluation of
the entire project. To Professor J. H.
Camin (University of Kansas) I am
deeply indebted for the generous loan of
equipment, facilities, and space during the
entire gestation of this revision.

Linda Gordh typed rough drafts of the
manuscript and rendered much clerical
assistance.

MATERIALS AND METHODS

The material studied came from three
sources: collections by the author from
Mexico during the summers of 1969 and
1970; material provided by William Rami-
rez from Central America; and represent-
atives of previously described species pro-
vided by J. T. Wiebes.

Collecting in Mexico was by the meth-
ods described by Ramirez (1970b). For a
tabulation of Central and North American
host trees, Blastophaga pollinators, and as-
sociated I dames parasites, see Table 1. For
each collection, ripened syconia of the
Ficits were preserved in 70% alcohol;
branches, including young figs and leaves,
were pressed. In addition, samples of all
fig species collected during 1970 were
halved, photographed, and close-up photo-
graphs of branches taken. Identification of
host figs was based on the classification of
Standley (1917). When applicable, the
work of DeWolf (1960) was consulted;
Condit's (1969) work also proved useful.
Plant material collected during the sum-
mer of 1969 was retained by Ramirez;
plant material collected during the sum-
mer of 1970 is retained by the author.

All wasp material was collected into

70% alcohol. Associated pollinators (aga-
onids) were given to Ramirez and Wiebes,
while voucher specimens of torymids col-
lected during the 1970 trip were given to
Professor Wiebes. The bulk of the mate-
rial remained with the author. For the
deposition of types, see Table 2.

Specimens were dissected in depression
slides filled with 70% alcohol. Dissections
were made with miniiten nadeln hooks
embedded in heat-treated glass rods. The
following parts were separated from intact
female wasps: antennae, head, pro-, meso-,
and metathoracic legs, pronotum, meso-
sternum, labiomaxillary complex, gaster
and wings. Parts separated from intact
male wasps included antennae, legs and
occasionally the gaster. When each dissec-
tion was completed, wings and the labio-
maxillary complex were mounted on a
slide in Hoyer's medium. Other structures
were cleared in 10% KOH at 50° C for
12-16 hours, depending upon size and
sclerotization. The parts were then trans-
ferred to 70% alcohol for several hours
and subsequently mounted in Hoyer's.
Material sent to the U.S. National Mu-
seum of Natural History and many para-
types were mounted in Canada balsam.

After Hoyer's medium slides were
dried for several days at 50° C (with the
aid of a slide warmer), each coverslip was
sealed with Zut's Ringing Medium. This
procedure proved excellent for preventing
deterioration of the mount due to evapo-
ration of water from the Hoyer's.

Illustrations were prepared with a mi-
croprojector and a binocular phase con-
trast Zeiss microscope with drawing tube
attachment.

THE EXTERNAL MORPHOLOGY
OF IDARNES

I. Female Head (Figs. 1-3)

Based upon the axis of the gnathal ap-
pendages and the position of the occipital
foramen, the cranium of female Idarnes

MdKPHOLOGY AND SysTEMATICS OF NkoTKOIMCAL PaUASITIC FiC WasP

393

may exhibit two distinct conditions: it
may be prognathous (as in /. hiicatoma)
or more generally hypognathous (as in /.
barbigera, I. galbina). The shape of the
cranium is highly variable, ranging from
elongate (/. biicatoma) to round (l. gal-
bin a).

The sculpture of the head capsule is
also variable, depending upon the degree
of sclerotization. Species such as /. galbina
have lightly sclerotized heads and hence
are nearly smooth; /. obtusijoUae females
have honeycomb-like sculpturing which
results in heavy sclerotization.

When viewed from above, 3 triangu-
larly arranged ocelli are visible; when
viewed from the front, the position of the
anterior ocellus is variable, rancins; from
the level of the upper third of the com-
pound eyes to the crest of the vertex. Posi-
tion of the ocelli seems correlated with the
axis of the cranium : species with pro-
gnathous heads have ocelli located higher
on the vertex than do species with hypo-
gnathous heads.

The epicranial suture represents a car-
dinal landmark of the generalized insect
head. Within the Hymenoptera an epi-
cranial suture has been identified in the
Symphyta (Pergidae), some braconids,
and evanoids. Michener (1944b) indicated
the suture was not present in bees.

The epicranial suture is not evident in
Idarnes or any other chalcid examined
morphologically. As a consequence stand-
ard regions of the cranium are poorly de-
fined and homology with more primitive
insect heads is difficult. Michener (1944b)
recognized this problem in bees and cre-
ated artificial boundaries for the area be-
tween the anterior ocellus and toruli. He
applied the term "supra-antennal" which
is noncommital with regard to homology;
however, this term has not been retained
in the present work because the scrobal
cavity and interantennal ridge are well de-
fined and are more familiar terms.

Immediately lateral to the scrobal cav-
ity and adjacent to the compound eye
inner margin is another ill-defined region
termed the "face" in the present work. It
is bounded by imaginary transverse lines
extending from the anterior ocellus dor-
sally, and the toruli ventrally. The region
is called paraocular by Michener (1944b),
parafrontal by Pratt (1940) and parietal
by Bucher (194S). The face and scrobal
cavity together are regarded as the fronto-
vertex by some chalcid taxonomists. How-
ever, because the scrobe is so well defined
in Idarnes it is again useful to separate
these regions terminologically.

The compound eyes of Idarnes are lat-
eral to the face. They vary in shape,
ranging from round (/. galbina) to oval
(/. carme), but are never emarginate. The
eves are never setose in Idarnes. thoutrh
this characteristic is common in other
chalcidoids such as Aphytis (Aphelinidae) .
Many Hymenoptera (some braconids,
evanoids, vespids, and bees) have a cir-
cumocular sulcus that circles the com-
pound eye. If the sulcus is present in
Idarnes. it must coincide with the margin
of the eye itself since no conspicuous ring
has been observed. Some chalcids (e.g.,
Brachymeria, Chalcididae) bear a distinct
line of favose punctations which surround
the eye margin. However, this condition
is also absent from Idarnes.

Heads cleared in KOH or bleach re-
veal the presence of a large, well defined
ocular sclerite. This feature has been
found in all chalcids examined (including
representatives of 12 families). Presum-
ably the ocular sclerite supports the com-
pound eye and provides a shelf upon
which the eye rests. The ocular sclerite of
Idarnes is peninsulate along the inner
margin, but the function of this feature is
unknown.

The gena (= malar space) is immedi-
ately beneath the compound eye. When
the head is viewed in lateral aspect it is

394

The University of Kansas Science Bulletin

Intcranrciiiuil

Ritigc

llpiis

Figs. 1-3. 1, Head of Idarnes female, anterior aspect: 2, iabiomaxillary complex of Idarnes female, external

aspect; 3, left mandible of Idarnes female, inner aspect.

bounded dorsally by the eye, ventrally by
the pleurostome, and anteriorly and pos-
teriorly by imaginary lines extending from
the lateral margins of the compound eye
to the mandibular articulations. In most

chalcids a subocular suture (= malar sul-
cus) bisects the gena, extending from the
ventral margin of the eye to the pleuro-
stome. In many encyrtids, eupelmids, and
pteromalids the suture is bold and pre-

Morphology and SvsTiiMATics of Neotropical Parasitic Fig Wasp

395

sumably serves as a secondarily developed
reinforcement for the cranium. In Mono-
dontoments spp. also it is well developed,
but all species of Idames lack the suture,
' In the related genus Critogaster the sub-
ocular suture is incipient, represented only
by a weak, dark, incomplete line.

The insertion of the antennae in
, Idames ranges from the midline of the
: compound eyes to their ventral margins.
Moreover, the toruli may be nearly con-
tiguous (I. biicatoma) or separated by one
to several times the diameter of a single
torulus (7. camini). The antennal suture
circling the torulus is well defined. The
antennifer is variable, being conspicuous
in some species (/. sirnits) and reduced in
others (I. galbina). Between the toruli
and anterior ocellus, a shallow depression
(scrobe) which receives the scapes when
the antennae are in repose is always pres-
ent. The scrobe is divided by a mesal
interantennal ridge of variable length and
height. When well developed, it creates
two distinct channels which merge at the
anterior ocellus.

Since the epicranial and frontal sutures
are absent from Idames, the boundaries of
the frons are not certain. The frons is
taken to be a region below an imaginary
transverse line below the toruli and be-
tween perpendicular lines extending from
the outer margin of each torulus to the
pleurostome.

The anterior tentorial pits are minute
and can be located only with a compound
microscope by tracing the course of the
anterior tentorial arms to the surface of
the head. The epistomal and subgenal
sutures are absent from Idames. Thus, the
boundary separating the ventral margin of
the frons from the dorsal margin of the
clypeus is arbitrary. For the purpose of
the present paper, it is taken to be an
imaginary transverse line between the an-
terior tentorial pits. Bucher (1948) noted
M. dentipes and James (1926) indicated

Harmolita graminicola have epistomal
and subgenal sutures.

The labrum of Idames consists of a
membranous flap concealed beneath the
clypeus. This condition seems common to
most chalcids though some species of Leu-
cospis exhibit a lightly sclerotized retractile
sclerite and in Perilampus this sclerite re-
sembles a comb.

The mandibles (Fig. 3) of Idames are
highly sclerotized, hi- or tri-dentate. The
toothed margin may be sharply (/. flavi-
collis) or weakly (/. galbina) incised.
Mandibles bear 2 hollow, internal, cone-
like cavities extending into the primary
and secondary teeth. The third tooth of
tri-dentate species lacks such a feature.
These cavities seem universal in the Chal-
cidoidea. In some Pteromalidae (Pachy-
crep Oldens vindemiae, Pteromaliis spp.)
and Perilampidae (Perilampus spp.) the
left mandible may have 3 cavities. How-
ever, in all cases the function of this
adaptation is unknown.

The maxillae and labium (Fig. 2) are
bound together by membrane, forming the
labiomaxillary complex. Maxillary com-
ponents include cardo, stipes, galea, la-
cinea, and segmented palpus. The cardo
shape is variable, but it always flexes
mesally along the posterior margin of the
elongate stipes, except in isolated instances
(some individuals of /. micheneri) where
it is fused to the stipes. Each stipes char-
acteristically bears a basal acuminate seta.
The maxillary palpus arises from the an-
tero-lateral margin of the stipes and con-
sists of a proximal palpifer and 2 distal
segments. Lengths of the palpifer and pal-
pal segments are variable and the palpifer
and basal palpal segment may be fused.
The basal segment bears a scolopophorous
sensillum at the distal end and the terminal
segment is characterized by sensory spines
and acuminate setae. The galea and
lacineae are membranous, enveloping the
ligula and paraglossae. These modifica-

396

The University of Kansas Science Bulletin

tions contrast sharply with the condition
found in aculeate Hymenoptera (Matsuda,
1965) but conform to other chalcidoids.

The labium consists of a medio-distal
ligula (fused glossae), prementum, para-
glossae, and lateral labial palpi. The ligula
is surrounded by a crown of pilus dentilae
(flattened pronged setae) and is arched
sharply forward, bearing 2 or 4 sensory
pegs along its anterior margin. (Sensory
pegs along the margin of the ligula seem
characteristic of all chalcidoids, though the
number varies considerably among species.
Encyrtids such as Comperia merceti have
as many as 10 pegs; 2 pegs seems to be the
minimum number.) The ligula is broadly
attached to the prementum (which is al-
ways larger than the ligula in Idarnes),
which also bears the labial palpi on the
distal margin. The 1-segmented labial
palpi are always shorter than the maxil-
lary palpi, with the palpiger irregular. The
apex of each palpus bears a long, acumi-
nate seta, and there is often a long basal
one.

The paraglossae arise from the pre-
mentum lateral to the labial palpi. Each is
blade-like, lightly sclerotized, and projects
anteriorly along the lateral margin of the
ligula. They are exceedingly difficult to
see except with high magnification
(> 250X). The mentum and submentum
are not dififerentiated, a feature noted by
Matsuda (1965) for other chalcidoids.

The posterior surface of the chalcid
head has undergone considerable modifi-
cation, making homology difficult or im-
possible. In Idarnes it may be flat or con-
cave, depending upon the species. The
occipital suture (preoccipital ridge of
Michener, crassa of Ross) is inconsistently
developed in the Hymenoptera and alto-
gether lacking in Idarnes. Bucher (1948)
indicated Monodontomenis bears a well
developed carina he chose to call an oc-
cipital suture, but it is doubtful that this

carina is homologus with the occipital su-
ture of more primitive insects.

The dorsal margin of the occipital
foramen of Idarnes exhibits a thickening
which diminishes laterally. The post oc-
cipitial suture is poorly developed and
does not circle the foramen. The dorsal
thickening may contain remnants of the
post occipital suture, but they have not
been observed in any species of Idarnes
yet examined.

In M. dentipes, H. graminicola, and
most other chalcids the occipital foramen
is bisected by an apodeme which origi-
nates at the posterior tentorial pits. This
structure has been termed the corporoten-
torium, body of the tentorium, or primary
tentorial bridge by various writers on
Hymenoptera morphology. It is lacking
from all species of Idarnes. Posterior ten-
torial arms do project from the posterior
tentorial pits to the ventral margin of the
foramen in Idarnes, but these are not
fused.

The posterior tentorial pits are small
and may be recognized by tracing the ten-
torial arms to the surface of the head. The
pits are ventro-lateral to the foramen.

The sclerite which separates the fora-
men from the proboscidial fossa exhibits
two conditions. In some species (/. ash-
lockj), faint post occipital sutures run
parallel from the posterior tentorial pits to
the proboscidial fossa. The sclerite thus
formed is a gula according to the defini-
tion of Snodgrass (1935). Bucher (1948)
has made a similar observation for Mono-
dontomerus. However, in each instance
the gula has developed on a hypognathous
head. Alternatively, a faint median suture
may be present {I. simus) and extend
from the ventral margin of the foramen to
the proboscidial fossa. The postgenae are
separated by a postgenal groove. This
mesad migration of weakly developed
postoccipital sutures has drawn the lateral

Morphology and Systlmatics of Nuotropical Parasitic Fig Wasp

397

walls of the proboscidial fossa together
thereby forming a hypostomal bridge.

IT. Female Antenna (Fig. 4)
Antennae are either 12- or 13-seg-
mented because there may be 1 or 2 annuli
at the base of the flagellum. The first an-
nulus is always constricted basally; the
second, when present, appears to be a
fragment of the first flagellomere. The
annuli have been counted in numbering
the flagellomeres.

Surface features of individual flagello-
meres are highly variable among species,
but constant intraspecifically. The most
conspicuous modifications include longi-
tudinal carinae and three forms of setae,
inflexible (obdurate) setae (Figs. 25, 27,
28), flexible setae (Figs. 29, 34), and
acicular setae (Fig. 26). Setal types may
be distinguished on the basis of appear-
ance when mounted under a coverslip and
viewed with a low power compound mi-
croscope objective. Flexible setae are re-
curved, appear hollow and are often
apically blunt; inflexible setae are not re-
curv'ed or hollow, and are shorter than
flexible setae. Acicular setae are long and
transmit light in such a way as to make
the base resemble the eye of a needle. Each
type of seta, when present, forms a whorl
around each flagellomere. These setae
usually originate at the base of a flagello-
mere, but minute inflexible setae may, in
addition, be randomly dispersed in some
species. For a given species, one setal form
predominates.

A club may or may not be present.
When present, it is composed of the last
3 flagellomeres, and may be compact (/.
biicatoma, 1. tnicheneri) or loosely formed
(I. ashlockj). Regardless of club formation,
the distal 3 flagellomeres (= clavus) always
show the carinal and setal patterns found
on previous segments. A nipple-like ter-
minal protuberance is always present, al-
though sometimes weakly developed. It

bears clusters of short, thick setae and
minute discoid sensilla. This feature seems
to be a characteristic of the family Tory-
midae.

III. Female Thorax (Figs. 6-9)
Aside from a few incomplete statements
regarding specific features (cf. Compere,
1962; Compere and Rosen, 1970; Richards,
1956; Matsuda, 1960, 1970) of the pleuro-
sternum, no attempt has been made to
homologize the chalcidoid thorax with
that of more primitive and advanced hy-
menopterous forms. Although Idarnes is
seemingly specialized biologically, it is a
member of a chalcidoid family considered
to be primitive (Malyshev, 1966; Breland,
1938). Morphological features of Idarnes
tend to bear out this hypothesis.

The central portion of the chalcidoid
body (mesosoma, sensit Michener, 1944b)
consists of four regions: pro-, meso-,
metathorax and propodeum. Morpholog-
ically, the last represents the first abdom-
inal segment which has fused with the
thoracic components, a condition charac-
teristic of all clistogastrous Hymenoptera.
The term thorax will be used to refer only
to the three anterior segments of the meso-
soma; when the term mesosoma is used
the propodeum is also included in the
discussion.

The Notum (Figs. 6, 7)

The pronotum (Figs. 6, 7) of female
Idarnes is rigid, collar-Uke, and envelops
the anterior edge of the mesoscutum. The
pronotum is not sharply angular at the
anterior end, but is latero-ventrally ex-
panded, forcing the proepisterna into ven-
tral positions; it usually possesses the notal
sculpture pattern only along the postero-
dorsal edge, with the remainder of the
sclerite smooth or bearing distorted stri-
ations.

The mesonotum (Fig. 6) comprises the
bulk of the thoracic notum. It has been

398

The University of Kansas Science Bulletin

MageUcjnierc

Pedicel

Scaj)e

0.25 mm

Fics. 4-5. 4, Left antenna of Idarnes female, inner aspect; 5, left antenna of Idarnes male, dorsal aspect.

subdivided into 7 distinct regions, sepa-
rated by sutures, sulci, and apodemes.^

^ Snodgrass (1962) has indicated that the first
two terms have been used interchangeably and that
it is often impossible to distinguish between them.
As a consequence, the term suture is used through-
out the text.

Anterior-most of the mesonotal regions is
the prescutum, which is not visible from
a dorsal view and may be seen only when
the pro- and mesosternum have been sepa-
rated. The prescutum is bounded dorsally
by the mesoscutum to which it is united
at the transverse notal suture. In Idarnes,

Morphology and Systematics of Neotropical Parasitic Fig Wasp

399

ronotum

Nuuuilix

lVans-.scui;i
Siitiiic

Pura -axillary
.Suture

Axilla

Axillarv Cord

SpiracI

ictanotmii

Propodciim

Mcrcpi>tcrnum
Pltriiral SutiiTc

cjijula

Prepectus

Aiiaplcurirc

\naplcural Suture

Mesepisternum

Mesepimeron

0.5 mm

Figs. 6-7. Mesosoma of Idarnes female in dorsal (6) and lateral (7) aspects.

however, this suture is poorly defined.
Moreover, the prescutum represents a
lightly sclerotized, pendulous lip which
projects posteriorly beneath the meso-
scutum and effectively forms an anterior
rim that contains the dorsal longitudinal
flight muscles.

The mesoscutum (in a restricted sense,
including scapulae) lies immediately be-
hind the pronotum and is bounded later-

ally by notaulices^ and posteriorly by the

^ Various authors have referred to these longi-
tudinal inflections of the mesonotum as "parapsidal
sutures." Tulloch (1929) indicated that parapsidal
sutures are distinct from notaulices, with the former
extending anteriorly from the scutoscutellar suture
while the latter extend posteriorly from the anterior
margin of the mesoscutum. Functionally, parapsidal
sutures represent the site of attachment for rudi-
mentary indirect flight muscles, while notaulices are
associated with the distribution of connective tissue
during metamorphosis (Daly, 1964). Some chalcids
bear both structures, but Idarnes has only notaulices.

400

The University of Kansas Science Bulletin

trans-scutal suture. The shape of the meso-
scutum in Idarnes is constant. In lateral
aspect the posterior margin is flat, but
where the notaulices begin to diverge an-
teriorly, the mesoscutum is arched (Fig.
7). The regions lateral to the mesoscutum,
bounded by the notaulices, are the scapu-
lae;^ In Idarnes these sclerites are large
and constant in shape. They are bordered
posteriorly by the lateral parts of the trans-
scutal suture.

The trans-scutal suture of Idarnes is
transverse and straight. The principal
landmark posterior to this suture is the
mesoscutellum which is characteristically
quadrate. As with the scutum, the term
here is used in a restricted sense, the lat-
eral fragments being named separately for
convenience. Laterally, the mesoscutellum
is bounded by a secondary suture which I
have chosen to call the para-mesosciitellar
suture. Posteriorly, the mesoscutellum is
bounded by a bold suture-like impression
which Graham (1969) terms the frenal
groove. Immediately posterior to the fre-
nal groove and anterior to the metanotum
is a narrow, transverse region termed the
postscutellar zone in contradistinction to
the postscutellum of Burks (1943). The
postscutellar zone extends laterally to the
base of the forewing and forms the axil-
lary cord, a fact which indicates that it is
a fragmentum of the scutellum.

Lateral to the mesoscutellum and pos-
terior to the trans-scutal suture another
secondary suture has been formed. Orig-
inating adjacent to the intersection of the
notaulices, trans-scutal and para-meso-
scutellar sutures, a para-axillary suture^
extends obliquely rearward to the base of

''Graham (1969), in agreement with general-
morphological usage, regards the entire region cephad
to the scutoscutcUar suture in chalcids as the meso-
scutum, with development of notaulices subdividing
the region into a mid- and two lateral sclerites. But
since the scapulae are so distinct in many chalcidoids,
it is convenient to separate them terminologically.

* Graham (1969) calls this the "scutello-axillar
suture," a term less descriptive. The sclerotized area

the forewing. The suture separates the
axilla from a region termed the para-
scutellum by Grandi (1921). Grandi rec-
ognized the presence of the para-axillary
suture, but chose to refer to it as a "rin-
forzo endoscheletrico della linea che se-
para I'ascella dal parascutello." Subsequent
taxonomic studies on chalcid parasites as-
sociated with figs have illustrated this
suture, but not indicated a name or its
morphological significance (cf. work of
Joseph, Hill, etc.). The suture separates
the para-scutellum and axilla into two
equal regions. Incipient para-axillary su-
tures exist in other families of chalcidoids,
notably Eulophidae (Brothers and Moran,
1969) and Agaonidae (Grandi, 1929).

The second phragma (Fig. 6) is a
large, lightly sclerotized, bilobed plate
which provides attachment for the dorsal
longitudinal flight muscles. In Idarnes
the sclerite may be viewed only when the
mesonotum is separated from the propo-
deum and metathorax, since it develops
beneath the metanotum and dorsum of
the propodeum. Weber (1924) and Snod-
grass (1910) both believed the post-
phragma to be double-walled; however, it
appears single-walled and arises from the
metanotal antecosta.

The metathorax represents the final
thoracic subdivision of the mesosoma; it is
a narrow transverse band when viewed
from above (the general condition in Hy-
menoptera), and upon closer inspection
appears superficially divided into three re-
gions.^ The distinctness of the areas is
emphasized by different sculpture patterns
found on the meson and lateral regions.
The meson (dorsellum of Graham, 1969)

just behind this suture was regarded by Graham as
the "Axillula." Hucher (1948) noted the presence of
an incomplete suture in this region of the meso-
notum (if M. dentipcs, but chose to call it "suture of
the postscutum."

■' Owing to the reduction of the metanotum,
homologies with the mesonotum are difficult. The
terms meson and lateron will be applied.

M()RP11()L(K,V AM) SVSTI.MATICS OF Nl.O TROPICAL PaR vSlTIC FiG WaSP

401

is characteristicallv elevated tiom the hit
eral areas (metanoium ot (irahain, 1%0).
The three regions are more or less equal
ill size. While plicae may be evident in
some species of Idarnes, no lonL,ntudinal
sutures are present to separate the three
areas into distinct sclerites. The metanotal
postphragma is obsolete.

The Plcnrostfrtnd Rci^io/i (Figs. 7-9)

Several theories regarding the origin
and development of the pleurosternal re-
gion of the insect thorax have been put
forward. The primitive pleural region
must have been membranous. To account
for the origin and development of sclero-
tization, Heymons (1890) developed the
subcoxal theory of the sternal and pleural
sclerites. While most subsequent mor-

phologists have subscribed to this concept
in principle, their opinions regarding ho-
mologies of various sclerites have diverged
sharply (Weber, 1924, 1928; Snodgrass,
1935; Ferris, 1940; Richards, 1956; Du-
Porte, 1965).

The pleurosternal region of Idarnes is
primitive compared to that of most chal-
cidoids. As in other pterygotes, the pro-
thorax is simple when compared with a
wing-bearing segment. Hanna (1935) in-
terpreted the pleuron of Euchalcidia car-
yobori (Chalcididae) to be exclusively
episternal and the proepimeron to be ab-
sent. Idarnes conforms to this pattern, for
reasons indicated below. Earlier workers
sometimes regarded the episternal plates
as the prosternum; however, this interpre-
tation is untenable for Idarnes.

I'repectiis
Aiijpleural Suture

Anajileurite

Pleural Suture

Prestcrnal Suture

Pre^tci

Si^inasternum

cto cinsternal

Suture

Meseiiistei'num
)iserimcn

sepinicron

Inirca-

FiGS. 8-9. 8, Prosternum of Idarnes female, ventral aspect with furca flexed cephalad; 9. mcsothoracic pleuro-
sternum of Idarnes female, ventral aspect with furca flexed caudad.

402

The University of Kansas Science Bulletin

Idarnes lacks lateral cervical sclerites.
Instead, the anterior cranial processes
(Snodgrass, 1910) of the proepisterna ar-
ticulate directly with the posterior surface
of the head. As in other Hymenoptera, it
is not possible to state conclusively if the
cranial processes represent a fusion of cer-
vical sclerites with the proepisterna or
whether the processes are modifications of
the anterior-most portion of the proepi-
sterna.

The mesal surfaces of the proepisterna
are flat and contiguous along their anterior
ends. The proepisterna are ventral in po-
sition, seemingly having been pushed ven-
tromesally by the expanding pronotum.

The prosternum (Fig. 8) is a small
sclerite which lies immediately behind the
posterior proepisternal margins and be-
tween the forecoxae. It is highly variable
in shape, ranging from nearly round to
trapezoidal. Superficially, the prosternum
may be rugose to smooth with scattered
setae projecting posteriorly.

The discrimen of the prosternum is
evident and variably developed. In one
species (1. bticatoma) it extends nearly the
entire length of the sclerite; in some spe-
cies it is barely evident. Internally, the
prosternum bears 2 conspicuous apodemes,
the prosternal furca.

■ Idarnes appears to retain only proepi-
sterna, proepimera being absent. Internally
a ridge extends from each cranial process
rearward, running the entire length of the
propleuron along its lateral margin. This
ridge has been called by Snodgrass the in-
ternal pleural ridge, in spite of the fact that
the upper ends leads to the cranial process.
At the level of the apex of each strenal
apophysis, the internal pleural ridge is ex-
tended further inward, forming a pleural
apophysis. By definition (Snodgrass,
1935), the episternum and epimeron are
separated by the pleural suture. Although
the pleural suture is evanescent (since it is

the lateral proepisternal margin), the
pleural apophysis and ridge are evident.

The mesothoracic pleurosternum (Fig.
9) (= mesopectus of Compere, 1962) of
Idarnes is large and conspicuous. The
mesothoracic legs arise ventrally along the
posterior edge of the mesepisterna and are
mesally contiguous. The points of articu-
lation with the mesothorax are unmistak-
able, and the pleural suture is quite dis-
tinct; consequently, homologizing the
various mesothoracic sternal components
is comparatively simple. Interestingly
enough, the coxal condyles are apparently
sternal in origin since they are continuous
apodemes which originate along the dis-
crimen and fuse with the posterior margin
of the mesepisternum. This condition has
been observed in all trichogrammatids,
aphelinids and encyrtids examined mor-
phologically.

The anterior-most region of the meso-
thoracic pleuron is a lightly sclerotized,
nearly membranous presternum. Imme-
diately posterior to this is the area com-
monly called the prepectus. An internal
submarginal ridge (= presternal suture)
separates the two structures.

The triangular prepectus (Figs. 7, 9) is
large and not completely separated from
the rest of the mesepisternum. Numerous
other species of chalcidoids have the pre-
pectus more nearly separated from the
sternal plate (Microterys, Encyrtidae). I
call the suture which separates the pre-
pectus from the mesepisternum the pre-
pecto-episternal suture, and this should
not be confused with the pleural suture
which lies posteriorly.

Immediately behind the prepectus is
the main mesepisternal plate, which com-
prises the bulk of the pleural region of the
mesothorax. The discrimen separates the
mesepisterna midventrally. Internally, a
large, conspicuous furca is evident on the
posteromesal surface of the episternum.
The spinasternum has fused to the epi-

MORPHOLOCJV AND SysTLMATICS OF NtOTROPICAL PaRASITIC FiG WaSP

403

sterna mesally at the anterior end of the
discrimen.

The mesothoracic pleural suture is
identifiable by the position of the articula-
tory process of the mesothoracic coxa. The
coxal articulation is on the posterior edge
of the mesepisternum, halfway between
the mesepimeron and the discrimen. The
pleural suture runs obliquely forward sep-
arating the epimeron from the episternum.
Prepectus, epimeron, and episternum are
moderately sclerotized and bear sculpture
patterns that range from favose to lightly
shagreened.

Lateral to the main mesepisternal plate
is a conspicuous anapleural suture (= pleu-
ral ridge of Bucher, 1948; = sternopleural
suture of Snodgrass, 1910, 1935). The
sclerite lateral to this suture has been
termed "epansione marginale anteriore del
mesopleurum terminante nel denza della
regione epimerale" by Grandi (1929) for
Blastophaga psenes. The anapleural su-
ture must be a secondary development and
consequently the sclerite formed must be
part of the mesepisternum. It is here
called the anapleurite, and it arches dor-
sally beneath the forewing, providing a
fulcrum for wing movement. The upper
part of the pleural suture which leads to
this fulcrum in primitive insects is absent.
Here the anapleurite may have an epi-
meral as well as an episternal component.

The metasternum of Idarnes is un-
usual. The hind legs are latero-ventrally
situated, and from a lateral view only a
metepisternum is visible. The metepister-
num meets the mesepimeron along the
lateral thoracic wall. However, the two
sclerites have not formed a suture between
one another since they are readily separa-
ble upon dissection. Together, the mete-
pisterna form a moderate sized oblique
plate which confines the mesothoracic
coxae posteriorly. The metepisterna ex-
tend ventrally and meet along the longi-
tudinal central axis of the body, forming

an internal keel between the hind coxae.
The keel splays out laterally, forming a
lightly sclerotized plate which separates
the hind coxae from the longitudinal flight
muscles.

The metathoracic pleural suture (Fig.
7) originates at the metathoracic leg ar-
ticulatory point (seen as a notch along the
dorsal surface of the coxa) and continues
obliquely forward, terminating beneath
the hind wing. Immediately behind the
pleural suture is the propodeum; the met-
athoracic epimera have been nearly oblit-
erated. All that remains of the metepimera
are two thin, transverse bands, one on
each side, which converge mesally behind
the metathoracic coxae and connect to the
mesepisternal keel.

IV. Wings (Figs. 10, 11)
The Forewing

The forewing (Fig. 10) of Idarnes con-
forms to the typical chalcidoid pattern:
veination is reduced to a single vein which
has been subdivided into submarginal,
marginal, postmarginal, and stigmal re-
gions (Howard, 1887). The submarginal
vein angles forward along its distal third
and at the extreme end is notched; at the
notch two campaniform sensilla are al-
ways present. The portion of the sub-
marginal vein between the bend and the
notch is often called the parastigma. The
notch represents the beginning of the mar-
ginal vein, which continues to the fork
where the stigmal and postmarginal veins
originate. The stigmal vein may have a
distal bulb which is called the stigma. The
stigma invariably bears campaniform sen-
silla. The degree of stigmal development
appears to be a species characteristic. The
postmarginal vein is variable in length
and usually bears a few small setae.

Close examination of the wing reveals
distinct setal tracts which converge along
the distal edge of the wing. Burks (1938),
using representatives of 8 chalcid families

404

The University of Kansas Science Bulletin

Mardnal \'ein

Stigmal Vein

P()stmai\mnal \'cin

%^ ■ • ■ ■

arginal Tract

Submarginal Tract
V^Discal Tract

Medial

'■:';:.^ Anal Tract

raci

0.5 nim

Figs. 10-11. Right forewing (10) and hindwing (11) of Idanics female.

(no torymids), attempted to correlate ex-
tensive setal tracts and obsolete venation
with the hypothetical forewing venation
of Ross. Unfortunately, the setal tracts of
the Idarnes forewing do not reach the
basal portion of the wing or connect with
existing wing venation. Thus it is not
possible to identify the setal tracts of
Idarnes with the system of Burks.

Recently, Doutt and Viggiani (1968)*

" The systems of nomenclature employed by
Burks and by Doutt and Viggiani are useful from a
taxonomic stantlpoint, but imply a vestige of wing
venation and hence homology. Since the tracts found
in Idarnes are not from the same region of the wing
and cannot be homologized with extant venation in
other Hymcnoptera, new terms seem necessary.
Names used in the present work refer to areas with
which the reader may be familiar and no presump-
tions of homology are intentlcil.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

405

have indicated that sctal tracts provide
taxonomic characters among species of
trichogrammatids. Oldroyd and Ribbands
(1936) found that these tracts of macro-
trichae in Tric/iografuma evanescens were
influenced by the size of the host from
which the wasps emerged. Taxonomists
of Torymidae have not used the tracts in
chissification, although they have been il-
lustrated. Tracts have not been observed
in other Sycophagini but in I dames have
proven quite reliable in separating species
and appear constant in their formation.
Some variation of the type found by Old-
royd and Ribbands has been noted, but
this has not proven a problem taxonom-
ically.

The setal tracts of Idarnes all seem to
merge with the conspicuous marginal
fringe. The anterior-most tract may be
called the marginal tract; it is always pres-
ent in Idarnes. In some species it extends
to the postmarginal vein (/. micheneri),
while in others it terminates abruptly well
in front of the postmarginal vein (/. oscro-
cata). Immediately behind the marginal
tract is sometimes found a very short "i^i"
which terminates behind the postmarginal
vein {I. simiis). A submarginal tract in
some species originates along the apical
wing margin and arches to the area of the
stigma (I. obtiisijoliae). Posterior to the
submarginal tract and perpendicular to
the distal edge of the wing is sometimes a
variable discal tract (I. barbigera). This
tract is absent from most species, and diffi-
cult to interpret when present owing to
the random dispersal of microtrichae over
the general wing surface. The micro-
trichae are widespread on the wing surface
behind and beyond the stigma. Idarnes
jimenezi, I. obtiisijoliae, and /. ashlochj
bear a tract composed of microtrichiae
that are densely arranged along an imag-
inary longitudinal line which would con-
form to a median vein, hence the name

medial tract. I. jimenezi also bears an
anal tract.

Adstigmal setae, long, acuminate and
located just mesad to the stigma, provide
an additional forewing feature. Species
tend to exhibit some variation in number,
but distinct trends (clusters vs. 1 or 2
setae) are useful in determining species.

The Hind wing

The hindwing (Fig. 11) of Idarnes has
proven to be of some taxonomic value. It
is slender and elongate. The entire pos-
terior margin bears a conspicuous fringe
which is longer than the corresponding
forewing fringe. Venation is much re-
duced from that of other Sycophagini,
consisting of a short, stubby, submarginal
vein. The region between the tip of the
submarginal vein and hamuli may be se-
tose and gives the appearance of a vestigial
marginal vein. Other genera of Sycopha-
gini have marginal veins. The entire wing
surface bears microtrichiae.

V. Female Legs (Figs. 12-15)

The coxae conform in part to the typ-
ical torymid pattern: the hind coxa (Fig.
15) is larger than the fore coxa (Fig. 12),
but the middle coxa (Fig. 14) is reduced,
globular, and easily detached from the
mesosoma. Sculpturing of coxae may be
evident. The trochanters usually have sen-
silla near the apical margins. The meso-
thoracic femur is slender and atrophied.
Additionally, it is basally constricted, giv-
ing the impression of a second trochanter
being present. Femora of all legs may be
mesally setose. The fore leg tibia posseses
an apical bifurcate calcar (Fig. 13). Sen-
sory spurs beneath the calcar form a comb
which presumably facilitates antennal
cleaning. Various authors (Grandi, Jo-
seph) have used the number of apical
tibial setae and spurs as a specific char-
acter. The present study has revealed the
number to be variable, and not a good di-
agnostic character for Idarnes (cf. Wiebes,

406

The University of Kansas Science Bulletin

1968). The middle tibia (Fig. 14) gener-
ally has a single large spur, while the rear
tibia generally has 2 subequal spurs. The
smaller spur may often be confused with
part of the setal tract which runs along the
length (mesad) of the hind tibia.

The ratio of basitarsal length to that of
other tarsomeres proves to be reliable in
separating species of Idanies. The basi-
tarsus of the fore leg possesses a mesal
setal comb (the number of setae is again
not constant) and invariably is shorter
than the other basitarsi. The middle basi-
tarsus is usually as long as the remaining
tarsomeres, while the hind basitarsus is
variable. All tarsomeres are clothed with
variable numbers of setae which are not
constant within species in number or ar-

rangement.

VI. Fetnale Abdomen (Figs. 6, 7, 23, 24)

The propodeum (Figs. 6, 7) is large
with spiracles along the antero-lateral mar-
gin. Subparallel carinae and costulae are
never present, but sculpturing may be de-
veloped. The gaster attaches to the propo-
deum via abdominal segment 2 (petiole),
but the attachment is somewhat broad
for clistogastrous Hymenoptera. The rela-
tively broad attachment restricts the
amount of flexibility possible for the gaster.

The gaster (Fig. 23) is comparatively
simple. Its general shape ranges from
ovoid with flat sterna to nearly spherical
with the sterna convex. Tergal and sternal
plates are conspicuous, though the extent
of development of the sterna may be some-
what variable. The surface of each tergal
sclerite may be lightly shagreened or
smooth. These plates may be glabrous or
setose along the posterior margin. The
posterior margins of terga 2-6 may be
either straight (/. fiavicoUis) or strongly
sinuate (I. obtitsijoliae). The latter condi-
tion appears to be an adaptation for move-
ment within the fig, since tergal margins
which are sinuate permit more flexibility
than straight margins. All sterna have pos-

terior margins which are straight; sternal
texture varies from coriaceous to smooth.
Abdominal pleurites are absent.

VII. Female Genitalia (Figs. 23, 24)

The most conspicuous feature of the
female torymid gaster is the ovipositor.
Morphologically, the external female
genitalia represent lateral projections
(gonostyli = sheath) and mesal extensions
(gonapophysis = ovipositor) from gono-
coxites of the eighth and ninth abdominal
segments (seventh and eighth metasomal
segments). Idarnes conforms to the rest
of the Pterygota in that the gonostyli of'
gonocoxite 8 have been lost. The gonostyli
of gonocoxite 9 are equal in length to the
ovipositor, serving as a sheath.

In the ovipositor gonapophysis 8 is
ventral, while 9 is dorsal. Smith (1969)
has observed that gonapophysis 9 is in-
verted and interconnected with 8 along
the long axis. The euventral surface of 9
provides a tongue (rachis) which fits
within a trough (aulax) of gonapophysis
8. Smith calls the track locking gonopo-
physis 8 and 9 together an olistheter. His
hypothesis is applicable to Idarnes since
gonapophyses 8 and 9 are connected along
their entire lengths.

In all species of Idarnes, the external
female genitalia (Fig. 23) are several times
longer than the gaster. The long oviposi-
tor is an adaptation for penetrating the fig
receptacle and depositing eggs inside the
gall flowers. Idarnes is not adapted for
entering the fig through the ostiole. Dur-
ing oviposition, only the gonapophyses
penetrate the receptacle of the fig while
the gonostyli diverge perpendicularly to
the axis of penetration of the ovipositor.
The mesal surface of each gonostylus ap-
pears transversely striated and is sparsely
setose. The extreme tip of each gonapo-
physis is serrated to facilitate penetration
of the receptacle.

As indicated by Smith (1969), insects
retaining both gonapophysis 8 and 9 are

Morphology and Systematics of Neotropical Parasitic Fig Wasp

407

12

14

15

0.5 mm

Figs. 12-15. 12, Right fore leg of Idarnes female; 13, enlarged calcar and strigil of Idarnes female fore leg; 14,
right middle leg of Idarnes female; 15, left hind leg of Idarnes female.

408

The University of Kansas Science Bulletin

limited in the distance 8 can be thrust in
relation to 9 (for movement of the egg),
and little vertical leverage can be applied
against the ovipositional surface. In
Idarnes, the problem is further compli-
cated by the length of the gonapophysis.
The former problem has been alleviated to
an extent in most Hymenoptera by the
development of an articulation of gono-
coxite 8 with tergum 9.

The pygostyli (Fig. 24) are well de-
veloped and situated at the apex of the
gaster. Three apical setae are present and
are presumably sensory, indicating to the
female precisely how deep the ovipositor
has been thrust into the fig.

VIII. Male Head (Fig. 16)

All male Idarnes exhibit moderately
sclerotized, depressed, subtrapezoidal, cra-
nial capsules. The occipital foramen is
located on the posterior surface of the head
which is prognathous. The mandibles are
located on the extreme anterior end of the
head and are articulated in a horizontal
plane. Dentition is variable among species,
but hollow cones (of the type mentioned
in females) are always present in the first
and second teeth (cf. Figs. 74-82). Imme-
diately posterior to the mandibles and
along the lateral margins of the head are
the compound eyes, identifiable only by
red pigment spots. The ocular suture and
ocelli are absent. Mesal to the compound
eye, along the base of the mandible, two
conspicuous, thick-walled toruli are pres-
ent.

IX. Male Antenna (Fig. 5)

Antennae are 4- or 5-segmented. The
basal segment is spoon-shaped and thick-
walled, with setae over the dorsal surface
and along the lateral margins. The distal
surface is mesally concave, allowing the
second segment (pedicel) to fit within.
The second segment is elongate with thick
walls, and often bears thickened setae
along the apical end. The third segment

is translucent (except /. biicatoma), lacks
setae or thickened walls, and resembles a
collar. Internally, a canal is visible which
communicates between segments 2 and 4.
In species with 5 segments, a small seg-
ment is present immediately distal to the
third; species that have 4 antennal seg-
ments have this region completely fused
with the club. The apical segment repre-
sents the fusion of 3 segments into a club.
Setae are randomly dispersed over the
club, with longitudinal carinae, thick se-
tae, and sensilla on the terminus.

The labiomaxillary complex is com-
pletely absent, leading to the conclusion
that males do not feed (a supposition sup-
ported by the comparatively short life of
the male).

X. Male Mesosoma (Fig. 17)

The mesosoma is dorso-ventrally com-
pressed, lightly sclerotized, and bears few
external surface features. Internally, how-
ever, numerous heavily sclerotized ridges
are conspicuous. Presumably these are for
structural support. The pronotum repre-
sents the largest single sclerite of the meso-
soma." It is sometimes invested with
setae, but more often is glabrous. The
posterior margin of the pronotum is indi-
cated by a distinct lateral constriction.
Setae are often found in this region of the
pronotum (cf. /. micheneri). A prescutum
is not present in the male, hence the pro-
notum attaches directly to the meso-
scutum.

Since males are apterous, the meso-
thorax is not fully developed as in the fe-
male. The lateral margins of the meso-
notum project outward and these margins
may or may not be setose. The remainder
of the mesonotum is devoid of surface

" Rcid (1941) has indicated that with the re-
duction of wings, the mesonotal region may become
reduced; concomitant with this change, other regions
of the thorax become enlarged. Idarnes conforms to
Rcid's observations in that the pronotum is much
larger in the male than in the female.

MoRPlUJLOGY AND SySTLMATICS OF NeOTKOPICAL PaKASITIC FiG WaSP 409

Craiiial Prkcss

Pleural Ridge

aire a

Propodeiim

Pronotum

MesonotLini

Metanotum

0.1 mm

Figs. 16-17. Head (16) and mcsosoma (17) of Idarncs male, dorsal aspects.

410

The University of Kansas Science Bulletin

features: notauli, scapulae, axillae, para-
scutella, and other features found on the
female are absent from the male. Inter-
nally, however, an exceedingly complex
network of apodemes may be noted.
These provide points for muscle attach-
ment which are presumably highly modi-
fied. In many species (1. micheneri) the
mesothoracic discrimen is reduced. How-
ever, the internal apodemes (= furca)
which it produces seem well developed.
The furca of the male projects forward,
joining other apodemes (see Fig. 17).

The posterior margin of the meta-
notum is marked by two very faint lines
which project obliquely rearward but do
not connect. The metanotum is slightly
smaller than the mesonotum and there are
numerous internal apodemes evident, but
no external surface features are present.
The metathoracic discrimen appears to be
more fully developed than the correspond-
ing mesothoracic structure. The furca is
attached to parallel apodemes that connect
to other apodemes of the mesothorax. The
propodeal spiracles are located near the
antero-dorsal edge of the propodeum.

Immediately beneath the pronotum lies
the propleuron which is subdivided into a
mesal proepisternum and a lateral proepi-
meron. Each region is about the same
size, being separated by a large, conspic-
uous, internal pleural ridge. The ridge
provides points of attachment for the ster-
nal apophysis (furca) as well as for coxal
pro- and remotors. This enlarged apo-
deme presumably provides strength for the
legs, which are used to rip open the gall
flowers in which the females develop. An-
teriorly, the pleural ridge forms a very
large bilobed cranial process. Unlike fe-
males, all males exhibit a ball-and-socket
type cranial process into which the cranial
lobe fits. The position of the paired cra-
nial processes limits movement of the head
to dorso-ventral flexion.

Immediately posterior to the mesally

contiguous proepisterna is a small proster-
num. It is variable in shape and bears
apophyses which serve as endoskeletal
support for the prothorax.

XI. Male Legs (Figs. 18-20)

The legs of the males are uniform in
appearance. In fact, the legs of virtually
all male fig-inhabiting wasps are strikingly
similar, differing only in number of tarsal
segments (cf. work of Joseph, Wiebes,
Hill, and Grandi), All coxae are large,
disc-like, and flattened dorso-ventrally,
with setules and campaniform sensillae
scattered over the surfaces. The trochan-
ters are fused to the femora, and each ap-
pears as a sclerotized articulatory process
on the basal portion of the femur. In some
males a faint impression of the surface
indicates the separation of the trochanter
from the femur. All femora are smaller
than the coxae and appear uniformly flat,
disc-like, and lightly invested with setules
and sensillae. Each femur is articulated in
such a way that the dorsal surface may be
rotated beneath and pressed against the
ventral surface of the coxa. This feature
permits the wasp to move among compact
flowers within the fig receptacle.

Each tibia is heavily sclerotized, elon-
gate, and invested with heavy spine-like
setae along the outer and apical margins.
These spines permit the wasp to wedge its
body into narrow spaces. The fore tibia
(Fig. 18) of the male lacks the strigil
found in the female.

Five tarsomeres are found in male
Idarnes (except in /. biicatoma which has
4). The basitarsus is more heavily sclero-
tized than the remaining segments and
often bears apical spines similar to those
found on the tibia. The remaining tarso-
meres are lightly sclerotized and easily
shed; many males collected from recep-
tacles lack tarsomeres 2 to 5 on the fore
legs. This probably results from tearing
open gall flowers which contain females.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

411

20

0.1

mm

V^ ^y-r-

\^^1

Fk;s. 18-2U. Right fore leg (18), middle leg (19), and hind leg (20) of Idanics male.

412

The University of Kansas Science Bulletin

XII. Male Gaster (Fig. 21)
The gaster is globular and white.
There are seven distinct terga and sterna
with spiracles only on the seventh (eighth
abdominal) tergum. Abdominal segments
8 and 9 are reduced and serve only as the
origin of genital components. The pos-
terior tergal margins of most species are
sinuous, corresponding to the condition
found in some females. Terga may be
either glabrous or with distinct transverse
rows of setae. The length and number of
setae vary from species to species. Sterna
are always glabrous.

XIII. Male Genitalia (Fig. 22)

The external genitalia of chalcidoids
are relatively featureless when compared
with other insects. The basic pattern of
Idarnes is indicated below and except for
size, little or no variation exists among
species.

The basal ring which supports the ex-
ternal genitalic components is not evident
in Idarnes. Instead, support seems drawn
from aedaegal rods. The parameres are
elongate, undivided lobes, united to the
aedaegus basally. Volsellae lie between
the parameres and the aedaegus.

The free portion of the aedaegus pro-
jects between the parameres and bears a
pair of lateral rods. Beck (1933) considers
these rods gonapophyses of the ninth ab-
dominal sternum in bees; Michener
(1944a) calls these "mesal basal processes
of the gonocoxopodites"; Snodgrass (1951)
calls them "lateral sclerotizations of the
aedaegus supporting the aedaegal apo-
demes."

The volsellae appear unique to the Hy-
menoptera. Each one consists of an elon-
gate, ventral plate which lies adjacent to
the paramere. At the distal end are the
digitus and cuspis. These are strongly
musculated and form pincers. Snodgrass
(1941, 1951) and Michener (1944a) have
indicated these are used by most Hyme-
noptera to grasp and spread female genital

membranes during copulation. They pre-
sumably function in a similar manner for
Idarnes.

SYSTEMATICS, MORPHOLOGY,
AND DISTRIBUTION OF FICUS

The genus Ficiis (Moraecae) includes
over 900 described species placed in four
subgenera: Urostigma (Banyans), Phar-
macosycea, Sycomorus, and Ficiis (Croi-
zat, 1952; Corner, 1958, 1965). Of these,
only the first two are native in the New
World. All members of the genus bear
conspicuous hollow syconia (when ma-
ture, these are the fig "fruits") which en-
close the flowers which line the syconial
cavities. Three types of flowers exist: male
flowers that contain anthers, female flow-
ers, and gall flowers. Condit and Flanders
(1945) indicated that the term gall flower
is a misnomer since this floral type repre-
sents short-styled female flowers. The
term will be retained, however, because it
is convenient and well established. The
arrangement and location of staminate or
pistillate flowers within the syconium pro-
vide the foundation of fig classification.
In monoecious species, male, female, and
gall flowers are randomly interspersed
within the same syconium; dioecious spe-
cies bear male and female flowers on sepa-
rate trees with male flowers usually ar-
ranged in rows around the ostiole, which
is the opening into the syconial cavity.

Most species of figs yield 3 or 4 crops
per year. A tree will usually have all fruits
in the same stage of development, but
within a region, trees of the same species
may be in different floral stages. This
adaptive feature insures pollination by
agaonids.

Ficiis is absolutely dependent upon in-
sects for pollination; all pollinators belong
to the family Agaonidae. Recent studies
(Ramirez, 1970a, 1970b; Hill, 1967a, 1967b,
1969; Joseph, 1966; Wiebes, 1964, 1966a)
have indicated a high degree of host spe-

Morphology and Systhmatics of Neotropical Parasitic Fig Wasp

413

21

22

0.1 mm

Figs. 21-24. 21, Metasoma of Idarnes male, dorsal aspect; 22, genitalia of Idarnes male, ventral a.spect: 23,
metasoma of Idarnes female, dorsal aspect; 24, pygostyle of Idarnes female, \ entral aspect.

414

The University of Kansas Science Bulletin

cificity between pollinating wasp and tree.

The life history and pollination ecology
of numerous agaonids have been estab-
lished (Baker, 1913; GaHl and Eisikowitch,
1968; Ramirez, 1969; Williams, 1928;
Eisen, 1896; Carmin and Scheinkin, 1931;
Cunningham, 1889). Within a mature
syconium the insects develop inside the
gall flowers. The wingless males emerge
first and move around within the sy-
conium in search of females of the same
species. The male locates a gall flower
containing a female, gnaws a hole in the
gall, and mates with the immobile female.
Emergence from the syconium is achieved
by the males tunneling through the fig
wall or ostiolar scales. Several males may
work in the tunnel simultaneously, and if
the tunnel is not completed, all the inhabi-
tants of a syconium will die.

The winged females, before leaving
their natal fig, gather pollen from the
male flowers which are ripe at this time.
Some female pollinators collect pollen in
sternal or coxal corbiculae. Upon emer-
gence from the fig, the females fly to a
tree of the same species in the proper floral
phase and enter the figs via the ostiole.

Once inside the fig, the agaonid wan-
ders about, pollinating long-styled female
flowers; subsequently, she oviposits in the
short-styled gall flowers. Female polli-
nators always die inside the receptacle.
The young of the next generation mature
during the male phase of the same fig.

THE PARASITES IN
FLORAL ECOLOGY

Many non-pollinating hymenopterous
inhabitants of figs, members of the fami-
lies Torymidae and Eurytomidae, have
been described. The relationship between
Idarnes and the pollinating agaonids has
not been completely established. Katta-
mathiathu (= Joseph) (1955) has indicated
that Philotrypesis caricae, another tory-
mid fig inhabitant, develops "cleptopara-

sitically" at the expense of Blastophaga
psenes, pollinator of the edible fig, Ficus
(Ficits) caricae. P. caricae lacks the so-
called "poison glands" which Grandi
(1930, 1961) indicated are necessary for the
preparation of gall flower endosperm. B.
psenes presumably induces gall formation
(hence "gall flowers") when its eggs are
deposited inside a short-styled flower.
Later, the parasite lays its eggs in the same
flower and both larvae begin to develop.
Both Blastophaga and Philotrypesis de-
velop through the second instar at which
time the parasite kills the pollinator but
does not feed upon it.

Concerning the development of
Idarnes, the following observations can be
made:

Examination of preserved syconia of
Urostigma figs reveals that the parasites
also develop within short-styled flowers.
Each flower appears to contain only one
pollinator or parasite. Wiebes (1968) has
shown that in rare instances two wasps
may develop within the same flower. A
similar observation has not been made for
Idarnes. It has not been established if the
male Idarnes mates with the female before
or after her emergence from the gall
flower or if the male parasites aid the male
agaonids in digging an exit tunnel through
the wall of the receptacle. One interesting
feature of the material examined appears
to be the constancy of the tunnel site. Each
species of fig appears to have a unique site
for the emergence hole.

Joseph (1956) has indicated that female
torymid parasites will live up to 35 days
in the laboratory. Idarnes females have
been kept alive at high humidity (+70%)
for 9 days without feeding.

Observations of /. oscrocata have re-
vealed that newly emerged females do not
immediately oviposit within receptacles of
proper age. Instead, they congregate on
the undersides of leaves and on branches
of the host tree. Branches and leaves hav-

Morphology and Systematics of Neotropical Parasitic Fig Wasp

415

ing quiescent Idarnes also bear fruits into
which pollinating agaonids are burrowing.
If the hypothesis of Grandi regarding
atrophied poison glands is correct, the
torymids would delay oviposition until
after pollination and egg laying by the
agaonids.

Females of Idarnes do not enter the fie
through the ostiole. Instead, they use their
long ovipositors to penetrate the receptacle
of the fig. Initially the female moves rap-
idly over the receptacle, but eventually she
slows down. Movement over the syconium
continues for a few minutes to half an
hour. Several females may be noted on
one receptacle. Upon selection of a suit-
able place for oviposition, the wasp moves
forward slightly, raises the gaster and ele-
vates the body by extending the legs. After
the tip of the ovipositor has been inserted
into the receptacle surface, the wasp moves
the body on an anterior-posterior axis,
forcing the ovipositor deeper into the re-
ceptacle. The wasp moves rearward as the
ovipositor penetrates more deeply. Occa-
sionally, females have been observed to dis-
continue drilling and search for another
oviposition site on the same fig. As the
ovipositor penetrates the receptacle, the
gonostyli project rearward or postero-later-
ally. After laying is completed, the female
pulls herself forward, extracting the ovi-
positor.

Developmental time of the parasite
varies with the species but is approximately
the same as that of the agaonid pollinators.
Ramirez (1970b) has observed that for
most New World species of Fiats, almost
every syconium on a tree is pollinated dur-
ing a single day. Oviposition by Idarnes
is not as restricted since many parasites
may be collected from trees several days
after oviposition by the pollinators.

HOST SPECIFICITY AND IDARNES

The following data indicate that
Idarnes develops exclusively within Uro-

stigma figs: (1) In 17 collections of figs
of the subgenus Pharmacosycea made
throughout Mexico during two successive
years, only parasites of the genus Crito-
gaster emerged from receptacles or were
found ovipositing into these figs. (2) In
56 collections of Urostigma figs made in
Mexico and Central America, 3 genera of
parasites were collected: Heterandriiim,
Physothorax, and Idarnes. Idarnes was al-
ways the most abundant. (3) At two lo-
calities (Izucar de Matamoros, Puebla,
Mexico and 12 miles west of Veracruz,
Mexico) species of Urostigma and Phar-
macosycea were growing sympatrically
and wasps were emerging from each fig
species. Idarnes aggregated on Urostigma
and Cri togas ter aggregated on Pharmaco-
sycea, even though the trees were sepa-
rated by only 15 to 20 meters. In personal
communications, Ramirez has indicated
his finding of Idarnes only in figs of the
subgenus Urostigma, not only in Mexico
and Central America but also in Vene-
zuela.

Host specificity is also indicated by a
study of ovipositor lengths. Ovipositors
must be long enough to penetrate the
syconial walls. Informal study, not sup-
ported by data on wall thicknesses, shows
that species of Idarnes with long oviposi-
tors inhabit figs with thick walls and those
with short ovipositors inhabit small figs
with thin walls. Obviously, the Idarnes
with short ovipositor could not successfully
infest figs with thick walls.

Recent authors (cf. Wiebes, Ramirez,
Hill, and Joseph, loc. cit.) have indicated
a high level of host specificity between
pollinator and species of Ficits. Although
the proof is not conclusive, a similar ten-
dency exists between Idarnes and Ficits.
Ten different species of Idarnes have been
collected from 10 different species of fig.
Moreover, where collections were made in
different areas from the same species of
fig, the wasps in all instances were pre-

416

The University of Kansas Science Bulletin

SLimably conspecific. At least they were
morphologically more similar to one an-
other than to different species of Idarnes
collected on diflferent fig species in the
same or adjacent areas.

It is entirely unknown whether host
specificity is based on host fig alone or also
on the pollinator species. Table 1 lists the
known Urostigma pollinators and their
parasites.

REDEFINITION OF IDARNES
Idarnes Walker, 1843

Idarnes Walker, 1843, Ann. Mag. Nat. Hist. 12:47,
9.

Type species: Idarnes airme Walker (monotypic).
Tetragonaspis Mayr, 1885, Verb. Zool. Bot. Gesell.

Wien 35.205, 9.

Type species: Tetragonaspis fhwicollis Mayr.
Ganosonia Mayr, 1885, Vcrh. Zool. Bot. Gesell.

Wien 35.204, $ .

Type species: Ganosoma rohusttim Mayr.
Idarnes Ashmead, F'04, Mem. Carnegie Mus. 1:238-

239; 305.

Female — Head, mesosoma metallic
green; gaster tan with metallic luster;
compound eye asetose, red; scape, pedicel
amber, remaining segments darker. Head
sculptured; antenna inserted below trans-
verse middle line of compound eyes;

scrobe cavity extending to anterior median
ocellus or vertex; inter-antennal ridge pres-
ent, but variable. Antenna 12-segmented
with 1 annulus or 13-segmented with 2
annuli; flagellomeres carinate and setose;
club variable, terminal protuberance pres-
ent. Mandible hi- or tridentate; stipes
with basal acuminate seta; maxillary pal-
pus 2-segmented, segment 1 with disc-
shaped sensillum along distal margin, ter-
minal segment with proximal acuminate
seta and distal margin excindate; labial
palp 1-segmented, ligula with 2 to 5 sen-
sory pegs; galea with numerous acuminate
setae.

Pronotum sculptured, collar-like; meso-
scutum sculptured, separated from scapula
by well developed notaulix; mesoscutellum
quadrate; forewing with parastigmal notch
sensillate; stigma sensillate; adstigmal
setae present; setal tracts conspicuous;
hindwing with submarginal vein only.
Fore leg with bifurcate calcar; basitarsus
with conspicuous strigil; middle leg with
coxa small, globular; middle femur slen-
der, with 1 or 2 subapical spurs; hind coxa
3 times as large as fore coxa; hind femur
setose, sculptured; hind tibia with 1 or 2
subapical spurs. Proepisterna sculptured,

Table 1. Neotropical Urostigma figs, pollinators (Blastophaga), and their known Idarnes

parasites.

Known

Tree

Pollinator

Parasite

Collections

distribution

Fictis aiirea Nuttall

B.

mexicana Grandi

carme Walker

2

West Indies, Florida

f.

cervantesiana Standi ey

bncatoma sp. nov.

1

Costa Rica

/•.

colitbrinae Standley

B.

orozcoi Ramirez

galbina sp. nov.

1

Costa Rica

F.

goldmanti Standley

oscrocata sp. nov.

2

Mexico, Costa Rica

F.

hemsleyana Standley

B.

tondiizi Grandi

barhigera sp. nov.

2

Mexico, Costa Rica

F.

isophelehia Standley

B.

iirhanae Ramirez

micheneri sp. nov.

1

Costa Rica

F.

jimenezii Standley

B.

jimenezii Grandi

jimenezi sp. nov.

1

Costa Rica

F.

lapitliijolia Liebmann

B.

agiiilari Grandi

simtis sp. nov.

1

Costa Rica

F.

ohttisijolia H. B. K.

B.

/loffmeyeri Grandi

ohttisijoliae sp. nov.

3

Mexico, Costa Rica

F.

ocrstediana Miguel

B.

standleyi Ramirez

cainini sp. nov.

2

Mexico, Costa Rica

F.

tuerckheimii Standley

B.
B.

niariae Ramirez
carlo si Ramirez

ashlocki ^P- nov.

4

Mexico, Costa Rica

F.

velutina Willd.

B.

torresi Grandi

/.

flaricollis Mayr

1

Brazil, Costa Rica

Morphology and Systematics of Neotropical Parasitic Fig Wasp

417

prepectus not completely separated from
mesepisternum; discrimen well developed;
anapleurite present.

Male — Head, mesosoma tan, depressed;
gaster white, globular; compound eye re-
duced to pigmented spot or absent; ocelli
absent; antenna 4- or 5-segmented, basal
segment with thick walls, segment 3 trans-
lucent; mandible bi- or tridentate, often
scythe-like; labiomaxillary complex absent.

Pronotum depressed, larger than re-
maining thoracic components; wings ab-
sent; coxae, femora disc-shaped, dorso-
ventrally flattened; tibiae with large con-
spicuous spines; basitarsi with apical spurs.

Propodeum reduced, demarcated from
metanotum by faint, transverse line, spira-
cles along antero-lateral margin; gaster
with tergites setose; posterior tergal mar-
gin often sinuate.

KEY TO SOME SPECIES OF
IDARNES

1. Head and mesosoma bronze; gaster
tan with metallic luster; antenna 12-
or 13-segmented; 3 ocelli and com-
pound eyes present; wings present;
ovipositor longer than remainder of
body Females 2

- Head and mesosoma tan; gaster
white; antenna with 4 or 5 segments;
ocelli, compound eyes, and wings ab-
sent Males 13

2. Mandible tridentate; clypeal margin
bilobed; antenna 13-segmented with 2
ring segments; mesoscutellum lightly
shagreened or smooth; posterior ter-
gal margin straight 3

- Mandible bidentate (except /. obtusi-
foltae); clypeal margin variable; an-
tenna 12-segmented with single ring
segment (except /. simus); mesoscu-
tellar sculpture variable; posterior
tergal margin strongly sinuate 6

3. Head prognathous; antennal scape
and pedicel strongly setose (Fig.
26); pedicel twice as long as wide;
flagellar setae not verticillate; club
compact; pronotum as large as meso-
scutum and scapulae combined; ad-
stigmal setae of forewing forming a
dense cluster (Fig. 54) /. bucatoma

- Head hypognathous; antennal scape
and pedicel bearing few setae; pedi-
cel not elongate; flagellomeres with
distinct whorls of setae hasally; club
not compact; pronotum smaller than
mesoscutum and scapulae combined;
forewing with fewer adstigmal setae
4

4. Vertex of head flat; frons weakly
rugose /. camini

- Vertex of head convex; frons smooth
5

5. Toruli just below middle line of com-
pound eyes; scrobe shallow, smooth;
clypeus bilobed; ovipositor 4.5 times
gaster length; gonostyli bearing only

a few small setae /. fiavicollis

- Toruli at lower margin of compound
eyes; scrobe cavity deep, rugose; clyp-
eus trilobed; ovipositor 3.5 times gas-
ter length; gonostyli bearing large
conspicuous setae along mesal sur-
face /. rnicheneri

6. Antennal flagellomeres, 2 times as
long as wide (Fig. 34); ovipositor 5
times gaster length /. simus

- Flagellomeres less than 2 times as
long as wide; ovipositor less than 5
times gaster length 7

7. Antenna inserted at or just below
middle line of compound eyes; meso-
scutellum lacunose or smooth; stigma

not tumid 8

Antenna inserted at level of ventral
margin of compound eyes; mesoscu-
tellum smooth; stigma tumid 9

8. Interantennal ridge acute; marginal
tract not extending to postmarginal
vein (Fig. 56); prepectus laterally
favose, mesally rugose; ovipositor 4
times gaster length; gonostylar setae
minute /. barbigera

- Interantennal ridge flat; marginal
tract reaching postmarginal vein
(Fig. 50); prepectus uniformly ru-
gose; ovipositor 2.5 times gaster
length; gonostylar setae conspicuous
/. galbina

9. Anal tract absent; ovipositor at least
3.0 times gaster length 10

- Anal tract present (Fig. 59); oviposi-
tor 2.5 times gaster length .... /. jimetjezi

10. Scape extending to vertex; frons
smooth; prepectus favose, pattern
fading mesally, mesepisternum later-
ally rugose, shagreened at discrimen;
ovipositor 6 times gaster length;

418

The University of Kansas Science Bulletin

gonostylar setae minute /. oscrocata

- Scape not reaching vertex; frons not
smooth (except /. cartne); ovipositor
several times more or less than 6
times gaster length 11

11. Vertex sinuous as seen from front;
frons, prepectus favose; mesepister-
num laterally favose, smooth at dis-
crimen; ovipositor 8 times gaster

length /. obtusijoliae

Vertex flat; frons not lacunose; pre-
jxctus rugose; mesepisternum later-
ally rugose; ovipositor considerably
less than 8 times gaster length 12

12. Frons favose; terga smooth; oviposi-
tor 4 times gaster length /. ashlocf(i

- Frons smooth; gastral terga sha-
greened; ovipositor 3 times gaster
length /. carme

13. Antenna 4-segmented (Fig. 43), seg-
ment 3 not translucent; legs with 4
tarsomeres (Figs. 116-118). .. /. bucatoma

- Antenna 4- or 5-segmented, segment
3 translucent; legs with 5 tarso-
meres 14

14. Antenna 4-segmented with no con-
striction at base of club 15

- Antenna 5-segmented or 4-segmented
with a constriction at base of club .... 18

15. Mandible tridentate, not scythe-like;

posterior tergal margin sinuate 16

Mandible bidentate, occasionally
scythe-like; posterior tergal margin
straight or sinuate 17

16. Antenna 4-segmented (Fig. 46), fla-
gellomeres moderately setose; middle
and hind tibiae with long apical

spurs (Figs. 105, 106) /. michenen

Antenna 4-segmented, flagellomeres
sparsely setose (Fig. 37); tibiae with-
out long apical spurs (Figs. 86-88)
/. galbina

17. Hind tibia with large apical spur;
middle and hind basitarsi spinose,
elongate (Figs. 99, 100); gastral

sterna with minute setae /. camini

Tibiae with apical spurs of inter-
mediate length; basitarsi with few

apical spines (Figs. 110-112)

/. jimenezi

18. Tibiae without large spurs 19

- Middle and/or hind tibiae with long
apical spurs 20

19. Hind basitarsus lacking apical spines,
nearly as long as tibia (Fig. 115);
clypeal margin with 2 small mesal
spines /. oscrocata

- Hind basitarsus with apical spines,
less than half as long as tibia (Fig.
91); clypeal margin lacking mesal
spines /• simus

20. Mandible tridentate (Fig. 75)

/. flavicollis

- Mandible bidentate 21

21. Mandible scythe-like; posterior tergal

margin straight 22

Mandible not scythe-like; posterior
tergal margin sinuate 23

22. Head with small setae uniformly dis-
tributed on dorsal surface; clypeal
margin straight; middle and hind
tibiae with apical spurs (Figs. 108,
109) /. barbigera

- Dorsal surface of head sparsely se-
tose; clypeal margin concave; tibiae
lacking long apical spurs (Figs. 95-
97) /. carme

23. Middle tibia with long apical spur
(Fig. 102) /. obtusijoliae

- Middle tibia without long apical spur
(Fig. 84) /. ashloc\i

SYSTEMATIC TREATMENT

All of the named species that fall in
Idarnes as here restricted are listed in this
section. Locations of type material and
voucher specimens for newly described
species or species treated in detail appear
in Table 2. Etymological origins of new
names are given in Table 3.

Characters are sequentially numbered
in each description for convenient com-
parison among descriptions. For example,
mandibular dentition may be ascertained
for females of all species by looking for
number 21 in each description.

Idarnes ashlocl{i, new species

Female — Body length 1.5 mm, oviposi-
tor 4.0 mm long. (1, 2) Head, mesosoma
metallic mossy green; (3) gaster bronze
with metallic luster; (4) scape, pedicel
smoky brown, remaining segments ebony;
(5) legs uniformly tawny; (6) prester-
num, mesosternum metallic green.

(7) Head hypognathous, in frontal as-
pect wider than long; compound eye
weakly protuberant; (8) entire head with

Morphology and Systematics of Neotropical Parasitic Fig Wasp

419

Table 2. Disposition of types and voucher specimens of Idarnes}^

K.U. U.S.N.M.N.H. Pret.

Lenin. B.M. Leiden Canberra Ramirez

iishIocl{i

9

H
A

1

9

P

9

P

harhigera

9

S

A

1

9

P

9

P

hiicatoma

9

H
A

1

9

P

9

P

c a mini

9

$

H
A

I

9

P

9

P

cunne

1

9

9

Nco

9

1

i

1

£

flavicollis

1

1

9

1

9

9

galhina

9

S

H
A

1

9

P

9

P

jimenezi

9

H
A

1
1

9
S

P

9

P

micheneri

9

i

H
A

I
1

9

P

9

P

obtitsifoUae

9

H
A

I
1

9

6

P

9

P

OS croc at a

9

H
A

1

9

P

9

P

simtis

9

6

H
A

1

9

P

9

P

I9P 19P 19P 19P 19P

1 9 P 19

1 9

1 9

9 P 1 9 P 1 9 P

19 P 19P I9P 19 P 19P

1 9 P 1 9 P 1 9 P

9 P 1 9 P

1 9

1 9

1 9

1 9

1 9

1 9

1 9

1 9

19P 19P 19P 19P 19P

19P 19P 19P 19P 19P

19P 19P 19P 19P 19P

19P 19P 19 P 19 P 19P

19P 19P 19P 19P 19P

19P 19P 19P 19P 19P

* Abbreviations used. Names in parentheses refer to individual to whom material was sent.

K.U. = Snow Entomological Museum, University of Kansas, Lawrence, Kansas (G. W. Byers).

U.S.N.M.N.H. = U.S. National Museum of Natural History, Washington, D.C. (B. D. Burks).

Pret. = Plant Protection Research Institute, Pretoria, Republic of South Africa (D. P. Annecke).

Lenin. = Zoological Institute, Academy of Science, Leningrad, U.S.S.R. (V. Trjapitzin).

B. M. = Bridsh Museum (Natural History), England (R. D. Eady).

Leiden = Rijksuniversiteit, Leiden, Netherlands (J. T. Wiebes).

Canberra = C.S.I.R.O., Canberra, Australia (E. Riek).

Ramirez = W. Ramirez.

H^Holotype; P^Paraty^x:; A = Allotype; Neo = Neotype.

Table 3. Etymological origin of new Idarnes species names.

Epithet

Origin

/. ashlocki
I. harhigera
I. hue atom a
1. catnini
I. galhina
I. jimenezi
I. micheneri
I. ohtusifoliae
I. oscrocata
I. simtis

Patronym of P. D. Ashlock, Hemipterist

barbigera = bearded

bu = large; catoma = shoulders

Patronym of J. H. Camin, Acarologist

galbina = jaundiced

Fictis /imcnezii

Patronym of C. D. Michener, Hymenopterist

Ficus ohtiisifolia

OS = mouth; crocatus = saffron yellow

simus = flat nosed

420

The University of Kansas Science Bulletin

bold, uniform favose sculpture; setae
sparse; (9) margin of vertex flat; (10) an-
tenna inserted at ventral margin of com-
pound eye; (11) toruli separated by twice
diameter of torulus; (12) scrobe with 2
favose, shallow channels converging at an-
terior ocellus; (13) interantennal ridge not
strongly convex. (14) Antenna 12-seg-
mented (Fig. 28); (15) scape shagreened,
setose, extending to median ocellus; (16)
pedicel shagreened, setose; (17) flagello-
meres as wide as long, verticillate; setae
inflexible; setae and carinae equal in num-
ber; (18) club weak; terminal protuber-
ance inconspicuous. (19) Frons favose,
asetose; (20) clypeal margin weakly bi-
lobed; (21) mandible (Fig. 63) tridentate,
tooth margin sharply incised. (22) Maxil-
lary palpus 2-segmented, palpifer not evi-
dent; (23) palpiger not evident; (24)
ligula with 2 sensory pegs; (25) paraglossa
enveloping 3/4 of ligular margin; (26)
galea strongly setose.

(27) Pronotum favose along postero-
dorsal margin, remainder shagreened; (28,
29, 30, 31) mesoscutum, scapula, parascu-
tellum and axilla favose; (32) mesoscutel-
lum lacunose; (33) meson of metanotum
favose, lateron smooth. (34) Forewing
(Fig. 51) with submarginal vein 3.5 times
as long as marginal vein; postmarginal
vein 2 times stigmal vein; (35) stigma
weakly tumid; (36) 2 adstigmal setae;
(37) marginal tract ending well before
postmarginal vein; Ri tract ending before
terminus of marginal tract; submarginal
tract reaching stigma; medial tract pres-
ent; discal, anal tracts absent. (38) Fore
and hind coxae lightly sculptured, middle
coxa smooth; (39) femora sculptured;
(40) middle and hind tibiae each with
single large apical spur; (41) fore basitar-
sus as long as tarsomere 2; middle basi-
tarsus as long as tarsomeres 2-5; hind basi-
tarsus as long as tarsomeres 2-4. (42)
Proepisternum bronze, lightly sculptured,
smaller than fore coxa, bearing single cra-

nial process; (43) prosternum triangular,
lightly shagreened; discrimen short; (44)
prepectus rugose; (45) mesepisternum
laterally rugose, pattern becoming sha-
greened at discrimen; (46) mesepimeron
uniformly rugose; (47) anapleurite lightly
shagreened.

(48) Propodeum 3.5 times as wide as
long; meson favose, with pattern laterally
evanescent; (49) gaster larger than head
and mesosoma; (51) sterna glabrous; (52)
ovipositor and gonostylus 4 times gastral
length; (53) gonostylar setae short.

Male— Body length 2.0 mm. (54, 55)
Head, mesosoma tan; mandible dark red;
(56) gaster milky white; (57) third an-
tennal segment translucent, remaining seg-
ments tan; (58) leg segments concolorous
with mesosoma.

(59) Head with lateral margin bearing
short setae; (60) mandible (Fig. 71) bi-
dentate, scythe-like; tooth margin not
sharply incised; (61) antenna (Fig. 39) 4-
segmented, basal portion of segment 4 con-
stricted, nearly forming a distinct segment;
(62) all legs with 5 tarsomeres (Figs. 83-
85) ; hind tibia with long apical spur, basi-
tarsi elongate. (63) Gastral terga smooth,
with long setae; posterior tergal margins
sinuate; (64) sterna smooth, asetose.

Material — Described from 16 females
and 3 males collected at San Rafael, He-
redia, Costa Rica on 23 May 1964 by Wil-
liam Ramirez; 14 females and 2 males
collected 7 miles northeast of Coscoma-
tepec, Veracruz, Mexico on 8 August
1969 by the author and Ramirez; 12 fe-
males and 1 male collected 10 August 1969
at Cordoba, Veracruz, Mexico by Rami-
rez; and 16 females and 4 males collected
by the author at Campeche, Campeche,
Mexico on 7 July 1970. In each instance
the host fig was Ficus titercJ^heimii
Standley.

Ficiis tuerchjieimn is distributed
throughout Central America and Mexico.
Standley (1917) indicates it is closely re-

Morphology and Systhmatics of Neotropical Parasitic Fig Wasp

421

lated to F. jimenezii, but differs in leaf
shape and venation.

Holotype — A dissected female mounted
in Hover's medium on 2 slides, each with
3 coverslips. Both hibels are inscribed:
"Idarnes ashlocl^i. San Rafael, Heredia,
Costa Rica. 23 May 1%4, leg. W. Ramirez,
ex. F. Uierc\heim'n. HOLOTYPE." Al-
lotype— One dissected male mounted in
Hoyer's medium on 1 slide under 3 cover-
slips. The label bears the same data as
indicated above. Paratypes — Nine females
mounted in the manner indicated for the
holotype above with a similar label in-
scription.

Variation — Idarnes ashlockj females are
quite constant in appearance. Some vari-
ation in the vertex shape (9), length of
setal tracts in the wings (37), number of
adstigmal setae (^6), and shape of the
prosternum (27) have been noted. Males
from the northern limit of the species may
have scythe-like mandibles (60).

Con7parative Comments— I. ashloc\i
seems most nearly related to /. obtnsifoliae,
but may be distinguished from this and
other species of Idarnes on the basis of the
following characters: females with favose
sculpture on head (8) and mesosoma (2<S,
29, 30, 31), antenna inserted at ventral
margin of compound eye (10), wings with
short marginal and Ri vein tracts (37),
and mesepimeron rugose (46).

Idarnes barbigera, new species

Female^Y^ody length 1.8 mm, oviposi-
tor 4.0 mm long. (1, 2) Head and meso-
soma brown with tawny luster; margin of
oral fossa light; (3) gaster chestnut brown;
(4) scape tan, remaining segments rust
red; (5) fore coxa basally tan, remaining
segments dull translucent white; (6) pro-
sternum translucent; mesosternum concol-
orous with gastral sterna.

(7) Head hypognathous, in frontal as-
pect wider than long; compound eye pro-
tuberant; (8) gena setose, smooth; face
rugose with pattern progressively bolder

dorsally, vertex favose; (9) vertex sinuate;

(10) antenna inserted just below middle
line of compound eye; antennifer large;

(11) toruli separated by single torulus di-
ameter; (12) scrobe deep, lacunose, ex-
tending to anterior ocellus; (13) inter-
antennal ridge acute. (14) Antenna (Fig.
25) 12-segmented; (15) scape with minute
setae along dorsal and mesal surface ex-
tending to vertex; (16) pedicel elongate,
smooth, with few setae; (17) flagellomeres
slightly longer than wide, verticillate, setae
inflexible; setae more numerous than cari-
nae; (IS) club terminal protuberance
weakly formed. (19) Frons glabrous with
marginal setae; (20) clypeal margin with
single lobe; (21) mandible bidentate (Fig.
61), tooth margin sharply incised. (22)
Maxillary palpus 2-segmented; palpifer
not evident; (23) palpiger not evident;
(24) ligula bearing 4 sensory pegs; (25)
paraglossa enveloping 5/6 of ligular mar-
gin; (26) galea with moderate number of
acuminate setae.

(27) Pronotum postero-dorsally favose,
remainder smooth; (28, 29, 30, 31) meso-
scutum, scapula, axilla, parascutellum fa-
vose; (32) mesoscutellum smooth; (33)
meson of metanotum lacunose, lateron
smooth. (34) Forewing (Fig. 56) with
submarginal vein 3 times as long as mar-
ginal vein; postmarginal vein 2 times stig-
mal vein; (35) stigma weakly tumid, sen-
sillae small; (36) 2 short adstigmal setae;
(37) marginal tract ending well before
postmarginal vein; Ri tract absent; sub-
marginal tract reaching stigma; discal
tract present; median and anal tracts ab-
sent. (38) Hind coxa sculptured, remain-
ing coxae smooth; (39) middle femur
with setal tract; (40) middle tibia with
single tibial spur; hind tibia with 2 setal
tracts, 2 subequal apical spurs; (41) fore
basitarsus as long as tarsomeres 2-4; mid-
dle basitarsus as long as remaining tarso-
meres; hind basitarsus as long as tarso-
meres 2-4. (42) Proepisternum smaller

422

The University of Kansas Science Bulletin

than fore coxa, shagreened, with single
large cranial process; (43) prosternum
semicircular, glabrous; discrimen extend-
ing half length of structure; (44) pre-
pectus laterally favose, pattern rugose mes-
ally; (45) mesepisternum rugose laterally,
pattern fading mesally, glabrous at discri-
men; (46) mesepimeron lacunose; (47)
anapleurite lacunose.

(48) Propodeum 4 times as wide as
long, smooth, median longitudinal carina
absent; (49) gaster ovoid, slightly larger
than mesosoma; (50) tergum shagreened,
pattern lightly incised; posterior tergal
margins sinuate; (51) strena coriaceous,
weakly setose; (52) ovipositor and gono-
stylus 4 times gaster length; (53) gono-
stylar setae along mesal surfaces minute.

Male— body length 1.2 mm. (54, 55)
Head, mesosoma tan; (56) gaster milky
white; (57) antennal scape, base of seg-
ment 2 amber, club pale; (58) coxae, fe-
mora concolorous with mesosoma, tibiae
marginally red.

(59) Head with small setae uniformly
distributed over dorsal surface; clypeal
margin straight, bearing 2 moderately
sized acuminate setae; (60) mandible
(Fig. 80) scythe-Uke, bidentate; (61) an-
tenna (Fig. 44) 4-segmented; club spindle-
shaped; (62) legs with 5-segmented tarsi;
tibio-tarsal complex (Figs. 107-109) with
middle and hind basitarsi elongate; (63)
terga asetose, posterior tergal margins
straight; (64) sterna smooth.

Material — Described from 16 females
and 4 males collected at Sarchi Garcia,
Alajuela, Costa Rica on 17 June 1964 from
Ficiis hemsleyana by William Ramirez,
and 241 females collected by the author at
4 miles south of Rio Hondo, Oaxaca, Mex-
ico on 6 July 1970 from F. hemsleyana.

Condit (1969) indicates F. hemsleyana
is synonymous with F. citrifolia. Although
Standley (1917) did not have material
available from Mexico, the voucher speci-
men collected in 1970 agrees with Stand-

ley's description. DeWolf (1960) lists 29
synonomous names for F. citrifolia, among
which is F. heinsleyana. He cites the dis-
tribution as Florida to Paraguay. Among
the major synonyms are F. laevigata Vahl,
F. jiiditnciilata Dryand, F. pyrijolia Des-
fontaines and F. brevijolia Nuttall. Col-
lections of Idarnes and Blastophaga from
these figs should be compared with the
type description to examine host specificity
and the validity of DeWolf's synonomy.

Holotype — A dissected female mounted
in Hoyer's medium on 2 slides each bear-
ing 3 coverslips. Both labels are inscribed:
''Idarnes harbigera. Sarchi Garcia, Costa
Rica. 17 June 1964, leg. W. Ramirez, ex.
Ficiis hemsleyana. HOLOTYPE." Allo-
type— A single dissected male mounted in
Hoyer's medium on 1 slide with the parts
under 3 coverslips. The inscription on the
label is the same as above. Paratypes —
Eight females are balsam-mounted in the
manner indicated above, with the same
label inscription, and the parts of each
insect are on 1 slide.

Variation — All females from Mexico
have strongly sculptured fore coxae (38),
unlike those from Costa Rica. Individuals
collected in Mexico are virtually identical
in other features to those collected in Costa
Rica.

Comparative Comments — Females may
be recognized by the following key char-
acters: gena smooth and setose (9), frons
glabrous with marginal setae (19), smooth
mesoscutellum (32), cranial process of
proepisternum very large (42), and mes-
epimeron lacunose (46). Males are recog-
nized by the short setae on the dorsal sur-
face of the head (59), scythe-like, bidentate
mandible (60), 4-segmented antenna (61),
and asetose terga (63).

Idarnes brevicolis (Mayr)

Tctiagonuipis hrcficolis Mayr, 1885, Verh. Zool. Bot.
Gesell. Wicn 35:209, $.

Mayr described /. brevicolis from the
female only. I have not examined the type

Morphology and Systkmatics or Neotropical Parasitic Fig Wasp

423

material but it was collected in Brazil. It
is considered a species of Idanies as here
restricted.

Idarncs bitcatoma, new species
F<fWrt/(f— Body length 3.4 mm, f)viposi-
tor 3.7 mm long. (1, 2) Head and meso-
soma brassy green; (3) gaster reddish
dark brown with iridescent sheen; (4)
scape, pedicel tan with remaining segments
darker; (5) coxae concolorous with head
and mesosoma; fore femur entirely dark
brown, hind femur apically dark brown;
remaining segments tan; (6) prosternum,
mesosternum concolorous with notum.

(7) Head subjrirognathous, in frontal
aspect longer than wide; compound eye
not protuberant, or weakly so; (8) face
and gena rugose, asetose; (9) vertex
weakly sinuate seen from front; (10) an-
tenna inserted at level of lower third of
compound eye; (ll) toruli nearly contig-
uous; (12) scrobe evanescent, lightly sha-
greened, extending to anterior ocellus;
(13) interantennal ridge acute, short, (14)
Antenna 13-segmented (Fig. 26); (15)
scape elongate, setose, reaching anterior
ocellus; (16) pedicel elongate, setose; (17)
flagellar segments as long as wide, bearing
more carinae than setae; setae irregularly
dispersed, not verticillate; each seta with a
distinct socket; (18) club distinct; termi-
nal protuberance conspicuous. (19) Frons
smooth; (20) clypeal margin trilobed,
mesal lobe small; (21) mandible tridentate
(Fig. 64), tooth margin sharply incised.
(22) Maxillary palpus 2-segmented; pal-
pifer evident; {IS) palpiger not evident;
(24) ligula with 4 sensory pegs; (25) para-
glossa enveloping basal 3/4 of ligular mar-
gin; (26) galea with numerous acuminate
setae.

(27) Pronotum as large as mesoscutum
and scapulae combined, shagreened,
sparsely setose; (28, 29) mesoscutum, scap-
ula rugose; (30, 31) axilla, parascutellum
weakly rugose; {lil) mesoscutellum elutely
shagreened; (33) meson of metanotum

rugose, lateron with longitudinal plicae.
(34) Forewing (Fig. 54) with submar-
ginal vein 4 times as long as marginal
vein; stigmal vein 2 times postmarginal
vein; (35) stigma weakly tumid, sensilla
evident; (36) dense cluster of adstigmal
setae conspicuous; (37) marginal tract
ending well before postmarginal vein; Ri
tract ending before marginal tract; sub-
marginal tract not reaching stigma; distal,
medial, and anal tracts absent. (38, 39)
Coxae, femora all lightly sculptured; (40)
middle and hind tibiae with 2 subequal
apical spurs each; (41) fore basitarsus as
long as tarsomeres 2-4; middle and hind
basitarsal longer than tarsomeres 2-5. (42)
Proepisternum as large as fore coxa, ru-
gose, bearing 2 apparent cranial processes;
(43) prosternum heart-shaped, shagreened;
discrimen extending entire length; (44)
prepectus laterally favose, mesally rugose;

(45) mesepisternum laterally favose, mes-
ally shagreened, smooth at discrimen;

(46) mesepimeron rugose; (47) ana-
pleurite smooth.

(48) Propodeum 3 times as wide as
long, rugose, with pattern distorted later-
ally; median longitudinal carina weakly
formed; (49) gaster ovoid, slightly larger
than mesosoma; (50) tergum lightly sha-
greened, rear half of each tergum weakly
setose, posterior margins straight; (51)
sterna coriaceous; (52) ovipositor and
gonostylus 2.5 times gastral length; (53)
mesal surface of gonostylus setose basally,
becoming sparsely setose distally.

Ma/f— Body length 2.3 mm. (54, 55)
Head and mesosoma chestnut; (56) gaster
dirty white; (57) basal antennal segment
dark reddish-brown, segment 2 tan, re-
maining segments dirty white; (58) coxae,
femora concolorous with mesosoma; re-
maining segments castaneous.

(59) Head with dorsal surface lightly
setose; clypeal margin with single con-
spicuous lobe; labiomaxillary complex rep-
resented by 2 large sensory pegs; (60)

424

The University of Kansas Science Bulleti

N

mandible tridentate (Fig. 74), tooth mar-
gin sharply incised; third tooth weakly
developed; (61) antenna 4-segmented
(Fig. 43) ; segment 3 not translucent or
collar-like; (62) all legs with 4 tarsomeres;
tibio-tarsal complexes (Figs. 116-118) with
basitarsi elongate; (63) each tergum bear-
ing transverse row of long setae at center
axis of plate; posterior tergal margins
straight; (64) sterna smooth, asetose.

Material — Described from 27 females
and 9 males collected at Turrialba, Car-
tago Province, Costa Rica by William
Ramirez on 29 July 1964 from Picas
cervantesiana.

Holotype — A dissected female mounted
in Hoyer's medium on 3 slides under 9
coverslips. Each label is inscribed: "Idarnes
biicatoma. Turrialba, Cartago, Costa Rica.
29 July 1964, leg. W. Ramirez, ex. Ficiis
cervantesiana. HOLOTYPE." Allotype
— One dissected male mounted in Hoyer's
medium on 1 slide under 3 coverslips.
The label is inscribed with the same
data indicated above. Paratypes — Ten dis-
sected females mounted in Hoyer's as the
above holotype, with a similar label in-
scription.

Comparative Comments — This species
is readily identifiable since the female an-
tennal scape, pedicel, and second annulus
are strongly setose (15, 16), the head is
subprognathous (16), the pronotum (26)
is exceedingly large; the adstigmal setae
(37) of the forewing form a dense cluster,
and the propodeum (46) is sculptured and
bears a longitudinal carina. Males are rec-
ognized by the combination of tridentate
mandible (60), 4 tarsomeres on each leg
(62), and long tergal setae at the middle
line of each plate, and straight posterior
tergal margin (63).

Idarnes biicatoma does not appear
closely related to any other Idarnes species,
yet it clearly falls within the generic lim-
its. The subprognathous head in the fe-
male presumably represents a derived

character. The male is the only Idarnes
which has a pigmented third antennal seg-
ment and 4-segmented tarsi.

Idarnes camini, new species

Pemale—V>ody length 1.4 mm, oviposi-
tor 3.7 mm long. (1, 2) Head and meso-
soma chestnut brown with sheen, margin
of oral fossa light; (3) gaster brown with
tawny luster, posterior margins of terga
each bearing dark transverse band; (4)
antenna tawny; (5) leg segments uni-
formly gold; (6) presternum translucent,
mesosternum concolorous with head and
mesosoma.

(7) Head hypognathous; in frontal as-
pect nearly round; compound eye not pro-
tuberant; ((S) face weakly rugose; gena
strongly asetose; (9) vertex flat; (10) an-
tenna inserted at middle line of compound
eye; (11) toruli separated by diameter of
single torulus; (12) scrobe cavity shallow,
smooth, lateral walls carinate, extending to
anterior ocellus; (13) interantennal ridge
not acute. (14) Antenna 13-segmented
(Fig. 3)3)\ (15) scape extending to vertex;
(16) pedicel with few setae; (17) flagello-
meres longer than wide, verticillate; setae
inflexible, more numerous than carinae;
(1(S) club present; terminal protuberance
conspicuous. (19) Frons weakly sha-
greened, setose; (20) clypeal margin bi-
lobed; (21) mandible tridentate (Fig. 67),
tooth margin sharply incised. (22) Maxil-
lary palpus 2-segmented; palpifer well
formed; {!?>) palpiger not evident; (24)
ligula with 2 sensory pegs; (25) paraglossa
enveloping basal 1/2 of ligula; (26) galea
moderately setose.

(27) Pronotum smaller than scapulae
and mesoscutum, campanulate, uniformly
shagreened; (28, 29) mesoscutum, scapula
rugose; (30, 31) axilla, parascutellum
strongly shagreened; (32) mesoscutellum
lightly shagreened; {3>3>) meson of meta-
notum smooth, lateron Hghtly sculptured.
(34) Forewing (Fig. 55) with submargi-
nal vein 3 times as long as marginal vein;

Morphology and Systematics of Neotropical Parasitic Fig Wasp

425

stigmal vein 3 times postmarginal vein;

(35) stigma tumid, sensilla conspicuous;

(36) 5 adstigmal setae; (37) marginal
tract ending before postmarginal vein; Ri
tract extending to postmarginal vein; sub-
marginal tract reaching stigma; discal,
medial, anal tracts absent. (38) Fore and
hind coxae lightly sculptured; (39) femora
lightly setose; (40) middle tibia bearing 1
subapical spur; hind tibia setose, bearing
1 apical spur; (41) fore basitarsus as long
as tarsomeres 2-5. (42) Proepisternum
slightly larger than fore coxa, shagreened,
bearing single cranial process; (43) pro-
sternum triangular, smooth, setose along
posterior margin; furca broadly attached
to posterior margin of presternum; (44)
prepectus rugose; (45) mesepisternum lat-
erally shagreened, smooth at discrimen;

(46) mesepimeron weakly shagreened;

(47) anapleurite smooth.

(48) Propodeum 3.5 times as wide as
long, meson smooth, lateral region sha-
greened; (49) gaster ovoid; (50) terga
shagreened, setose; posterior tergal mar-
gins straight; (51) sterna coriaceous; (52)
ovipositor and gonostylus 3.5 times gas-
ter length; (53) gonostylar setae along
mesal surface large.

Male— Body length 1.1 mm. (54, 55)
Head, mesosoma tan; mandible red; (56)
gaster milky white; (57) antennal scape
red; (58) legs straw colored.

(59) Head with long setae along poste-
rior and lateral margins; (60) mandible
(Fig. 79) bidentate with tooth margin
sharply incised; (61) antenna (Fig. 38) 5-
segmented; (62) legs with 5 tarsomeres
(Figs. 98-100) ; middle and hind tibiae with
single long apical spur; fore basitarsus short,
middle and hind basitarsi spinose, elon-
gate. (63) Gastral terga asetose, posterior
margins sinuate; (64) sterna with minute
setae.

Material — Described from 24 females
and 3 males collected by William Ramirez
at La Virgen, Heredia, Costa Rica on 31

May 1964 and 9 females and 2 males col-
lected 15 June 1969 by the author and
Ramirez 7.1 miles northeast of Coscoma-
tepec, Veracruz, Mexico. Ficits oerstedi-
ana Mic]uel was the host tree in both
instances.

Ficiis oerstediana is distributed through-
out Mexico, Central America, and into
Colombia.

Holotype — One dissected female
mounted in Hoyer's medium on 2 slides
under 6 coverslips. Each label is inscribed :
''Idarnes camini. La Virgen, Heredia,
Costa Rica. 31 May 1964, leg W. Ramirez,
ex. Ficus oerstediana. HOLOTYPE." Al-
lotype— One dissected male mounted in
Hoyer's medium on a single slide under 3
coverslips. The label is inscribed as is the
holotype label. Paratypes — Eight dissected
females mounted in the manner indicated
above for the holotype. Each label is in-
scribed as above.

Variation — Some females have setae on
the vertex (9) ; the frons may be smooth
and glabrous (19); 4 ligular spines (24)
and 4 adstigmal setae (36) may be present.
The series of males was too small to show
variation.

Comparative Comments — Idarnes ca-
mini females may be recognized by the
combination of setose gena (8), distinctive
antennal appearance (14-18), mesoscutum
lightly shagreened (32), long postmargi-
nal vein (34), and large setae on the gono-
stylus (53). Males are identifiable by the
long setae on the head (59), bidentate
mandible (60), 5-segmented antenna (61),
and sinuous posterior tergal margins {63).

Idarnes carme Walker

Idarnes carme Walker, 18-13, Ann. Mag. Nat. Hist.
12:47, 9.

Female — Body length 1.5 mm, oviposi-
tor 2.2 mm long. (1, 2, 3) Head, meso-
soma, and gaster reddish brown with faint
sheen; (4) antenna concolorous with head;
(5) legs tan; (6) presternum nearly yel-
low, mesosternum reddish brown.

426

The University of Kansas Science Bulletin

I

(7) Head hypognathous, in frontal as-
pect distinctly wider than long; compound
eye not protuberant; (8) face lightly setose,
favose, pattern bolder toward vertex; gena
setose, lacunose; (9) vertex flat; (10) an-
tenna (Fig. 32) inserted at ventral margin
of compound eye; (11) toruli separated by
twice diameter of torulus; antennifer
small; (12) scrobe evanescent, lacunose,
not extending to anterior ocellus; (13) in-
terantennal ridge setose, broad, nearly
smooth; (14) antenna 12-segmented; (15)
scape short, weakly setose; (16) pedicel
not elongate, smooth, lightly setose; (17)
flagellomeres longer than wide, longitu-
dinal carinae more numerous than inflex-
ible setae; (18) club weakly formed; ter-
minal protuberance inconspicuous. (19)
Frons smooth, setose; (20) clypeal margin
straight; (21) mandible (Fig. 72) biden-
tate, tooth margin moderately incised.
(22) Maxillary palpus 2-segmented; pal-
pifer weak; (23) palpiger absent; (24)
ligula with 4 sensory pegs; (25) paraglossa
enveloping 3/4 of ligular margin; (26)
galea with numerous acuminate setae.

(27) Pronotum with postero-dorsal sur-
face favose, remainder shagreened; (28)
mesoscutum favose; (29) scapula lacunose,
weakly setose; (30, 31) axilla, parascutel-
lum rugose; axilla weakly setose; (32)
mesoscutellum lacunose, nearly smooth,
lateral margin setose; (^3) meson of meta-
notum lacunose, nearly smooth; lateron
shagreened. (34) Forewing (Fig. 53) with
submarginal vein 3 times as long as mar-
ginal vein; postmarginal vein over 2 times
stigmal vein; (35) stigma tumid, sensillae
large; (36) single adstigmal seta present;
(37) marginal tract ending before post-
marginal vein; Ri tract short; submargi-
nal tract reaching to stigma; discal, medial,
and anal tracts absent. (38) Coxae all
lightly shagreened, weakly setose; (39)
femora lightly shagreened, weakly setose;
(40) middle and hind tibiae each with
single apical spur; (41) fore basitarsus as

long as tarsomeres 2-3; middle and hind
basitarsi as long as tarsomeres 2-4. (42)
Proepisternum larger than forecoxa, sha-
greened, with single cranial process; (43)
prosternum hemispherical, smooth; discri-
men short; (44, 45) prepectus, mesepister-
num laterally rugose, mesally smooth;
(46) mesepimeron shagreened; (47) ana-
pleurite smooth.

(48) Propodeum 3 times as wide as
lone, smooth; median longitudinal carina
absent; (49) gaster as large as mesosoma;
(50) terga lightly shagreened, setose; pos-
terior tergal margins sinuate; (51) sterna
smooth; (52) ovipositor and gonostylus 3
times gaster length; (53) gonostylus with
setae along mesal surface, progressively
more numerous distally.

Male— Body length 0.8 mm. (54, 55)
Head, mesosoma tan; (56) gaster milky
white; (57) antenna pale, nearly white;
(58) legs concolorous with mesosoma,
tibial margins red.

(59) Head with few short setae along
dorsal surface; clypeal margin concave;
(60) mandible (Fig. 73) bidentate, tooth
margin not sharply incised; (61) antenna
(Fig. 40) 4-segmented, with few setae;
(62) all legs with 5 tarsomeres (Figs. 95-
97); tibiae lacking long apical spurs; basi-
tarsi elongate, with few apical spines. (63)
Gastral terga asetose, posterior tergal mar-
gins straight; (64) sterna smooth, asetose.

Material — Redescribed from 6 females
collected in part by H. H. Smith on St.
Vincent and Barbados, West Indies (date
of collection not indicated), and 27 fe-
males and 3 males collected by R. E. Beer
on 10 April 1968 from the Florida Keys.
In each instance the host tree was Ficus
aitrea.

Walker's type material should be in
the British Museum (Natural History)
but cannot now be located. Material bor-
rowed from the U.S. National Museum of
Natural History through the courtesy of
B. D. Burks had been collected in the

Morphology and SvsThMATics of Neotropical Parasitic Fig Wasp

427

West Indies and identified as Idurnes
curme by Ashmead and Girault. Walker
did not describe the male of this species,
nor are any males present in the U.S. Na-
tional Museum of Natural History from
the West Indies. The females collected by
Beer in Florida are considered conspecific
with the material from the West Indies.

The epithet "aurea" has been used to
describe varieties of Ficiis elastica Rox-
burgh and F. macrophylla Desfontaines.
However, these are not to be confused
with F. aurea Nuttall, the Florida Stran-
i^ling Fig. The species is found in south-
ern Florida and the Bahamas, in addition
to the West Indies.

Neotvpe—A dissected female mounted
in Canada balsam on 2 slides with 3 cover-
slips each. Both labels are inscribed:
"Idarnes carme Walker. St. Vincent, W.
I., female, leg. H. H. Smith 209. NEO-
TYPE."

Variation — Some variation may be
noted in the sculpturing of the female
head (8) and mesosoma (28-33). Males
were not collected in large enough num-
bers to detect variation.

Comparative Comments — I dames carme
may be distinguished from other species
by females with brunneous color (1, 2, 3),
antenna inserted at level of ventral margin
of compound eyes (10), shallow scrobe
(12), setose interantennal ridge (13),
mesosoma with setae on dorsal surface
(29-32), and setae of the gonostylus be-
coming progressively more numerous dis-
tally (53). Males have a bidentate man-
dible (60), 4-segmented antenna (61), and
straight posterior tergal margins (63).

Idarnes coriaria (Mayr)

Tetragonaspis coriaria Mayr, 1885, Verh. Zool. Bot.
Gesell. Wicn 35:209.

This species was adequately described
by Mayr from Brazilian material collected
bv Fritz Miiller. The male remains un-
described and the Ficus host is unknown.
The species belongs in Idarnes as here lim-

ited but differs from the more northern
species described herein.

Idarnes flavicollis (Mayr)

Tetragonaspis flavicollis Mayr, 1885, Verh. Zool. Bot.

Gesell. Wien 35:207-208, $.
Ganosoma rolntstttm Mayr, 1885, Verh. Zool. Bot.

Gesell. Wien 35:204, c? .

Female — Body length 2.5 mm, oviposi-
tor 5.0 mm long. (1, 2) Head, mesosoma
metallic dark green; (3) gaster smoky
brown; (4) scape basally tan, becoming
progressively darker distally; pedicel and
remaining segments reddish brown; (5)
fore and hind coxae tawny brown; remain-
ing leg segments dirty white; (6) prester-
num, mesosternum concolorous with head
and mesosoma.

(7) Head hypognathous, in frontal as-
pect oval, nearly round; (8) face progres-
sively more rugose above toruli; vertex
favose; gena rugose along margin of com-
pound eye, remainder smooth; genal sur-
face setose; (9) vertex convex; (10) an-
tenna (Fig. 29) inserted just below midline
of compound eye; (11) toruli separated by
1.5 times diameter of torulus; (12) scrobe
shallow, glabrous, terminated at anterior
ocellus; (13) interantennal ridge short.
(14) Antenna 13-segmented; (15) scape
setose, shagreened, extending beyond ver-
tex; (16) pedicel not elongate, setose; (17)
flagellomeres longer than wide, verticillate,
with flexible setae; (18) club not distinct;
terminal protuberance conspicuous. (19)
Frons smooth, asetose; (20) clypeal margin
bilobed; (21) mandible tridentate (Fig. 69),
tooth margin not sharply incised. (22)
Maxillary palpus 2-segmented; palpifer
evident; (23) palpiger evident; (24) ligula
with 4 sensory pegs; (25) paraglossa en-
veloping 2/3 of ligular margin; (26) galea
with numerous acuminate setae.

(27) Pronotum with postero-dorsal sur-
face rugose, remainder shagreened; (28,
29, 30, 31) mesoscutum, scapula, axilla,
parascutellum uniformly rugose; (32)
mesoscutellum shagreened with pattern
lightly incised; lateral margin asetose;

428

The University of Kansas Science Bulletin

(3)3) meson of metanotum rugose, lateron
smooth, setose. (34) Forewing (Fig. 48)
with sLibmarginal vein 3 times as long as
marginal vein; stigmal vein 2 times post-
marginal vein; (35) stigma tumid with
large sensillae; (36) adstigmal setae vari-
able; (37) marginal tract ending at post-
marginal vein; Ri tract absent; submargi-
nal tract reaching stigma; discal, medial,
anal tracts absent; (3S) all coxae sculp-
tured, setose; (39) femora setose; hind
femur and coxa subequal in size; (40)
middle tibia with single apical spur; hind
tibia with 2 subequal apical spurs; (41)
fore basitarsus as long as tarsomeres 2-3;
middle basitarsus longer than remaining
tarsomeres; hind basitarsus as long as tar-
someres 2-5. (42) Proepisternum as large
as fore coxa, asetose, rugose, bearing single
elongate cranial process; (43) prosternum
subtrapezoidal, rugose; discrimen ex-
tending to center; (44) prepectus laterally
rugose, pattern fading mesally; (45)
mesepisternum uniformly shagreened;
spinasternum smooth, well developed;
(46) mesepimeron laterally rugose, sha-
greened at pleural suture; (47) anapleurite
smooth.

(48) Propodeum 3 times as wide as
long, smooth; area antero-lateral to spira-
cles setose; median longitudinal carina ab-
sent; (49) gaster larger than head and
mesosoma combined; (50) terga lightly
shagreened; posterior half of each tergum
uniformly setose; posterior tergal margins
straight; (51) sterna smooth, glabrous;

(52) ovipositor 4.5 times gaster length;

(53) gonostylus with few small setae along
mesal surface.

Male— Body length 1.9 mm. (54, 55)
Head, mesosoma tan; (56) gaster pale,
nearly dirty white; (57) antenna tan; (58)
legs concolorous with mesosoma, tibial
margins reddish.

(59) Dorsal surface of head smooth;
clypeal margin weakly concave; (60) man-
dible tridentate (Fig. 75), front tooth mar-

gin sharply incised; (61) antenna (Fig.
47) 5-segmented; (62) each leg with 5
tarsomeres; tibio-tarsal complex (Figs. 92-
94) with fore basitarsus short, middle and
hind tibiae with long spurs. (63) Gastral
terga asetose, posterior margins straight;
(64) sterna smooth.

Material — Twenty-one female and 8
male wasps collected at La Caiiada, Car-
tago, Costa Rica by William Ramirez on
31 January 1963 from Ficiis vehitina Willd.

Comparative Comments — Superficially
this species resembles Idarnes simiis, but it
may be distinguished on the basis of the
following female characters: /. favicollis
has a rugose face (8), the antenna 13-
segmented (14), frons asetose (19), lateral
margin of mesoscutellum asetose (32), lat-
eron of metanotum setose (33), the stig-
mal vein twice as long as the postmarginal
vein (34), the prosternum subtrapezoidal
in shape (43), the anapleurite smooth
(47), and the posterior tergal margins
straight (50).

Mayr (1885) described Ganosoma ro-
biistiim from Brazil as developing in re-
ceptacles of the same tree as Idarnes flavi-
collis. The material from Costa Rica
reveals two distinct male morphs, neither
of which conforms to G. robiistiim very
well. The marked male dimorphism may
be due to nutrition, which could also ac-
count for the discrepancy between the
material examined from Costa Rica and
Mayr's G. robustum. Since the larger form
is more common, and since the smaller
foriTL could conceivably be the male of
another symbiont, I have described above
only the larger of the two morphs. The
problem of association of sexes is still not
satisfactorily solved, and the Costa Rican
material may represent a new species. The
female from Costa Rica, however, agrees
with /. flavicollis from Brazil.

Mayr failed to give the species of fig
from which Idarnes flavicollis emerged.
Ficits veliitina Willd. if found throughout

Morphology and Systematics of Neotropical Parasitic Fig Wasp

429

Central America, but it is uncertain

whether or not it grows in South America.

Type Locality — Santa Catarina, l^razil.

Idarnes jorticornis (Mayr)

Tctragonaspis jorticornis Ma\r, 1S85, Vcrli. Zool.
Hot. Gcscll. Wicn 35:208, 9.

This species was described from mate-
rial collected in Brazil. The male remains
undescribed and the Ficiis host unre-
corded. This is a species of Idarnes as
here restricted.

Idarnes galbina, new species
Female — Body length 1.0 mm, oviposi-
tor 1.9 mm long. (1, 2, 3) Head, meso-
soma, gaster shining brown; (4) antenna
concolorous with body; (5) legs uniformly
straw colored; (6) presternum chestnut
brown; mesosternum shining brown.

(7) Head hypognathous, in frontal as-
pect nearly round; compound eye weakly
protuberant; (8) face shagreened at toruli,
pattern becoming progressively bolder to-
ward vertex; (9) vertex convex; gena his-
pidous, smooth; (10) antenna inserted just
below midline of compound eye; (11)
toruli separated by 2 times torulus diam-
eter; (12) scrobe cavity short, not reaching
anterior ocellus; lateral walls carinate;
(13) interantennal ridge broad, flat. (14)
Antenna 12-segmented (Fig. 31); (15)
scape weakly setose, smooth, reaching
lower margin of anterior ocellus; (16)
pedicel with very few setae; (17) flagello-
meres as long as wide, carinae long; cari-
nae more numerous than setae; (18) club
nearly absent; terminal protuberance not
conspicuous. (19) Frons lightly shag-
reened, marginally setose; (20) clypeal
margin convex; (21) mandible bidentate
(Fig. 68), tooth margin not sharply in-
cised. (22) Maxillary palpus 2-segmented;
palpifer not evident; (23) palpiger not evi-
dent; (24) ligula with 2 sensory pegs; (25)
paraglossa enveloping entire ligular mar-
gin; (26) galea with few acuminate setae.

(27) Pronotum lightly shagreened;

(28) mesoscutum favose along scuto-scu-
tellar suture, pattern distorted cephalad;
(29, 30, 31) axilla, scapula, parascutellum
rugose; (32) mesoscutellum glabrous;
(33) meson of metanotum lacunose, lat-
eron weakly sculptured. (34) Forewing
(Fig. 50) with submarginal vein 3 times
as long as marginal vein; postmarginal
vein 2 times stigmal vein; (35) stigma not
tumid, sensillae small; (36) 2 adstigmal
setae; (37) marginal tract ending at post-
marginal vein; Ri tract present, ending
behind postmarginal vein; submarginal
tract not reaching stigma; discal, anal
tracts absent; medial tract present. (38)
Fore coxa lightly shagreened, hind coxa
rugose; (39) femora shagreened, setose;
(40) middle tibia with medial setal tract,
single apical spur; hind tibia bearing setal
tract, several subequal apical spurs; (41)
fore basitarsus as long as tarsomeres 2-3;
middle and hind basitarsi as long as tarso-
meres 2-4. (42) Proepisternum as large as
fore coxa, glabrous, bearing single elon-
gate cranial process; (43) prosternum
nearly round, glabrous; discrimen only at
base of structure; (44) prepectus rugose;
(45) mesepisternum shagreened, pattern
evanescent at discrimen; (46) mesepimeron
uniformly shagreened; (47) anapleurite
smooth.

(48) Propodeum 5 times as wide as
long, lacunose; median longitudinal carina
absent; (49) gaster ovoid; (50) posterior
tergal margins sinuate; each tergum with
transverse row of small, inconspicuous se-
tae; (51) sterna smooth; (52) ovipositor
and gonostylus 2.5 times gaster length;
(53) gonostylus with setae becoming pro-
gressively longer, more numerous distally.

Male— Body length 0.9 mm. (54, 55)
Head and mesosoma chestnut; (56) gaster
dirty white; (57) scape light tan, club
white; (58) all leg segments concolorous
with head and mesosoma.

(59) Head with postero-dorsal surface
setose; clypeal margin sinuate; (60) man-

430

The University of Kansas Science Bulletin

dible tridentate (Fig. 77). tooth margin
moderately incised; (61) antenna 4-seg-
mented (Fig. 37), inserted at base of man-
dible; segment 3 translucent, collar-like;
lateral walls of scape not thick; (62) all
legs with 5 tarsomeres; tibio-tarsal com-
plexes (Figs. 86-88) as figured; (63) each
tergum bearing transverse row of very
long setae; posterior tergal margins sinu-
ate; (64) sterna smooth.

Material — Described from 42 females
and 16 males collected at Puerto Viejo,
Heredia, Costa Rica, by William Ramirez
on 16 May 1964 from Ficiis colubrinae.

Holotype — A dissected female mounted
in Hoyer's medium on 2 slides, each bear-
ing 3 coverslips. Both labels are inscribed:
"Idarnes galbina. Puerto Viejo, Heredia,
Costa Rica. 16 May 1964, leg. W. Ramirez,
ex. Ficiis colubrinae. HOLOTYPE." Al-
lotype— A single male, dissected, on 1 slide
with the parts under 3 coverslips. The in-
scription on the label is the same as above.
Paratypes — Eight females, Canada balsam
or Hoyer's medium, mounted in the man-
ner indicated above for the holotype, with
a similar label inscription.

Variation — While this species is exceed-
ingly small, occasionally a large (1.4 mm)
female may be found; the antennal scape
may bear some setae (15), the ligula has a
variable number of sensory pegs (24), and
the mesoscutellum may bear a shagreened
sculpture pattern (32). Males also exhibit
some variation in size, and in setal num-
ber on the antennal club (61).

Comparative Comments — This species
of Idarnes may be distinguished from oth-
ers on the basis of the following key fe-
male characters: smooth, hispidous gena
(8), 12-segmented antenna (13), flagello-
meres with more carinae than setae (17),
clypeal margin convex (20), mesoscutum
with distorted favose sculpture (28), me-
dial vein tract present {37), and a thin,
transverse, lacunose propodeum. Males
may bear long setae on the dorsum of the

head (59), and antenna (61) may appear
5-segmented, with the base of the club
constricted.

Idarnes galbina seems most closely re-
lated to /. carme. The female antennae of
both species are similar, the sculpturing
patterns of the mesosoma overlap, and the
labiomaxillary complexes resemble one
another.

Idarnes gracilicornis (Mayr)

Ganosoma atteniiatiim Mayr, 1885, Verh. Zool. Bot.

Gescll. Wicn 35:204, ^ .
Tetragonaspis gracilicornis Mayr, 1885, Verh. Zool.

Bot. Gcsell. Wien 35:208, $.

Ashmead (1904) originally associated
the female T. gracilicornis with the male
G. atteniiatnm , but oflfered no reason for
doing so. Thus some question remains as
to the correct association between the spe-
cies synonomized under /. gracilicornis. I
have not had an opportunity to examine
the material Ashmead used and from the
original descriptions it is not possible to
correctly associate the sexes.

Idarnes jimenezi, new species
Fetnale — Body length 2.1 mm, oviposi-
tor 2.3 mm long. (1, 2) Head, mesosoma
dark metallic green; (3) gaster rusty red;
(4) antenna reddish brown; (5) coxae
concolorous with mesosoma; remaining
leg segments tan; (6) prosternum gold.

(7) Head hypognathous, in frontal as-
pect nearly round; eye weakly protuber-
ant; (8) vertex favose, pattern evanescent
over remainder of head; setae irregularly
dispersed over face; (9) margin of vertex
strongly sinuate; (10) antenna inserted at
ventral margin of compound eye; (11)
toruli separated by diameter of torulus;
(12) scrobe cavity deep, lacunose, lateral
margins carinate; (13) interantennal ridge
acute. (14) Antenna 12-segmented (Fig.
36); (15) scape extending to anterior ocel-
lus; (16) pedicel not elongate, setose; (17)
flagellomeres slightly longer than wide,
verticillate, setae inflexible; setae, carinae
equal in number; (18) club absent; ter-

Morphology and Systematics of Neotropical Parasitic Fig Wasp

431

niinal protuberance inconspicuous. (19)
Frons favose, glabrous; (20) clypcal mar-
gin straight; (21) mandible bidentate
(Fig. 65), tooth margin deeply incised.
(22) Maxillary palpus 2-segmented; pal-
[lifcr small; (23) palpiger not evident; (24)
ligula very reduced, bearing 2 sensory
pegs; (25) paraglossa enveloping basal 3/4
of ligular margin; (26) galea with mod-
erate number of acuminate setae.

' (27) Pronotum favose along postero-

dorsal margin; (28) mesoscutum rugose,
pattern evanescent along posterior margin
near scapula; (29, 30) scapula, axilla sha-
greened; (31) parascutellum rugose; (32)
mesoscutellum smooth; (33) meson of
metanotum rugose, lateron smooth. (34)
Forewing (Fig. 59) with submarginal vein
3 times as long as marginal vein; post-
marginal vein 1.5 times stigmal vein; (35)
stigma tumid, with small sensilla; (36) 2
adstigmal setae; (37) marginal tract end-
ing at postmarginal vein; Ri tract absent;

I submarginal tract not reaching stigma;
discal tract absent; medial, anal tracts pres-

[ ent. (38, 39) Coxae, femora shagreened,
setose; (40) middle tibia with single apical
spur; hind tibia with 2 subequal large
apical spurs; (41) fore basitarsus as long
as tarsomeres 2-3; middle basitarsus as long
as tarsomeres 2-4; hind basitarsus as long
as tarsomeres 2-3. (42) Proepisternum as
large as fore coxa, shagreened, bearing sin-
gle cranial process; (43) prosternum semi-
circular, glabrous; discrimen long; (44)
prepectus rugose, pattern evanescent mes-
ally; (45) mesepisternum shagreened, pat-
tern absent at discrimen; (46) mesepi-
meron rugose; (47) anapleurite smooth.

(48) Propodeum 3 times as wide as
long, meson favose, lateral regions smooth;
median longitudinal carina absent; (49)
gaster larger than head and thorax; (50)
tcrgum glabrous, posterior tergal margins
sinuate; (51) sterna glabrous; (52) ovi-
positor and gonostylus 2.5 times gaster

length; (53) gonostylus glabrous or bear-
ing minute setae mesally.

A/^//<f— Body length 1.8 mm. (54, 55)
Head and mesosoma amber; (56) gaster
white; (57) scape, pedicel reddish brown;
segment 3 translucent; remaining seg-
ments tan. (58) Coxae concolorous with
mesosoma; tibiae lacteous.

(59) Head with short setae; clypeal
margin arched inward, with single large
mesal spine; (60) mandible (Fig. 78)
scythe-like, bidentate, tooth margin sharply
incised; (61) antenna (Fig. 41) 4-seg-
mented, basal part of segment 4 nearly a
distinct segment; (62) all legs with 5 tar-
someres; tibio-tarsal complexes as figured
(Figs. 110-112), basitarsi with apical
spines. (63) Each tergum bearing a few
long setae; posterior margins sinuate; (64)
sterna smooth, asetose.

Material — Described from 31 females
and 6 males collected by William Ramirez
on 24 August 1964 at Santa Domingo,
Heredia, Costa Rica from Ficus jimenezii.

Holotype — One dissected female
mounted in Hoyer's medium on 2 slides
under 6 coverslips. Each label is inscribed:
"Idarnes jimenezi. Santa Domingo, He-
redia, Costa Rica. 24 August 1964, leg. W.
Ramirez, ex. Ficus jimenezii. HOLO-
TYPE." Allotype— A dissected male
mounted in Hoyer's medium under 3
coverslips on 1 slide. The label is inscribed
as indicated above. Paratypes — Eight dis-
sected females each mounted on 1 slide
under 3 coverslips in Canada balsam. Each
label is inscribed as indicated above.

Variation — Variations in the facial
sculpturing pattern (8. 19), setae on the
coxae and femora (38, 39), and propodeal
sculpture pattern (48) are exhibited by fe-
males. Males were not collected in large
enough numbers to determine the extent
of variation.

Comparative Comments — Females may
be distinguished from other species of
Idarnes on the basis of the following char-

432

The University of Kansas Science Bulletin

acters: antennal configuration (14-18),
mesoscutellum smooth (32), and the pres-
ence of an anal vein tract (37). All males
examined have the scythe-like mandible
(60) and a 4-segmented antenna (61) in
which the club base is constricted.

I dames michenen , new species

Female — Body length 1.8 mm, oviposi-
tor 3.9 mm long. (1, 2, 3) Head, meso-
some, gaster brassy green; margin of oral
fossa tan; (4) antenna uniformly dark
brown; (5) all coxae, hind femur shining
brown; remaining segments tan; (6) pro-
sternum, mesosternum fusco-rufeous.

(7) Head verging on prognathous, in
frontal aspect nearly round; (S) face pro-
gressively more rugose above toruli; gena
smooth, setose; (9) vertex weakly convex;
(10) antenna inserted just above ventral
margin of compound eye; (11) toruli sep-
arated by twice diameter of torulus; (12)
scrobe cavity moderately deep, extending
to anterior ocellus at level of upper 1/3 of
compound eye; (13) interantennal ridge
not acute. (14) Antenna (Fig. 27) 13-
segmented; (15) scape lightly setose, ex-
tending to anterior ocellus; (16) pedicel
lightly setose, elongate; (17) flagellomeres
as long as wide, verticillate, setae inflex-
ible; (18) club present; terminal protuber-
ance not conspicuous. (19) Frons smooth,
often lateral margin setose; (20) clypeus
trilobed; (21) mandible tridentate (Fig.
60), tooth margin not sharply incised. (22)
Maxillary palpus 2-segmented; palpifer
conspicuous; (23) palpiger large; (24) li-
gula with 4 sensory pegs; (25) paraglossa
enveloping basal 1/3 of ligular margin;
(26) galea with numerous acuminate setae.

(27) Pronotum postero-dorsally rugose,
remainder shagreened; (28, 29) mesoscu-
tum, scapula rugose; (30, 31) axilla, para-
scutellum shagreened; (32) mesoscutellum
eleutely shagreened; (M) meson of meta-
notum shagreened, lateron smooth. (34)
Forewing (Fig. 58) with submarginal
vein 3 times as long as marginal vein;

stigmal vein 2 times postmarginal vein;
(35) stigma tumid, sensillae conspicuous;
(i6) 3 adstigmal setae present; (37) mar-
ginal tract reaching postmarginal vein; Ri
tract ending behind postmarginal vein;
submarginal tract reaching stigma; discal
tract conspicuous; medial, anal tracts ab-
sent. (38) Fore and hind coxae shagreened;
hind coxa larger than hind femur; (39)
femora setose, lightly shagreened; (40)
middle and hind tibiae each with 2 sub-
equal apical spurs; hind tibia with setal
tract, large apical spurs; (41) fore basi-
tarsus as long as tarsomeres 2-3; middle
and hind basitarsi as long as tarsomeres
2-5 of respective legs. (42) Proepisternum
larger than fore coxa, weakly shagreened,
setose along posterior margin, bearing sin-
gle elongate cranial process; (43) proster-
num campanulate, shagreened, setose; dis-
crimen extending cephalad 1/4 length of
structure; (44) prepectus rugose; (45)
mesepisternum shagreened; (46) mesepi-
meron rugose; (47) anapleurite smooth.

(48) Propodeum 3 times as wide as
long, eleutely shagreened, median longitu-
dinal carina absent; (49) gaster 2.5 times
longer than wide; (50) terga lightly setose,
strongly shagreened; posterior tergal mar-
gins straight; (51) sterna coriaceous; (52)
ovipositor and gonostylus 3.5 times gaster
length; (53) gonostylus bearing long con-
spicuous setae.

Male—Body length 1.4 mm. (54, 55)
Head and mesosoma tan; (56) gaster
white; (57) scape, pedicel tan with margin
reddish; (58) coxae, femora concolorous
with mesosoma; remaining segments red-
dish.

(59) Head with long conspicuous se-
tae; clypeal margin concave; (60) man-
dible tridentate (Fig. 76), tooth margin
sharply incised; (61) antenna 4-segmented
(Fig. 46); base of club nearly separated
from distal portion; (62) all legs with 5
tarsomeres (Figs. 104-106); middle and
hind tibiae each with 1 long apical spur;

Morphology and Systematics of Neotropical Parasitic Fig Wasp

433

basitarso spinose; (63) all terga smooth,
setose, posterior margins sinuate; (64)
sterna smooth, lightly setose.

Material— Described from 83 females
and 12 males collected by William Rami-
rez at Ciruelas, Alajuela, Costa Rica on 20
August 1964 from Ficus isophlebia.

Ficits isophlebia is related to F. ji?nen-
ezii according to Standley (1917). Both
species of Ficus grow throughout Central
America and northward into Mexico.

Holotvpe — One dissected female
mounted in Hover's medium on 2 slides
under 6 coverslips. Each label is inscribed:
"Idarnes micheneri. Ciruelas, Alajuela,
Costa Rica. 20 August 1964, leg. W. Rami-
rez, ex. Fiais isophlebia. HOLOTYPE."
Allotype — One dissected male mounted in
Hoyer's medium on 1 slide under 3 cover-
slips. The label is inscribed as indicated
above. Paratypes — Eight dissected females
mounted in Canada balsam or Hoyer's
medium, each on 2 slides under 6 cover-
slips. Each label is inscribed as above.

Variation — Some female wasps from
Mexico have the hind coxae and femora
bronze (4), the genae shagreened (7), a
minute mesal clypeal lobe (20), and fewer
than 4 adstigmal setae {'?>6). Males some-
times have a setose pronotum and some
variation in the development of setae on
the sternum of the gaster (64).

Comparative Comments — Idarnes mich-
eneri females may be recognized by the
rugose face ((S), 13-segmented antenna
(14), smooth, setose frons (19), tridentate
mandible (21), large labial palpiger (23),
well developed submarginal and discal
vein tracts (37), straight posterior tergal
margin (50), and long, conspicuous gono-
stylar setae (53). Males can be recognized
by the tridentate mandible (60), 4-seg-
mented antenna (61), and lightly setose
gastral sterna (64).

Idarnes obtiisifoliae, new species
Female — Body length 2.4 mm, oviposi-
tor 7.0 mm long. (1, 2) Head and meso-

soma brassy green; (3) gaster chestnut
brown; (4) antenna uniformly shining
brown; (5) coxal bases brassy green, legs
otherwise dirty white; (6) presternum
castaneous; mesosternum concolorous with
head and mesonotum.

(7) Head hypognathous, in frontal as-
pect slightly wider than long; compound
eye not protuberant; (8) face asetose, uni-
formly favose; gena favose; (9) vertex
sinuous; (10) antenna inserted just above
ventral margin of compound eye; (11)
toruli separated by 3 times diameter of
torulus; antennifer conspicuous; (12)
scrobe with 2 shallow channels, favose;
lateral walls not carinate; (13) interanten-
nal ridge not acute. (14) Antenna 12-seg-
mented (Fig. 30); (15) scape shagreened,
setose, elongate, but not reaching vertex;
(16) pedicel shagreened, setose; (17) fla-
irellomeres 1.5 times as long as wide, ver-
ticillate; setae inflexible, more numerous
than carinae; (18) club absent; terminal
protuberance evident. (19) Frons favose;
(20) clypeus bilobed, meson incised; (21)
mandible tridentate (Fig. 62), tooth mar-
gin sharply incised. (22) Maxillary palpus
2-segmented; palpifer evident; (23) pal-
piger not evident; (24) ligula with 2 sen-
sory pegs; (25) paraglossa enveloping en-
tire margin of ligula; (26) galea with
numerous acuminate setae.

(27) Pronotum with postero-dorsal sur-
face favose, remainder shagreened; (28,
29, 30, 31) mesoscutum, scapula, axilla,
parascutellum lacunose; (32) mesoscutel-
lum lacunose; (33) meson of metanotum
favose, lateral regions with numerous par-
allel plicae. (34) Forewing (Fig. 52) with
submarginal vein 4 times as long as mar-
ginal vein; postmarginal vein 2 times
stigmal vein; (35) stigma tumid, sensilla
small; (36) 1 adstigmal seta present; (37)
marginal tract not extending to postmar-
ginal vein; Ri tract absent; submarginal
tract reaching stigma; medial tract pres-
ent; discal, anal tracts absent. (38) Fore

434

Thb University of Kansas Science Bulletin

and middle coxae mesally setose; hind
coxa asetose; (39) middle femur shorter
than middle tibia; hind femur setose; (40)
middle tibia with single apical spur; hind
tibia with 2 mesal, longitudinal setal tracts,
2 subequal apical spurs; (41) fore basi-
tarsus as long as tarsomere 2; middle basi-
tarsus as long as tarsomeres 2-5; hind
basitarsus as long as tarsomeres 2-4. (42)
Proepisternum smaller than fore coxa, ru-
gose, bearing single elongate cranial proc-
ess; (43) prosternum oval, glabrous, lightly
setose; discrimen extending along basal
third of structure; (44) prepectus favose;
(45) mesepisternum laterally favose,
smooth at discrimen; (46) mesepimeron
rugose; (47) anapleurite favose.

(48) Propodeum 4 times as wide as
long, lacunose; medial longitudinal carina
present; (49) gaster ovoid; (50) tergum
shagreened, small setae irregularly distrib-
uted; posterior margin of each tergum
sinuate; (51) sterna smooth; (52) oviposi-
tor and gonostylus 8 times gaster length;
(53) gonostylus with small setae along
mesal surface.

Male— Body length 2.1 mm. (54, 55)
Head and mesosoma tan; (56) gaster
milky white; (57) scape tan with margin
reddish; segments 2, 4, and 5 tan; (58)
coxae, femora concolorous with mesosoma;
tibiae, tarsomeres reddish.

(59) Head prognathous; lateral margin
with short, inconspicuous setae; (60) man-
dible bidentate (Fig. 70), not scythe-like,
tooth margin not sharply incised; (61)
antenna 4-segmented (Fig. 45); clypeal
margin flexed inward; (62) all legs with
5 tarsomeres; tibio-tarsal complexes (Figs.
101-103) with elongate basitarsi, apical
basitarsal spines; (63) all gastral terga
smooth, setose; posterior margins sinuate;
(64) sterna smooth.

Material — Described from over 100 fe-
males and 38 males collected by William
Ramirez at Playon Aguirre, Puntarenas
Province, Costa Rica on 15 August 1964

from Ficiis obtitsi folia H. K. B.; by the
author and Ramirez from Coiclagus, Vera-
cruz, Mexico, 11 August 1969 on Ficits
obtitsi folia; and by the author at El Salto,
San Luis Potosi, Mexico, 12 June 1970,
also from Ficiis obtusifolia.

Ficus obtusifolia ranges throughout
Mexico and Central America. DeWolf
(1960) indicates Ficus bonplandiana
(Liebm.) Miquel is a synonym of Ficus
obtusifolia H. B. K. The association be-
tween Ficus obtusifolia and Idarnes ob-
tusifoliae appears to be one of the better
documented examples of host specificity
between torymid and fig, since the wasp
has been collected from this host at 3
widely separated localities on 3 separate
occasions.

Holotype— One dissected female
mounted in Hoyer's medium on 2 slides
under 6 coverslips. Each label is inscribed:
^'Idarnes obtusifoliae. Playon Aguirre,
Costa Rica. 15 August 1964, leg. W. Rami-
rez, ex. Ficus obtusifolia. HOLOTYPE."
This locality is inland from Parrita, Pun-
tarenas Province. Allotype — A dissected
male mounted on 1 slide under 3 cover-
slips in the manner indicated above and
similarly labeled. Paratypes — Eight dis-
sected females slide-mounted in the man-
ner indicated for the holotype above, with
labels inscribed in a similar manner.

Variation — Females from Mexico some-
times have a unilobed clypeus (20), an-
tennae inserted a little higher on the head
(10), head with flat vertex (9), and the
propodeal carina absent (48).

Comparative Comments — Idarnes ob-
tusifoliae may be easily recognized by the
following combination of female charac-
ters: Head uniformly favose (8), toruli
widely separated (11), 12-segmented an-
tenna (14), flagellomeres as illustrated
(17), and forewing with medial setal tract
(37). Males are difficuk to separate from
other species since some have a scythe-
like mandible (60), and some have an an-

Morphology and Systematics of Neotropical Parasitic Fig Wasp

435

tenna that appears 5-segmented (61). Sev-
eral large setae are arranged in transverse
rows along each tergum (63) and appear
to he a disthictive, constant male character.

Idarnes oscrocata, new species
Female— ?>ody length 1.9 mm, oviposi-
tor 5.0 mm long. (1, 2) Head and meso-
soma brassy green; margin o£ oral fossa
tan, extending to occipital foramen; (3)
gaster brown; (4) scape tan with luster;
pedicel brunneous; remaining antennal
segments bronze; (5) all leg segments tan;
(6) prosternum tan.

(7) Head hypognathous, in frontal as-
pect wider than long; compound eye mod-
erately protuberant; (8) head above toruli
favose; gena lacunose, weakly setose; (9)
vertex flat; (10) antenna inserted at mid-
line of compound eye; (11) toruli sepa-
rated by twice diameter of torulus; anten-
nifer conspicuous; (12) scrobe cavity deep,
favose, lateral walls carinate; (13) mter-
antennal ridge acute, short. (14) Antenna
12-segmented (Fig. 35); (15) scape asetose,
elongate, extending to vertex; (16) pedicel
asetose, 2 times longer than maximum
width; (17) flagellomeres 1.5 times longer
than wide; (18) club present but weak;
terminal protuberance conspicuous. (19)
Frons smooth, setose; (20) clypeal margin
straight; (21) mandible bidentate (Fig.
66), tooth margin moderately incised.
(22) Maxillary palpus 2-segmented; pal-
pifer evident; '(23) palpiger not evident;

(24) ligula elongate with 4 sensory pegs;

(25) paraglossa enveloping 2/3 of ligular
margin; (26) galea with numerous acumi-
nate setae.

(27) Pronotum postero-dorsally favose,
remainder shagreened; (28, 29, 30, 31)
mesoscutum, scapula, axilla, parascutellum
favose; (32) mesoscutellum lacunose; {^i)
meson of metanotum favose, lateron with
numerous longitudinal plicae. (34) Fore-
wing (Fig. 57) with submarginal vein 3.5
times as long as marginal vein; post-
marginal vein 2 times stigmal vein; (35)

stigma tumid, sensilla small; (36) 3 ad-
stigmal setae present; (37) marginal tract
with few microtrichae, ending well before
terminus of postmarginal vein; Ri tract
absent; submarginal tract reaching stigma;
discal, medial, anal tracts absent. (38)
Fore and hind coxae sculptured; middle
coxa smooth; (39) femora setose, hind fe-
mur smaller than hind coxa; (40) middle
tibia with single apical spur; hind tibia
with 2 subequal apical spurs; (41) fore
basitarsus slightly longer than tarsomere
2; middle basitarsus as long as tarsomeres
2-5; hind basitarsus as long as tarsomeres
2-4. (42) Proepisternum bronze, rugose,
slightly smaller than fore coxa, bearing
single elongate cranial process; (43) pro-
sternum heart-shaped, smooth, setose on
posterior quarter; (44) prepectus favose
laterally, smooth mesally; (45) mesepister-
num rugose laterally, shagreened mesally;
(46) mesepimeron smooth along pleural
suture, laterally rugose; (47) anapleurite

rugose.

(48) Propodeum 3 times as wide as
long, glabrous, asetose; median longitu-
dinal carina absent; (4Q) gaster slightly
larcrer than head and mesosoma combined;
(50) terga lightly shagreened, posterior
margins sinuate; (51) sterna coriaceous;
(52)" ovipositor and gonostylus 6 times
gaster length; (53) gonostylus bearing
small inconspicuous setae along mesal
surface.

Mrf/^— Body length 1.5 mm. (54, 55)
Head, mesosoma tan; (56) gaster light
tan, nearly white; (57) antenna pale; (58)
all leg segments concolorous with meso-
soma; tibial spines amber.

(59) Head with dorsal surface lightly
setose; clypeal margin concave, bearing 2
small mesal spines; (60) mandible biden-
tate (Fig. 82), tooth margin not sharply
incised; large spines on mesal surface of
mandible; (61) antenna (Fig. 42) 5-seg-
mented; (62) all legs with 5 tarsomeres;
tibio-tarsal complexes (Figs. 113-115) with

436

The University of Kansas Science Bulletin

tibiae lacking apical spurs; basitarsi lack-
ing thickened walls; (63) terga smooth,
asetose; posterior tergal margins sinuate;
(64) sterna smooth, asetose.

Material — Described from 6 females
collected on 6 August 1969 and 87 females
and 4 males collected on 20 August 1969,
2 miles north of Izucar de Matamoros,
Puebla, Mexico, along the Rio Balsas, by
the author and William Ramirez. In both
instances the host tree was Ficiis gold-
ma nil.

FiciiS goldmanii Standley ranges
throughout Mexico and into Central
America. In Puebla this species grows in
a riparian situation; in Sinaloa it was ob-
served growing in a xeric habitat. Idarnes
was not recovered from figs taken in
Sinaloa.

Holotvpe — One dissected female
mounted in Hoyer's medium on 2 slides
under 6 coverslips. Each label is inscribed:
''Idarnes oscrocata. 2 mi. N. Izucar de
Matamoros, Puebla, Mexico. 20 August
1969, leg. Gordh and Ramirez, ex. Ficiis
goldmanii. HOLOTYPE." Allotype—
One dissected male mounted in Hoyer's
medium under 3 coverslips on a single
slide, with the label inscribed as indicated
above. Paratypes — Eight dissected females,
each on one slide under 3 coverslips in
Canada balsam.

Variation — Some females have 2 sen-
sory pegs on the ligula (24), the frons may
be lacunose (19), the medial vein tract is
sometimes obscure (37), the metanotal
meson may be lacunose {i^), and the pro-
podeum may be mesally and postero-later-
ally rugose (48). Aside from the fourth
antennal segment being nearly fused to
the club (61), little variation was detected
in males.

Comparative Cojiiments — Females of
this species may easily be recognized by
the conspicuous tan margin of the oral
fossa (1), straight clypeal margin (20),
elongate ligula with 4 sensory pegs (24),

medial vein tract (37), rugose anapleurite
(47) and coriaceous sterna (51). Males
are identifiable by their bidentate mandi-
ble (60), 5-segmented antenna (61), and
asetose terga with sinuate posterior mar-
gins (63).

Idarnes oscrocata appears most closely
related to /. obtiisijoliae and may be dis-
tinguished from that species on the basis
of facial sculpture pattern (8), and clypeal
margin conformation (20).

Idarnes parallela (Mayr)

Gaiiosonia panillehim Mayr, 1885, Verh. Zool. Bot.
Cc-cll. Wien 35:204, '$ .

The female of /. parallela remains un-
described. From the original description
by Mayr, it is not possible to state with
certainty that a synonymy has not been
made in the present work or that a previ-
ously described species is not synonymous
with /. parallela. Until further collections
have been made in Brazil and /. parallela
females are collected and associated with
the males, positive statements regarding
the taxonomic status of /. parallela cannot
be made.

Idarnes punctata (Mayr)

Tctrugonaspis punctata Mayr, 1885, Verh. Zool. Bot.
Gesell.Wien 35:209, $.

The female of this species was ade-
quately described by Mayr, but the male
remains undescribed. It is a species of
Idarnes as here restricted, from Brazil.

Idarnes simiis, new species

Female — Body length 2.0 mm, oviposi-
tor 6.2 mm long. (1, 2) Head, mesosoma
bright metallic green; margin of oral fossa
pale; (3) gaster dorsally tawny; sterna tan
with pale transverse stripes; (4) scape
brownish yellow, remaining segments
chestnut brown; (6) prosternum tan;
mesosternum pale at discrimen, remainder
concolorous with head and mesonotum.

(7) Head hypognathous, in frontal as-
pect slightly wider than long; compound
eye not protuberant; (8) face favose above

Morphology and Svstematics of Neotropical Parasitic Fig Wasp

437

level of toruli; gena smooth, strongly
setose; (9) vertex weakly sinuate; (10)
antenna inserted at lower third of com-
pound eye; antennifer conspicuous; (11)
toruli separated by 3 times diameter of
torulus; (12) scrobe nearly absent, 2 widely
separated channels, short, smooth; (13)
interantennal ridge broad, not acute. (14)
Antenna (Fig. 34) 13-segmented; (15)
scape elongate, setose, extending to vertex;
(16) pedicel not elongate, bearing few se-
tae; (17) flagellomeres 2 times as long as
wide, verticillate; setae flexible, more nu-
merous than carinae; (18) club absent;
terminal protuberance conspicuous. (19)
Frons smooth, setose; (20) clypeal margin
straight; (21) mandible bidentate, tooth
margin sharply incised. (22) Maxillary
palpus 2-segmented; palpifer small; (23)
palpiger not evident; (24) ligula with 2
sensory pegs; (25) paraglossa enveloping
basal 2/3 of ligular margin; (26) galea
with numerous acuminate setae.

(27) Pronotum postero-dorsally favose,
remainder shagreened, pattern lightly in-
cised; (28, 29, 30, 31) mesoscutum, scapula,
axilla, parascutellum favose; (32) scutel-
lum lacunose, nearly smooth; (33) meson
of metanotum favose, lateron bearing sev-
eral longitudinal pHcae. (34) Forewing
(Fig. 49) with submarginal vein 3 times
as long as marginal vein; postmarginal
vein 2 times stigmal vein; (35) stigma
tumid, sensilla small; (36) 2 adstigmal
setae; (37) marginal tract ending at post-
marginal vein; Ri tract absent; submargi-
nal tract ending at stigma; discal, medial,
anal tracts absent. (38) Fore and hind
coxae lightly shagreened; (39) femora sha-
greened, setose; (40) middle tibia heavily
setose, bearing single apical spur; hind
tibia with setal tract, 2 subequal apical
spurs; (41) fore basitarsus as long as tarso-
meres 2-3; middle basitarsus longer than
remaining tarsomeres; hind basitarsus as
long as tarsomeres 2-4. (42) Proepisterna
smaller than fore coxa, shagreened, bear-

ing setae basally, single wide cranial proc-
ess; (43) prosternum nearly round,
smooth, bearing few setae; discrimen ex-
tending 1/3 the length of structure; (44)
prepectus favose mesally, smooth laterally;
(45) mesepisterna setose, laterally favose
with pattern centrally rugose, absent at
discrimen; (46) mesepimera favose; (47)
anapleurite rugose.

(48) Propodeum 4 times as wide as
long, mesally rugose, laterally smooth;
median longitudinal carina absent; (49)
gaster ovoid, slightly larger than head and
mesosoma combined; (50) terga lightly
shagreened; posterior tergal margins
strongly sinuate; (51) sterna coriaceous;
(52) ovipositor and gonostylus 5 times
gaster length; (53) gonostylus bearing
small setae along mesal surface.

Male— Body length 1.2 mm. (54, 55)
Head and mesosoma tan; (56) gaster
white; (57) antenna, except segment 3,
amber; (58) legs honey colored with tibial
margins red.

(59) Head with small setae sparsely
arranged over dorsal surface; clypeal mar-
gin strongly concave, without setae or
spines; (60) mandible bidentate (Fig. 81),
not scythe-like; (61) antenna (not figured)
5-segmented; (62) tarsi 5-segmented; tibiae
(Figs. 89-91) without large apical spurs;
basitarsi with few apical spurs; (63) terga
asetose; posterior tergal margins sinuate;
(64) sterna smooth, asetose.

Material — Described from 31 female
and 14 male specimens collected by Wil-
liam Ramirez at Rio Aguilar, Hatillo, San
Jose, Costa Rica on 14 June 1964 from
Ficiis lapithijolia. This species of fig is
widespread in Mexico and Central Amer-
ica.

Holotype — One dissected female
mounted in Hoyer's medium on 2 slides
under 5 coverslips. Each label is inscribed:
''Idarnes siwiis. Rio Aguilar, Hatillo,
Costa Rica. 14 June 1964,"^ leg. W. Rami-
rez, ex. Ficiis lapithijolia. HOLOTYPE."

438

The University of Kansas Science Bulletin

Allotype — One dissected male mounted on
1 slide under 3 coverslips in the manner
indicated above. Paratypes — Eight dis-
sected females, slide mounted in the man-
ner indicated for the holotype, with a sim-
ilar label inscription.

Variation — In some females the face
may be rugose at the level of the toruli
and become progressively more favose to-
wards the vertex (8), the entire pronotum
may be shagreened (27), the mesosternum
may be asetose (45, 46) or the anapleurite
may be shagreened (47). Males exhibit
some size variation, but appear remark-
ably constant morphologically. Occasion-
ally, a male may have a mandible verging
on scythe-like (60).

Comparative Comments — This wasp is
easily recognized by the following combi-
nation of female characters: widely sepa-
rated toruli (11), broad, flat interantennal
ridge (13), flexible, blunt setae on each
flagellomere (17), smooth frons bearing
setae (19), straight clypeal margin (20),
and mesopectus favose (44, 45, 46). Males
have small setae on the head (59), biden-
tate mandible (60), and asetose terga (63).

SPECIES INCORRECTLY
ATTRIBUTED TO IDARNES

The species listed below have at one
time or another been placed in Idarnes,
but this assignment is probably or certainly
incorrect as explained in each case.

Idarnes austyalis Froggatt, 1901), Agr. Gazette N.S.W.
11:452, ?.

The illustration of this species indicates
the forewing lacks setal tracts and adstig-
mal setae, and the hindwing bears a fully
developed marginal vein. Froggatt's de-
scription makes the species difficult to
place, but it has been recognized by Mayr
(1906) as Sycoryctes aitstralis. It is an
Australian species and for that reason
alone unlikely to fall in Idarnes as that
genus is here defined.

Otitcsella gtiaphalocarpac Risbec, 1951, Mem. Inst.
Fran^ais d'Afr. Noire 13:332, $.

I have not had the opportunity to ex-
amine this species. Wiebes (1970) regards
it as an Idarnes species. It was described
from Senegal emerging from Ficus
gnaphalocarpa and for these reasons prob-
ably does not belong to Idarnes as here
restricted,

Idarnes graalie Wiebes, 1968, Zool. Meded. 42:318,

Described from Africa, this species
bears a strong resemblance to Idarnes.
However, I consider the species extra-
limital not only because of its continent of
origin but because: (1) the forewing lacks
setal tracts; (2) the hindwing has a com-
pletely developed marginal vein; (3) the
terminal segment of the maxillary palpus
is not excindate; (4) an acuminate seta is
found on the first maxillary palpal seg-
ment rather than at the base of the ter-
minal one; and (5) as pictured, the female
lacks the discoid sensillum on the base of
the first palpal segment of the maxilla.

Idarnes nigra Risbec, 1951, Mein. Inst. Francais
d'Afr. Noire 13:123, 9,5.

Wiebes (1970) examined Risbec's ma-
terial and indicated that the females are
probably Sycoscapter and the males Apo-
crypta. The continent of origin is suffi-
cient to exclude the material from Idarnes
as here defined.

Idarnes oricntalis Walker, 1875, Entomologist 8:17,
9.

I have not had the opportunity to ex-
amine the material. Walker indicated /.
orientalis was recovered from Hindustan
and Ceylon. On the basis of distribution
it is excluded from Idarnes as here defined.
Grandi (1928) considered /. orientalis a
synonym of Sycoscapter stabilis.

Idarnes pteromaloides Walker, 1871, Notes on Chal-
cidiae 4:63, 9 .

Wiebes (1967) has considered the ma-
terial as Micransia pteromaloides (Walk-
er). The material described by Walker
came from Hindustan, and hence is ex-
cluded from Idarnes in the present work.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

439

hiarnes stahilis Walker, 1871, Notes on Chalcidiac
4:62, 9.

Wiebes (1967) has demonstrated that
/. stahilis is Sycoscapter stabilis (Walker).
The original material came from India
and is also excluded from Idarnes on the
basis of distribution.

Idaniomorplia siibaenea Girault, 1915, Mem. Queens-
land Mus. -1:282, 9, $.

hiarnes snbaenea Girault, 1917, Insec. Inscit. Menstr.
5:37, 9.

Girault (1917) considered Idarno-
morpha to be synonymous with Idarnes;
thus the Australian species snbaenea was
transferred to Idarnes. Girault considered
the antenna 13-segmented, counting the
terminal protuberance as a distinct seg-
ment. Girault indicates the female is 1.75
mm and the male is 2.8 mm long. In all
described species of Idarnes, the female is
larger than the male. If the sexes are cor-
rectly associated, a typographical error has
probably been made. Assuming the sexes
to be correctly associated, Idarnes snbaenea
must be considered extra-limital both geo-
graphically and morphologically. Idarno-
morpha is probably a good genus on the
basis of the male antenna being 3- and
tarsi 5-segmented. The wasps were reared
from figs with Ceratosolensis ficop/iaga
Girault. Idarnes has been shown in all
other instances to parasitize tropical Amer-
ican Blastophaga; hence, biological infor-
mation also suggests that Idarnomorpha is
a valid genus.

Idarnes transiens Walker, 1871, Notes on Chalcidiae
4:62, 9.

Idarnes transiens was described from
the female from Ceylon and India. Mor-
phologically, the species is clearly not
Idarnes, and Wiebes (1967) has called it
Ph ilotrypesis tra n si ens.

Tetragonaspis testacea Mayr, 1885, Verh. Zfxjl. Bot.

Gesell. Wien 35:157, 9.
Enkpebelea testacea Mayr, 1906, Wicn. Ent. Ztg.

25:165, 9.
Sycophagella agraensis Joseph, 1953, Agra Univ. }.

Sci. 2:54, 9,5.
Idarnes testacea (Mayr), Joseph, 1956, Ann. Soc. Ent.

Fr. 125:103, 9,6.

Illustrations do not accompany the
original description of Tetragonaspis tes-
tacea. Joseph (1953) described Sycopha-
gella agraensis from the receptacles of
Ficns glomerata collected in India and
illustrated the female. In 1956, he consid-
ered S. agraensis synonymous with Idarnes
testacea (Mayr). If Joseph's description is
correct, and if he has correctly associated
S. agraensis with T. testacea, then this
material cannot be considered Idarnes for
the following reasons: (1) the maxillary
palpus is unisegmented; (2) the forewing
lacks a parastigmal notch and sensilla; (3)
the forewing lacks setal tracts; and (4) the
hindwing bears a fully developed marginal
vein. Geographically, also, this species is
outside the known range of Idarnes as
here delimited.

LITERATURE CITED

AsHMEAD, W. H. 1904. Classification of the chalcid
flies. Mem. Carnegie Mus. 1:225-555.

Baker, C. F. 1913. A study caprification in
Ficits nota. Philippine J. Sci. (D) 8:63-83.

Beck, D. E. 1933. A morphological study of the
male genitalia of various genera of bees. Proc.
Utah Acad. Sci. 10:89-137.

Breland, O. p. 1938. Phylogeny of some callimomid
genera parasitic Hymenoptera). J. New York
Ent. Soc. 46:355-398.

Brothers, D. J., and V. C. Morax. 1969. A new
species of Tetrastichits Haliday, 1844 (Hy-
menoptera :Eulophidae) parasitic on the
nymphs of Patirocephala calodendri Moran
(Homoptera:Psyllidac). Proc. Roval Ent. Soc.
London (B) 38:40-46.

15UCHER, G. E. 1948. The anatomy of Monodonto-
mcriis dentipes Boh., an entomophagus chal-
cid. Canadian J. Res. 26:230-281.

Burks, B. D. 1938. A study of chalcidoid wings
(Hymenoptera). Ann. Ent. Soc. .America 31:
157-161.

. 1943. The North American parasitic wasps

of the genus Tetrastichits — a contribution to
biological control of insect pests. Proc. U.S.
Nat. Mus. 93:505-608.

. 1969. Redefinitions of two genera of chal-

cidoids from figs, with new Florida species
(Hymen.). Florida Ent. 52:115-122.

Butcher, F. G. 1964. The Florida fig wasp, Se-
ciindeisenia mexicana (Ashm.), and some of
its hymenopterous svmbionts. Florida Ent. 47:
235-238.

Carmin-, J., AND D. ScHEiNKiN. 1931. The fauna of
Palestinian plants: 2. Ticus sycomoriis Linne.
Bull. Soc. Royale Ent. d'Egypte 15:164-187.

Compere, H. 1962. The reality of sternites in the
mesothorax of Hymenoptera. Proc. Ent. Soc.
Washington 64:224-228.

440

The University of Kansas Science Bulletin

Compere, H., and D. Rosen. 1970. The prescutum
in Hymenoptera. Proc. Royal Ent. Soc. Lon-
don (A) 45:91-97.

CoNDiT, 1. J. 1969. Ficiis: The Exotic Species. Univ.
California (Berkeley) Div. Agri. Sci. 363 pp.

CoNDiT, 1. J., AND S. E. Flanders. 1945. "Gall-
flower" of the fig, a misnomer. Science 102:
129-130.

Corner, E. J. H. 1958. An introduction to the dis-
tribution of Fiiiis. Rcinwardtia 4:325-355.

. 1965. Check-list of Ficiis in Asia and Aus-
tralia, with keys to identification. Gardens
Bull. Singapore 21:1-186.

Croizat, L. 1952. Manual of Phytogeography. W.
Junk, Den Haag. 587 pp.

Cunningham, D. D. 1889. On the phenomena of
fertilization in Fiais roxhitrgliii Wall. Ann.
Royal Bot. Garden Calcutta 1:13-51.

Daly, H. V. 1964. Skeleto-muscular morphogenesis
in the thorax of the Hymenoptera. Proc. XII
Int. Congr. Ent. London: 151.

DeWolf, G. P., Jr. 1960. Fictts (Tourn.) L. Ann.
Missouri Bot. Garden 47:146-165.

DouTT, R., AND G. ViGGiANi. 1968. The classifica-
tion of the Trichogrammatidae ( Hymenop-
tera :ChaIcidoidea). Proc. Califf)rnia Acad.
Sci. 35:477-586.

DuPoRTE, E. M. 1965. The lateral and ventral
sclerites of the insect thorax. Canadian J.
Zool. 43:141-154.

EisEN, G. 1896. Biological studies on figs, caprifigs,
and caprification. Proc. California Acad. Sci.
5:897-1003.

Ferris, G. FMO. The myth of the thoracic sternites
of insects. Microentomology 5:87-90.

Froggatt, W. W. 1900. Insects living in figs, with
some account of caprification. Agri. Gaz.
New South Wales 11:447-456.

Galil, J., AND D. EisiKowiTCH. 1968. On the polli-
nation ecology of Ficus svcomorns in East Af-
rica. Ecology 49:259-269'.

GiRAULT, A. A. 1913. Australian Hymenoptera
Chalcidoidea IV-Vl. Mem. Queensland Mus.
2:140-334.

. 1915. Australian Hymenoptera Chalcidoidea

XII. Mem. Queensland Mus. 4:275-309.

. 1917. Some new Australian chalcid-flies,

mostly of the family Encyrtidae. Insec. Inscit.
Menstr. 5:29-37.

Graham, M. W. R. de V. 1969. The Pteromalidae
of northwestern Europe (Hymcnoptera:Chal-
cidoidea). Bull. Brit. Mus. (Nat. Hist.) Ent.
Suppl. 16:1-908.

Grandi, G. 1921. Richerche sul gen. Philotiy pests

Forst. (Hymenoptera :Chalcididae). Boll. Lab.

Zool. Portici 15:33-190.
. 1928. Hymenopteres sycophiles recoltes

dans rinde par le Frere E. Gombert. Bull.

Soc. Zool. France 53:69-82.

. 1929. Studio morfologico e biologico della

Blastopliugd psencs (L.). Boll. Lab. Ent. Bo-
logna 2:1-147.

. 1930. Monografia del gen. Philotiypesis

F5rst. Boll. Lab. Ent. Bologna 3:1-181.

. 1961. The hymenopterous insects of the

superfamily Chalcidoidea developing within
the receptacles of figs. Their life history,
symbioses, and morphological adaptations.
Boll. Ent. Univ. Bologna 26:1-13.

Hanna, a. D. 1935. The morphology and anatomy
of Ettchalcidia caryobori Hanna (Hymenop-
tera-Chalcidinae). Bull. Soc. Royale Ent.
d'Egypte 19:326-364.

Heymons, R. 1899. Beitrage zur Morphologie und
Entwicklungsgeschichte der Rhynchoten. Nova
Acta. Leop. -Carol. Deut. Akad. 74:349-456.

Hill, D. 1967a. Figs (Fiais spp.) and fig-wasps
(Chalcidoidea). J. Nat. Hist. 1:413-434.

. 1967b. The Figs (Ficiis spp.) of Hong

Kong. Hong Kong Univ. Press. 154 pp.

. 1969. Revision of the genus Liporrhopahtm

Waterston, 1920, (Hymenoptera, Chalcidoi-
dea, Agaonidae). Zool. Verhandl. 110:1-36.

Howard, L. O. 1887. A brief consideration of cer-
tain points in the morphology of the family
Chalcididae. Proc. Ent. Soc. Washington 1 :
65-75.

James, H. C. 1926. The anatomy of a British phyto-
phagus chalcidoid of the genus HannoUta
(Isosoma). Proc. Zool. Soc. London, Pt. 1:
75-182.

Joseph, K. J. 1953. Contribution to our knowledge
of fig insects (Chalcidoidea :Parasitic Hy-
menoptera) from India. III. Description of
three new genera and five new species of
Sycophagini, with notes on biology, distribu-
tion and evolution. Agra Univ. J. Res. 2:53-
82.

. 1*^56. Contributions to our knowledge of

fig insects (Chalcidoidea :Parasitic Hymenop-
tera) from India. VII. Descriptions of fifteen
new and revision of some old species. Ann.
Soc. Ent. France 125:97-133.

. 1964. A proposed revision of the classifica-
tion of the fig insects of the families Agaonidae
and Torymidae (Hymenoptera). Proc. Royal
Ent. Soc. London (B) 33:63-66.

. 1966. Taxonomy, biology, and adaptations

in the fig insects (Chalcidoidea). Proc. II
All-Indian Congr. Zool., Varanasi (1962)
2:400-403.

. 1967. Newer trends in the taxonomy of the

fig-inhabiting Chalcidoidea. Symp. Newer
Trends Tax. 34:257-262.

Kuttanl\ihiathu, J. J. 1955. The biology of
Pliilotyypesis caricae (L.), parasite of Blasto-
pliaga psenes (L.) ( Chalcidoidea :Parasitic
Hymenoptera). XV Int. Congr. Zool. Sec.
VIII: 1-2.

Malyshev, S. I. 1966. Genesis of the Hymenoptera
and the Phases of Their Evolution. Methuen
and ("o. Ltd., London. 319 pp.

Matsuda, R. 1960. A new interpretation of the
pleurosternal region of the hymenopterous
thorax. Acta Hymenopterologica 1:109-113.

. 1965. Morphology and Evolution of the In-
sect Head. Mem. Amer. Ent. Inst., Ann Arbor,
Michigan. 334 pp.

. 1970. Morphology and evolution of the in-
sect thorax. Mem. Ent. Soc. Canada 76:1-431.

Mayr, G. 1885. Feigeninsekten [Hymenoptera].
Verb. Zool. Bot. Gesell. Wien 35:147-250.

. 1906. Neue Feigen-Insekten. Wien Ent.

Zeitges. Wcin 25:153-187.

Michener, C. D. 1944a. A comparative study of
the appendages of the eighth and ninth ab-
dominal segments of insects. Ann. Ent. Soc.
.America 37:336-351.

Morphology and Systematics of Nlotropical Parasitic Fig Wasp

441

. 19-l-tli. Comparative external morphology,

ph\Iogeny, and a classification of the bees
(Hvmcnoptcra). Hull. Aiiier. Mus. Nat. Hist.
82:'l51-326.

Oi.DROYD, H., and C. R. Ribisands. I''?6. On the
validity of trichiation as a systematic character
in Trichogiamma (Hymcnoptera, Chalcitli-
dae). Proc. Roval Ent. Soc. Lontlon (R)
5:148-152.

Pratt, H. D. 19-10. Studies on the Ichncumonidae
of New England, part 1. The external mor-
phology of Arotes amocntis Cresson. J. New
York Ent. Soc. 48:155-193.

Ramirez, W. 1969. Fig wasps: mechanism of pollen
transfer. Science 163:580-581.

. 1970a. Host specificity of fig wasps

(.•\gaonidae). Evolution 24:680-691.

. l''7l)b. Taxonomic and biological studies of

neotropical fig wasps (Hymenoptera:Agaoni-
dae). Univ. Kansas Sci. Bull. 49:1-44.

Reid, J. A. 1941. The thorax of the wingless and
short-winged Hymenoptera. Trans. Royal
Ent. Soc. London 91:367-446.

Richards, O. W. 1956. An interpretation of the
ventral region of the hymenopterous thorax.
Proc. Royal Ent. Soc. London (A) 31:99-104.

Risbec, J. 1951. Les Chalcidoides d'A. O. F. Mem.
Inst. Francais d'Afrique Noire 13:1-409.

Smith, E. L. 1969. Evolutionary morphology of ex-
ternal insect genitalia. 1. Origin antl rela-
tionships to other appendages. Ann. Ent.
Soc. America 62:1051-1079.

SxoDGRAss, R. E. 1910. The thorax of the Hy-
menoptera. Proc. U.S. Nat. Mus. 39:37-91.

. 1935. Principles of Insect Morpholog\'. Mc-
Graw-Hill Rook Co., New York. 667 pp.

. 1941. The male genitalia of Hymenoptera.

Smithsonian Misc. Publ. 99:1-86.

. 1951. A revised interpretation of the ex-
ternal reproductive organs of male insects.
Smithsonian Misc. Publ. 135:1-60.

. 1962. Suture or sulcus? Proc. Ent. Soc.

Washington 64:222-223.

Standley, p. C. 1917. The Mexican and Central
American species of Ficiis. Contr. U.S. Nat.
Herb. 20:1-35.

TuLLocH, G. S. 1929. The proper use of the terms
parapsides and parapsidal furrows. Psvche
36:376-382.

Walker, F. 1843. Descriptions of Chalcidites dis-
covered in St. Vincent's Isle by the Rev. Lans-
down (juilding. Ann. Mag. Nat. Hist. 12:46-
49.

. 1871. Notes on Chalcidiae. Pt. 4, 55-70.

E. W. Janson, London.

. 1875. Descriptions of new genera and spe-
cies of parasites, belonging to the families
Proctotrupidae and Chalcididae, which attack
insects destructive to the fig in India. Ento-
mologist 8:15-18.

Weber, H. 1924. Das Grundschema des Pterygoten-
thorax. Zool. Anz. 60:17-37. 57-83.

. 1928. Die Gliederung der Sternopleural-

region des Lepidopterenthorax. Eine ver-
gleichende morphologische Studie zur Sub-
coxaltheoric. Z. Wiss. Zool. 131:181-254.

Wiebes, J. T. 1964. Host specificity of fig wasps

(Hymenoptera Chalcidoidca, Agaonidae). Proc.

XII Int. Congr. Ent. London: 95.
. 1966a. Rornean fig wasps from Fictts sttt-

penda Miquel (Hymenoptera, Chalcidoidea).

Tiidsch. Ent. 109:163-192.

. 1966b. The structure of the ovipositing

organs as a tribal character in the Inclo-Aus-
tralian svcophagine Torymidae (Hvmenop-
tera, Chalcidoidea). Zool. Meded. 41:151-159.

. 1967. Redescription of Sycophaginae from

Ceylon and India, with designation of lecto-
types, and a world catalogue of the Otitesel-
lini (Hymenoptera Chalcidoidea, Torymidae).
Tiidsch. Ent. 110:339-442.

. 1968. Fig wasps from Israeli Fictis syco-

monis and related East African species (Hy-
menoptera, Chalcidoidea). 2. Agaonidae
(concludetl) and Svcophagini. Zool. Meded.
43:307-321.

. 1970. Revision of the Agaonidae described

by J. Risbec, and notes on their torymid
symbionts (Hymenoptera, Chalcidoidea).
Zool. Meded. 45:1-16.

Williams, F. X. 1928. Studies in tropical wasps —
their hosts and associates (with descriptions
of new species). Rull. Hawaiian Sugar Exp.
Sta., Ent. Scr. 19:1-179.

WoLcoTT, G. N. 1951. The insects of Puerto Rico.
J. Agr. Univ. Puerto Rico 32:749-882.

442

The University of Kansas Science Bulletin"

0.4 mm

Figs. 25-28. Left antenna of Idarncs females, inner aspect. 25, /. barbigcra; 26, /. bitcatoma; 27, /. micheneri;

28, 7. ashlocJii.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

443

0.4 mm

Figs. 29-33. Left antenna of Iduines females, inner aspect. 29, 7. flaricolhs; 30, /. obtiisijoliac; 31, I. galbina;

32, /. carme; 33, /. cami/n.

444

The University of Kansas Science Bulletin

0.4 inni

Figs. 34-40. Left antenna of Idanies females (34-36) and males (37-39), inner aspect; right antenna, male,
dorsal aspect (40). 34, /. s/miis; 35, /. oscrocata; 36, /. liniciiezt: 37, /. galbina: 38, /. ciunini; 39, /. aililocl{i;

40, /. canne.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

445

.t.^'C/^r

i<

^^

0.25 mni

Figs. 41-47. Antennae of Idarnes males, dorsal aspect; 45, 47 right antenna, remainder left. 41, 7. jimenezi;
42, /. oscrocata; 43, /. bticutoma; 44, /. barbigera; 45, /. obtiisijoliae; 46, /. micheneri; 47, /. flaiicollis.

446

The University of Kansas Science Bulletin

48

49

52

53

^

""""T^n-&rrr^jv.\tx^'

]^..,^-

<^

0.6 mm

Figs. 48-53. Right forewing of Idaines females. 48, /. flaricollis; 49, /. siiniis; 50, /. galbina; 51, /. ashloclii,

52, /. ohtitsijoliae; 53, /. aume.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

447

54

55

.««''

r^r^""

'rrrniTT""'"''^

56

"Tirnnrmri^^T^'

^

<o

<<^

:;^

57

TniTTTTrm^

58

59

'r^.

""TTmtTTTTmT'.T'^

().() mm

Figs. 54-59. Right forewing of Idanies females. 54, /. bv.catoma; 55, /. Lainini; 56, /. bavhigera; 57, /. oscro-

cata; 58, /. michenen; 59, /. jimenezi.

448

The University of Kansas Science Bulletin

64

0.25 mm

Figs. 60-68. Right mandible of Iduines females, inner aspect. 60, /. imchencn: 61, /. barhigera; 62, /. obttisi-
joliae; 63, /. ashlockr, 64, /. hiuutoma; 65, /. fwienezi; 66, /. oscrocata: 67, /. camini; 68, /. galbma.

Morphology and Svstematics of Neotropical Parasitic Fk; Wasp

449

72

>.25 mm

Figs. 69-73. Right mandible ot Iduines females (69, 72) and males (70, 71, 73), inner aspect. 69, /. fluricollis

70, /. ohtiisijoluic : 71, /. aihlockj; 72, /. carmc; 73, /. caimc.

450

Thk University of Kansas Sciknce Bulletin

0.25 m m

Figs. 74-82. Right mandible of Idanies males, inner aspect. 74, /. bitcatoma; 75, /. jiavicolUs; 76, /. michencn:
77, /. galhina; 78, /. jiniciiezi; 79, /. cainini; 80, /. barbigera; 81, /. sunns; 82, /. oscrocata.

Morphology and Systematics oi-' Neotropical Parasitic Fig Wasp

451

V

s

87 A

o

0.1 mm

Figs. 83-91. Fore, middle and hind tibiotarsi of Idanics males, inner aspect. 83-85, /. ashlocki; 86-88, /. galhina;

89-91, /. simns.

452

I'm; University of Kansas Science Bulletin

^

1-

^/'t

96

^;

\^':

97

^

98

)

\~y7K

0.1 nun

Figs. 92-100. Fore, middle and liind tibiotarsi of Idanies males, inner aspect. 92-94, /. flavicolUs: 95-97, /.

carme; 98-100, /. camini.

Morphology and Svstkmatics of Neotropical Parasitic Fig Wasp

453

102

0.1 mm

Fk;s. 101-106. Fore, middk- antl liin.l tilnntarsi of Idaifies males, inner aspect. 101-103,7. ohtiisijoUae; 104-106-

/. michencri.

454

The University of Kansas Science Bulletin

\_ 110

111

m

Jk^ ■'

I \_

/ A

O.l

115

Figs. 107-115. Fore, middle and hind tibiotarsi of Idarnes males, inner aspect. 107-109,/. barbigera; 110-112,

/. jimcnezi: 113-115, /. oscrocata.

Morphology and Systematics of Neotropical Parasitic Fig Wasp

455

117

116

■X

I

1^

mh

S'.

Figs. 116-118. Fore, middle and hind tibiotarsi of Idarnts hncatoina male, innei' aspect. (Hind tibiotarsus

0.2 mm.)

s

s

g

I

§

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

iMUs. ^

w

L;Br?ARy

APR 1 5 /S7o

HARVARD

^■'^ivr c- TV

i

5

THORACIC POLYMORPHISM IN

MESOVELIA MULSANTI
(HEMIPTERA: MESOVELIIDAE)

«

;s

By
JANET E. GALBREATH

&

g

S

>;:

Vol. 50, No. 10, pp. 457-482

Aprils, 1975

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out at
the University of Kansas or by University faculty and students. Since its incep
tion, volumes of the Bulletin have been variously issued as single bound volumes,
as two or three multi-paper parts or as series of individual papers. Issuance is at
irregular intervals, with each volume prior to volume 50 approximately 1000
pages in length. Effective with volume 50, page size has been enlarged, reducing
the length of each volume to about 750 pages.

The supply of all volumes of the Kansas University Quarterly is now ex-
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased at
$20,00 per volume. Where some of these volumes were issued in parts, individual
parts are priced at the rate of 2 cents per page. Current policy, initiated with
volume 46, is to issue individual papers as published. Such separata may be
purchased individually at the rate of 3 cents per page, with a minimum charge
of $1,00 per separate. Subscriptions for forthcoming volumes may be entered at
the rate of $20,00 per volume. All communications regarding exchanges, sales
and subscriptions should be addressed to the Exchange Librarian, University
OF Kansas Libraries, Lawrence, Kansas 66045.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

The International Standard Serial Number of this publication is US ISSN
0022-8850.

Editor
Charles R. Wyttenbach

Editorial Board

William L. Bloom
Philip W. Hedrick

Rudolf Jander

Harvey Lillywhite

Charles D. Michener

Norman A. Slade

Henry D. Stone

George W. Byers, Chairman

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. 50, No. 10. pp. 457-482 April 8, 1975

Thoracic Polvinorphism in Mesovelia rnulsanti
(Hemiptera: Mesoveliidae)

Janet E. Galbreath

Thoracic Polymorphism in Mcsovelia muLsanfi
(Hemiptera: Mesoveliidae)^

Janet E. Galbreath-

ABSTRACT

Three thoracic murphs are recognized: winged, wingless and intermediate (new term
to identity flightless adults that develop from nymphs with wing pads). The wingless morph
may be recognized in the iourth nyinphal instar by the absence of wing pads. An individual
of the fourth instar with wing pads must be observed in the fifth instar to identify the morph.
In an intermediate morph of the fifth instar, the wing pads may reach to the second abdominal
segment, remain short, become vestigial, or be absent. A winged morph of the fifth instar has
long wing pads which extend to the third abdominal segment or beyond.

Results of crossing studies gave little evidence to suggest that anv one of the nine kinds
of crosses (three morphs, sexes taken reciprocally) would yield significantly more of any one
kind of morph. The occurrence of morphs (as nymphs) was related to field temperatures,
called "the estimated mean air temperature" and derived from the climatologist's "march of
monthly mean temperatures." Fourth and fifth instars of the wingless rnorph were collected
when the estimated mean air temperature ranged from a spring low of 16.7° C to a summer
high of 28.3° C and to a fall low of 8.3° C. Fourth and fifth instars of the winged and in-
termediate morphs were found primarily, but irregularly, when estimates of the mean air
temperature were 21.S° C or higher.

Nymphs have one or two morph-determining points in development. The first, whether
or not wing pads develop, occurs in all nymphs. The sentitive stage for the first critical point
is unknown, but it is thought to occur prior to the late third instar. In the field, high temper-
atures and unknown factors during the first critical period are apparently necessary for a high
frequency (11-45°.) of nymphs with wing pads. In the laboratory, a high frequency of
nymphs with wing pads was reared from a parental stock with a frequency of 11 to 45% and
when the fluctuating temperatures averaged 25.6° C or higher. The second critical period
occurs early to midway in the fourth stadium of nymphs with wing pads. A fourth instar
with wing pads is more likely to become a winged morph in the field than in a laboratory
rearing, but it is not known why.

INTRODUCTION

third term, short-winged or brachypterous,
has been used in two ways: to name the

Historical background. Thoracic poly- wingless adults mentioned above (Jaczew-

morphism is typical of members of the ski, 1928; Gupta, 1963) and to denote

superfamily Gerroidea; Usinger (1956) winged adults with broken wings, an

gave examples of it in six families. In error first recognized by Horvath (1915).
Mesovelia. thoracic polymorphism occurs The literature concerning Mesovelia

in both sexes. Heretofore, two morphs shows that the winged morph is rare in

have been recognized: the common wing- areas where the average monthly mean

less, or apterous, morph lacking wing pads for July is 21° C or lower. The relative

as a nymph and wings as an adult, and abundance of the winged and wingless

the winged, or macropterous, morph morphs of M. j areata Mulsant and Rey

known to have long wing pads in the was observed between 59° and 45° north

fifth instar and wings in the adult. A latitude in Europe where the average

77 Z ■ , ,--, r , r. monthly means of Julv are 17° C to about

Contribution number 13/ 4 rrom the Department ^ ^ . '

of Entomology, University of Kansas, Lawrence, 21 C, respectively. In Sweden, where

Kansas 66045. this species is univoltine, wingless adults

"608 SkvHne Drive, Carbondale, Illinois 62901 or , • j j i. ■

^/ p, ' . rv , c 1 Til ■ Ti 'ire common; onlv one wmged adult is

c/o Department or Zoology, Southern Ilhnois Univer- ' ^

sity, Carbondale, Illinois 62901. recorded (Ekblom, 1930; Wesenberg-

459

460

The University of Kansas Science Bulletin

Lund, 1943). Leth (1943) reported 20
(= 7%) winged adults from the southern
half of Denmark. In Germany and
France, where there are one or two an-
nual generations, the winged morph is
infrequent. Jordaii (1931, 1936), who
studied in Oberlausitz, found only wing-
less adults in 1925 to 1927, five winged
among many hundred wingless adults in
1928, and three specimens of the winged
morph (one adult, two nymphs) in 1934.
Apparently no more winged adults were
collected until 1950, when he was aston-
ished to find 16 (Jordan, 1951). Schu-
macher (1919) found one winged adult
near Brodown, Germany; Miiller (1919)
reported winged adults and the more
common wingless adults on July 5, 1911,
on a pond near Walkenried, Thtiringia.
Poisson (1922, 1933) stated that he had
never collected the winged morph in the
Department of Calvados.

Likewise near Pellston, Michigan (45°
34' north latitude; average monthly mean
of July, 18.8° C), Hungerford (1953) re-
ported few winged (none, 1923-1950; a
total of five, 1951-1952) and a preponder-
ance of wingless adults of M. doiiglasen-
sis Hungerford (taxonomy not clear; pos-
sibly a synonym of M. amoena Uhler).
In contrast, he gave data frcjm pinned
specimens from southern states (37'"' to
25° north latitude; average monthly means
of July, 25.5° C or higher) showing a fre-
quency of 20^ c. (N = 174) for the winged
morph.

The literature pertaining to A/, iiuil-
santi White near Pellston gives some
information on the morphs of adults.
Hussey (1919) listed only the wingless
morph taken in general collecting in the
years 1913, 1914, and 1918. Hoffmann
(1932) for the year 1930 and Neering
(1954) for the year 1950 reported that
winged adults occurred with the more
common wingless morph. Pinned adults
from this area in the Snow Entomological

Museum, University of Kansas, include
no winged (N = 63) in seven years (1923-
1925, 1927, 1930-1931, 1937) and 54 winged
(= IL'o) in three years (1939, 1941, 1950).

In laboratory crossing experiments,
only the wingless morph of the European
M. f areata has been reared (Poisson, 1922;
Jordan, 1931). Poisson used wingless par-
ents and Jordan used winged and wingless
parents. Hoffmann (1932) worked with
wingless parents and obtained wingless
progeny in rearings of the North Ameri-
can M. cryptophila Hungerford and M.
doiiglasensis. One record of three winged
adults reared from eggs of M. japonica
Miyamoto was reported by Miyamoto
(1964); the morphs of the parents were
not known.

Questions to be answered. This paper
describes the developmental pathways of
morphs and discusses the extent of en-
vironmental and genetic causation of
morphs in M. midsanti. To understand
the developmental divergence of the
thorax, observations were made on the
form of the first three instars, of the
fourth instar with wing pads and of the
fifth instar and adult of the heretofore
unrecognized intermediate morph. Data
were accumulated on the percentages of
morphs among pinned adults, nymphs
and adults in field samples in three geo-
graphical areas (Fig. 1), and nymphs in
laboratory rearings. Study of these data
suggested three tests. Field data on the
occurrence of the winged morph as
nymphs and as adults showed the need
to observe whether or not migration could
be influencing the frec]uencies of winged
adults. Field data on the occurrence of
morphs as nymphs indicated that a test
could be made on whether or not the
occurrence of nymphs with wing pads
could be related to high temperatures.
Data from laboratory crosses were tested
to determine if any one kind of cross

Thoracic Polymorphism in Mesovelia mttlsanti

461

would yield significantly more of any one
kind of morph.

In the field, other problems included
the sequence of the life history stages,
where these stages lived, variations of
temperatures and how temperature in-
fluenced the rate of nymphal develop-
ment. In the laboratory, records were
made on the duration of the nymphal

stage.

ACKNOWLEDGMENTS

This work, which initially was directed
liy the late H. B. Hungerford, was com-
pleted under the supervision of C. D.
Michener. I am grateful to my husband,
E. C. Galbreath, for his dedicated interest
in this study and assistance with the prepara-
tion of this manuscript. The opportunity to

- 37'44' N

I J A 5 O N D

Fig. 1. Map of the central United States showing

where populations of A^. mtdsanti were studied, and

graph of the average monthly mean air temperatures

for these three areas.

carry on research at the University of Michi-
gan Biological Station was made possible
by a Summer Fellowship from the National
Science Foundation and by an Ida H. Hyde
Scholarship from the University of Kansas.
I am grateful for use of facilities at the Uni-
versity of Kansas and at Southern Illinois
University at Carbondale.

MATERIALS AND METHODS

Field worli. Data were accumulated
on the percentages of morphs among
adults (15,000) and nymphs (9,765 fourth
and fifth instars) in field samples. A dip
net and an aspirator were used for col-
lecting. The insects survived collecting
best in an aspirator with a piece of mois-
tened paper toweling; otherwise they often
died by desiccation if left too long in dry
containers and by drowning if caught in
a condensation of moisture in wet con-
tainers.

In making a field count, a stereoscopic
microscope with a magnification of 45 X
was used to identify the morphs of
nymphs. Fourth instars were recognized
as having wing pads (Fig. 2A) or as the
wingless morph (Fig. 2B). Fifth instars
were recognized as the winged morph
(Fig. 2C), as intermediates (Fig. 2D, E,
F) or as without wing pads (Fig. 2G, H),
the latter being grouped as the wingless
morph. Adults were recognized as winged
(Fig. 21) or flightless, because some indi-
viduals of the intermediate morph (Fig.
2L) are not distinguishable from the wing-
less morph (Fig. 2M). When the count
was finished, nymphs and adults not used
in other studies were returned to the lo-
cality where they had been found.

During a field count, fourth instars
with wing pads and fifth instars of the
intermediate morph (Fig. 2D, E, F) were
placed in separate containers for rearing
in the laboratory. A fourth instar with
wing pads was reared until the fifth in-
star to recognize the winged or the inter-
mediate morph. When an individual did

462

The University of Kansas Science Bulletin

not survive, a dissection was made to
learn if the exoskeleton of the iifth instar
was sufficiently developed for identifica-
tion. Often such an individual was early
in the instar and was listed as "morph
unknown" in the tables containing field
data. Fifth instars of the intermediate
morph were reared to observe the thoracic
form of the adult. Unless used for rearing
studies, each specimen with its exuviae
was preserved in a separate vial of alcohol.
Routine field counts were made once
a week and frequently more often. In the
vicinity of the University of Michigan
Biological Station near Pellston, Michigan,
between June 18 and August 10, 1955, 29
counts were made at 15 places (Cheboygan
County: two open beach pools at Black
Lake; Black River; Bryant's Bog; vicinity
of Douglas Lake — marsh near mouth of
Bessey Creek, mouth of Bessey Creek,
open beach pool at Hook Point, Maple
River Bay, two closed beach pools at Sedge
Point; two drainage ditches near Duncan
Bay; mouth of Indian River; Emmet
County: Maple River Swamp; Pellston
Pool). Most collections and observations
were made in the habitats around Douglas
Lake because they were most accessible.
Near Lawrence, Kansas, field samples
were made regularly at The Pond, Uni-
versity of Kansas Natural History Reser-
vation, from September 30 to October 19,
1955; July 12 to November 18, 1956; and
May 1 to September 6, 1957; with a few
collections at Eudora Pond in 1956 and
1957 and one sample at Bagsby Pond in
1956. All of these places are in Douglas
County. Near Carbondale, Illinois, rou-
tine field collections were made at Fisher
Pond during periods of seasonal activity
from April 28, 1959 through September
6, 1962. More frequent collections were
made in 1961 : 22 samples of nymphs from
May 18 to June 18 and 23 samples of adults
from May 26 to June 18. Each day all
unmarked adults were marked on the

thorax with model paint applied with a
crow quill pen. In addition, some samples
were made at LaRue Swamp in 1959
through 1961, at Lewis Swamp in 1959,
and at Lake-on-the-Campus and at Tar
Hill Pond in 1961. LaRue Swamp is in
Union County; the other collecting places
are in Jackson County.

Records were kept on the duration of
seasonal activity, i.e., when first instars
hatched in the spring and when nymphs
and adults were last observed in the fall.

The temperature data came from the
following sources: for Europe, from the
United States Department of Agriculture
(1941); and for the areas of study, from
"Climatological Data: Michigan" (1955),
"Climatological Data: Kansas" (1955-
1957), "Climatological Data: Illinois"
(1910-1962), "Local climatological data.
Carbondale, 1910-1962," the operating rec-
ords of the Murphysboro Water Plant,
Murphysboro, Illinois, and from a thermo-
graph placed at Fisher Pond, Carbondale,
during the seasonal activity of M. mulsanti
in 1961-1962. Freeze probabilities were
determined from the dates of threshold
freezes of 0° C or lower for Carbondale
for the years 1910-1962. Freeze probabil-
ities are statistics used by climatologists
(Thom and Shaw, 1958) and are derived
from threshold freezes, the last vernal and
the first autumnal freeze.

To relate seasonal temperatures to the
dates of collection, a graph of the monthly
mean air temperatures was made for each
year of research from the records of the
U.S. Weather Bureau Station closest to
the area of study. An estimated mean air
temperature for any specific date was ob-
tained by reading the temperature indi-
cated by the intersection of a line drawn
through the date and the line connecting
the monthly mean temperatures. For ex-
ample, in Figure 4, line C (for September
1) intersects the line of the mean air tem-
perature at 2i.i° C. Estimates of the mean

TiioRAcic Polymorphism in Mesovelid mulsanti

463

temperature of pond water were deter-
mined for 1961-1962 from graphs of the
monthly means from data obtained at
Fisher Pond.

Laboratory ivorl(. Data were accumu-
lated on the percentages of morphs among
pinned adults (3,000 specimens. Snow
Entomological Museum, University of
Kansas) and among nymphs (2,355)
reared primarily from virgin crosses in
the laboratory.

The laboratory rearings rec]uired han-
dling over 8,000 individuals within one
month's time. To rear M. mulsanti.
freshly killed insects for food, a rearing
container with a cover, a moist environ-
ment, and a matrix for the insertion of
eggs must be provided. Care was needed
in handling of insects used for food (pri-
marily Drosophila) to be sure that the
ether fumes had evaporated and the bodies
remained moist for piercing and sucking.
First, second, and third instars required
one fruit fly per day; fourth and fifth
instars and adults needed two per day.
Circular, plastic, covered containers, 35
mm in diameter and 28 mm tall, were
used. A humid environment was main-
tained in two ways. The eggs and the
first through third instars were kept in
containers partially filled with distilled
water. For fourth and fifth instars and
for adults, two layers of paper toweling
moistened with distilled water were placed
on the bottom of the rearing containers.
The toweling also served as a matrix into
which the ovipositor was inserted and to
which the eggs were usually attached.
The toweling was not likely to in-
fluence development. Abnormal develop-
ment in some pyrrhocorids was recognized
to be caused by a "paper factor," now
known to be juvabione (Gilbert and
King, 1973). Carlisle and Ellis (1967)
observed that toweling did not modify
growth when the manufacturer had used
a chemical process or if the wood pulp

dill not include the balsam fir. The
toweling used in my work was made by a
chemical process from wood pulp with no
balsam fir (personal communication from
the manufacturer).

Some progeny were reared from eggs
laid by females collected as adults in the
field (Table 5: Pcllston, expt. 1-4; Law-
rence, expt. 1-2; Carbondale, expt. 4); in
these studies the morph of just the one
parent was known. In 1955, virgin crosses
were set up from parental stocks in which
the percentages of nymphs with wing pads
were unknown. Beginning in 1957, cross-
ing experiments were set up with parental
stocks taken when field counts indicated
0 to 10^ c or alternatively 11 to 45^ □
nymphs with wing pads. Differing con-
ditions of temperature and light were used
(Table 5). Each experiment started with
collecting fourth and fifth instars from
the field and rearing each nymph in isola-
tion in the laboratory. When the adults
emerged, single-pair crosses were estab-
lished. The matrix of toweUng with eggs
from one female for one day of oviposi-
tion was placed in a separate rearing con-
tainer. In about five days (average tem-
peratures 25° C or higher) the eggs were
examined for embryos; if there were few
embryos, the eggs and parents of the par-
ticular cross were discarded. When the
fertility rate was high, the toweling was
separated to place five or six eggs in a
container — a precaution to reduce future
cannibalism. For each group of eggs
records were kept of the number and
kinds of morphs produced.

During the rearings, observations were
made on the thoracic form of the first
three instars. Dead third instars were
dissected to observe in late instars the
presence or absence of wing pads. Fourth
instars of the wingless morph were pre-
served. Fourth instars with wing pads
were isolated and reared as long as possi-
ble; those which died were dissected to

464

Stage of
life hislfiry

The University of Kansas Science Bulletin
Winged Morph InlemediaU Morph

Wingless Morph

Adult

Fifth
mslar

Fourth
instar

u'lth u'lng pads

u'lthoul u'ing pads

Fig. 2. Diagrams to show developmental (.livergence in M. tnnlsanti. A, Fourth n\inphal instar with wing pads,
inorph unknown. Winged morph. C, Fifth instar; I, young adult female. Intermediate morph. D-G, Thor-
axes of fifth instars: D, with short wing pads; E, with shorter wing pads; F, with vestiges of wing pads;
G, without wing pads. J-L, Thoraxes and heads of adults: J, with three vesdges of wing pads and reduced
ocelli; K, with two vestiges of wing pads and no ocelli; L, without vestiges of wing pads and ocelli. The
individual illustrated as a fifth instar, D, and as an adult, J, represents the maximum development toward
wingedness observed in the intermediate morph. Wingless morph. B, Fourth instar; H, Fifth instar (thorax as
in G); M, adult female (details as in L). (Parts B, C, I and M are from Hungcrford, 1917.)

observe in late instars the winged or the
intermediate morph. Some nymphs with
wing pads, the dissected late third and
early fourth instars, were grouped as
"morph unknown" in the tables of progeny
counts. Each individual of the winged and
intermediae morphs was preserved as a
fifth instar or as an adult with its exuviae
in a separate vial.

Measurements were made of the wing
pads of exuviae of the fourth instar of
each preserved specimen known to be a

winged or an intermediate morph. The
thorax of the exuviae of the fifth instar
and of the adult of the same intermediate
morph individual was studied from pre-
served specimens.

The duration of the nymphal stage was
observed at 23.9° and 30.0° C.

Near the end of the summer, the dura-
tion of the incubation period was used to
determine whether the eggs were non-
diapause or diapause (Galbreath, 1973).

Thoracic Polymorphism in Mesovelta mulsanti

465

MORPHS

Developniental divergence. Three tho-
racic morphs are recognized: winged, in-
termediate and wingless (Fig. 2). Indi-
viduals do not seem to deviate from the
norm of five nymphal instars. The meso-
tcrgum and the metatergum remain un-
dilTerentiated through the third instar of
all nymphs and through the adult of the
wingless morph (Fig. 2B, H, M). Wing
pads are present in some fourth instars
(Fig. 2A) and may he observed in dissec-
tions of some late third instars, htit the
morph of stich individuals cannot be rec-
ognized. Measurements of the length and
width of the wing pads on exuviae of
lourth instars of preserved specimens
known to be either a winyed or an inter-
mediate morph did not show a morph-
related difference. However, differences
in the development of the wing pads dur-
ing the fourth stadium are evident after
the molt into the fifth instar: the growth
is normal (winged morph, Fig. 2C), some
growth occurs (intermediate, Fig. 2D),
growth is suppressed (intermediate, Fig.
2E), or reduction occurs (intermediates,
Fig. 2F, G). These thoracic differences
may also be observed in dissections of late
fourth instars with wing pads. Adults
have wings (winged morph, Fig. 21),
vestiges of wing pads (intermediates, Fig.
2], K), or no sign of wings (intermediate.
Fig. 2L; wingless morph. Fig. 2M). Ves-
tiges of wing pads (Fig. 2F, J, K) retain
the mesal position characteristic of wing
pads and in adults lack the differentiation
of wings.

Each morph has a stage or stages when
it may be recognized.

The winged morph may be identified
as a fifth instar (Fig. 2C) by the long
wing pads which reach to the third ab-
dominal segment or farther, and as a
young adult (Fig. 21) by the well-de-
veloped wings themselves, as well as by
two prominent ocelli, the pronotal lobe,

and the exposed mesonotal and metanotal
scutella. In older individuals, the apices of
the hemelytra and wings are broken and
the indirect flight muscles undergo autoly-
sis.

The one feature common to all inter-
mediates is the presence of wing pads in
the fourth instar. The rarest intermediate
lacks vestiges of wing pads in the fifth
instar and may not be distinguished from
the fifth instar of the wingless morph
(Fig. 2G, H, respectively). Fifth instars
with short wing pads (Fig. 2D, E) and,
more fret]uently, with vestiges of wing
pads (Fig. 2F) are recognizable as inter-
mediates. A few of the adults may be
identified as intermediate by having ocelli
or vestiges of wing pads; some have both
oi these features (Fig. 2J). However, the
majority of intermediate adults (Fig. 2L)
are inseparable from wingless adults (Fig.
2M).

To recognize an individual of the
wingless morph, one must know that
wing pads are absent in the fourth instar
(Fig. 2B). This knowledge is important
in designing rearing experiments with
virgin crosses.

In flightless adults, the intermediate
and wingless morphs, the pronotal lobe is
absent; the mesonotum and the meta-
notum remain simple, subequal segments;
flight muscles are absent; and the phrag-
mata are reduced. The wings of some
intermediates are reduced to stubs called
vestiges of wing pads or to minute, dis-
torted areas visible only with a stereo-
scopic microscope at a magnification of
60X.

Discussion. "Polymorphism can usually
be interpreted as the outcome of develop-
mental divergence after a critical point
which may be anywhere from the tgg to
the adult stage. The fate of the indi-
vidual is thought to become fixed or deter-
mined at this critical point. However, as
in studies of growth and development of

466

The University of Kansas Science Bulletin

tissues and organs, words like 'determined'
often turn out to have relative rather than
absolute meanings . . . Determination does
not always occur at the same stage in all
individuals of a species" (Michener,
1961:47).

My research with M. miilsanti and
study of papers by Ekblom (1949), Miya-
moto (1953), Sprague (1956) and Brink-
hurst (1963) on developmental divergence
in members of three other families of
semiaquatic Hemiptera show that there
are several ways that the development of
the nymphs may differ: the nymphs may
or may not have wing pads, the wing pads
may not appear in the same instar, the
wing pads may show differences in rates
of growth or amounts of regression, the
number of kinds of flightless morphs may
vary, and the morphs may not be recog-
nizable at the same stage of development.
The wingless morph of M. mulsanti does
not have wing pads; all nymphs of the
species reported on in the literature cited
above have wing pads. In Hydrometra
7nartini Kirkaldy, wing pads are found
in the third instar; in Microvelia dtlitta
Distant, wing pads appear in the fourth
instar. Most of the differences in the de-
velopment of wing pads appear to be a
result of different rates of growth, al-
though regression or loss of wing pads
occurs in some intermediate nymphs of
M. mulsanti. The kinds of morphs may
vary from two {H . martini with one
winged and one flightless) to several (one
winged and several flightless). In some
species (M. midsanti, M. dihita, Gerris
lateralis Schummel = Gerris asper Fieber
of Ekblom) the thoraxes of flightless
morphs show much variation as nymphs
and adults. The stage when the morphs
first become recognizable ranges from the
third instar (H. martini) to the adult
{Gerris odontogaster [Zetterstedt]).

A critical point has been suggested for
only one species, G. odontogaster. The

hypothesized critical point of the egg dur-
ing vitellogenesis (Brinkhurst, 1959) was
not substantiated. Instead, the critical
point is thought to occur during the first
four stadia (Vepsaliiinen, 1971b).

Study of the developmental divergence
of M. midsanti suggests that all individuals i
could undergo a critical point sometime |
before the late third instar, when the path- !
way for the wingless morph may be recog-
nized in dissections. Then it is possible I
that fourth instars with wing pads un- I
dergo a second critical point which deter-
mines whether the individual becomes a
winged or an intermediate morph. See
later sections for answers to these ques-
tions. I

RESULTS OF FIELD STUDIES

Seasonal temperatures. To study the
effect of seasonal temperatures on M.
mulsanti, one must know where the indi- i
viduals live during their various stages.
It is now known that the tgg overwinters
(Galbreath, 1969). Diapause eggs em-
bedded in plant tissue overwinter below
the surface film and post-diapause nymphs
hatch below the film and swim upward
to attain their position on the surface.
Non-diapause eggs, inserted in aquatic
plant tissue just above the water line,
must remain in a moist environment.
Then, depending upon the position of
the plant, a nymph hatches from a non-
diapause egg and either climbs onto the
surface of the floating plant or emerges
below the surface and swims upward as
described above. Nymphs and adults live
on the surface film but in the air.

This study is concerned with the sub-
surface water temperatures, surface film
temperatures, and air temperatures influ-
encing nymphal development. However,
daily fluctuations in temperatures mask
the overall point of view that is needed to
relate the occurrence of morphs to differ-
ent temperatures. The climatologist's

Thoracic Polymorphism in Mesovelia miilsanti
1 1

T

T

T

467

-T

25

20

15

■D

o

O)

c

0)

U

0)

t 10

Q

0-.

o stream water
^ pond water

M

M

J

J

O

N

Fig. 3. Comparison of the monthly mean temperatures of air, stream water, and water of Fisher Pond for 1961
at Carbondale, Illinois. Data on air temperatures were obtained from the U.S. Weather Bureau Station at Car-
bondale. Temperatures for stream water were compiled from daily records of the Big Muddy River made at
the Murphysboro Water Plant. The mean for the month was the sum of the daily temperatures divided by
the number of days in the month. Water temperatures for Fisher Pond were obtained from a thermograph
for April 23 to July 28 and September 15 to November 21. The monthly mean temperatures were calculated
from the average of the maximal water temperature plus the average of the minimal water temperature divided

by two.

"march of monthly mean air tempera-
tures" smooths these variations.

The graphs and data of Harmeson
and Schnepper (1965) record a marked
similarity in the average monthly mean
temperatures of air and surface waters in
Illinois for 38 stations, 27 flowing streams
and 11 impoundments of flowing streams.
Analvsis of their data shows that the

average monthly means of surface waters
are generally (90% of the 456 sets of water
and air data compared) equal to or higher
than those of air. Figure 3 is a graph
plotted after Harmeson and Schnepper
but comparing the monthly mean tem-
peratures of air and stream water to those
of the water in Fisher Pond where field
studies were made. From April li

468

The University of Kansas Science Bulletin

30

Percent probability of threshold freezes 0°Cor lower
occurring after date in spring and before date in fall

100% 50% 0%

0% 50%

100%

Fig. 4. Graph relating seasonal activity of M. miihanti, probabilities of threshold freezes, and the average
monthly mean air temperatures at Carbondale, Illinois. First records of first instars (A) were as early as April
28 and as late as May 2. First records of adults (B) were as early as May 12 and as late as May 26. The last
record of nymphs or adults (D) occurred as early as October 22 and as late as November 14. The earliest
record of eggs in diapause (C) was obtained from eggs laid in the laboratory on September 1, 1971, by females

collected as adults on August 31 (unpublished data).

through July 28, the average temperatures
of pond water were warmer than those of
stream water and air. The biggest differ-
ence occurred in June (25.7° C for pond
water, 22.8° C for stream water, 22.0° C
for air). From September 15 until No-
vember 21, the average temperatures of
the pond water were cooler than those
of the stream water but warmer than those
of air.

Because water temperatures were avail-
able for but two out of seven years in-
volved in this study, seasonal temperatures

were determined from air data because of
the close similarity in the monthly mean
temperatures of air and water.

Seasonal activity at Carbondale. Sea-
sonal activity refers to the time when
nymphs and adults are found on the sur-
face film. Figure 4 shows the variations
in seasonal activity observed at Fisher
Pond from 1959 to 1962. Hatching of the
first instars occurred after the last vernal
freeze of each year, when the freeze proba-
bility was 10% or lower (Fig. 4A) and

Thoracic Polymorphism in Mesovelia mulsanti

469

Table 1. Field records of the first generation at Fisher Pond, Carbondale, Ilhnois. All
temperatures are °C. (Average monthly mean of May: 19.3°.)

Year

Monthly

First

Estimates

First

Estimates

Duration

mean temp.

records

c

)f mean

records

c

f mean

of

of May

of first
instars

temp

of

temp.

Surface

nymphal

Surface

adults

stage

Air

water

Air

water

in days

1959

21.5

Apr. 28

17.2

May 16

21.4

18

1960

17.6

Mav 2

16.7

..

May 22

20.8

..

20

1961

16.7

May 2

14.4

16.0

May 26

20.9

21.1

24

1962

23.7

Apr. 28

17.2

19.4

May 12

22.8

23.8

14

the estimated mean air temperature was
14.4° C or higher (Table 1). The first
records of adults were obtained during
the freeze-free period (Fig. 4B) and when
the estimated mean air temperature was
20.8° C or higher. There are five or more
successive generations per year. In this
study, records of diapause eggs were ob-
tained in October. Figure 4C shows the
earliest record of diapause observed in sub-
sequent work. Also, see section on dia-
pause under results of laboratory studies.
The lowest estimated mean air tempera-
ture on the last date of observed activity
was 8.3° C. Much variation occurred in
the last dates of seasonal activity (Fig.
4D). Surely this variation must be related
to the kind of freeze and the freeze re-
sistance of M. mulsanti. In 1960, only two
nymphs in poor physiological condition
were found after the first freeze of -2.7° C;
in 1961, a few nymphs and adults were
alive after eight freezes, the lowest being
-6.7° C.

Influence of climate and weather. An
understanding of the differences in climate
in the three areas of study and of features
of the weather of the specific years of re-
search is an integral part of the interpreta-
tion of the field data. As shown in Figure
1, Pellston, Michigan, normally has cool
summers; whereas Lawrence, Kansas, and
Carbondale, Illinois, have warm to hot
summers. Rainfall is important to the
ecology of M. mulsanti because it deter-

mines the area and the number of places
that can be inhabited. At Pellston and
Carbondale the rainfall is relatively de-
pendable; Lawrence is characterized by
intermittent periods of drought.

Seasonal activity at Lawrence has the
same duration as at Carbondale (average
duration of freeze-free periods: 196 and
186 days, respectively; average monthly
means of July: 27.0° C and 26.3° C, re-
spectively). Study of the records of tem-
peratures for Pellston suggest that seasonal
activity would begin there in late May or
early June and continue into September
with one or more generations per year
(average duration of freeze -free period:
92 days; average monthly mean of July:
18.8° C). The approximate changes in
day length during the periods of seasonal
activity are: at Lawrence and Carbondale,
13.9 hr on May 1 to 15.0 hr on June 15 to
10.0 hr on November 15; at Pellston, 15.4
hr on June 1 to 15.6 hr on June 15 to 11.7
hr on October 1 (Beck, 1968).

At Pellston, in 1955, the summer was
warmer and drier than average. It is prob-
able that the warm temperatures of June
caused an early molt into adults. I found
adults on June 18, the first date of collec-
tion, whereas the earliest dates on the
labels of pinned adults from this area in
the Snow Entomological Museum show
June 28 in 1950 and June 30 in 1923 and
1939. In 1955, the continued warm tem-
peratures in July and August may have

470

The University of Kansas Science Bulletin

resulted in three generations; in the lab-
oratory adults of the second generation
emerged on July 21 and those of the third
generation on August 26. The dryness
brought about low water levels in aquatic
habitats in late July (an accumulated de-
ficiency of 14.1 cm of precipitation) ; shal-
low habitats were dry in early August.

Near Lawrence, in 1955 and 1956, a
widespread drought, the longest and most
severe on record (Climatological Data:
Kansas, 1957), reduced The Pond at the
Natural History Reservation to three small
pools. In contrast, heavy rains in May and
June of 1957 raised the water level of The
Pond causing it to overflow the dam.

At Carbondale during the period of
study from 1959 through 1962, precipita-
tion was normal. Whether the tempera-
tures were warmer or cooler than average
during May was observed to influence the
duration of the nymphal stage and the
time when the first generation became
adults (Table 1).

Use of data to determine frequencies
of morphs. One problem was to decide
which data were the most meaningful to
determine if a relationship could exist be-
tween any one kind of morph and the
environment.

Percentages of the winged morph often
showed great disparities. For example,
between May 18 and June IS, 1961, at
Fisher Pond, Carbondale, among 577
nymphs and 920 adults, none of the
nymphs were the winged morph but 54%
of the adults were winged. In this series
of data, the unmarked adults were counted
each day, marked and released. Because
this sampling started at the beginning of
seasonal activity, it was possible to observe
when the different stages were most nu-
merous: fourth instars on May 22, fifth
instars on May 28, flightless adults on June
1 and winged adults on June 11. Two
facts, the absence of nymphs of the winged
morph and the numerical dominance of

the winged adults when there were few
fourth and fifth instars, suggest the possi-
bility of long distance migration. These
results showed that the frequencies of the
morphs as nymphs would be more reliable
to show ratios among kinds of morphs.

Data from all field samples of fourth and
fifth instars showed a range of 0 to 21%
for the winged morph, 0 to 24° o for the
intermediate morph and 55 to 100% for
the wingless morph.

However, the primary problem was
whether or not the environment could
influence the development of wing pads.
A percentage based on the number of
nymphs with wing pads, i.e., fourth in-
stars with wing pads and fifth instars of
the winged and intermediate morphs, in
the total sample was used to compare the
results obtained in different field samples.
A secondary problem was whether or
not the environment could influence the
development of wings; the percentage was
based on the number of nymphs of the
winged morph among all of the nymphs
with wing pads.

The first critical period. "It must, in-
deed, be common ground that the diverse
potentialities latent in the gene system
become manifest only when the cell or
organism are in the appropriate environ-
ment or at the appropriate stage of de-
velopment" (Wigglesworth, 1961:112). I
think that the frequency of nymphs with
wing pads in M. midsanti must be related
to temperature during the first critical
period.

The stage of the insect when the de-
velopment of the thorax is sensitive to
the environment is unknown. The first
critical point docs not appear to occur at
fertilization because results of laboratory
crossing experiments did not suggest a
genetic relationship between the morphs
of the progeny and those of their parents.
The embryo does not appear to be the
sensitive stage because there are records

liioRAcic Polymorphism in Mesovelia mulsanti

471

from the field of high frequencies of
nymphs wiih win^ pads when embryonic
development took place at low as well as
at high temperatures. For example, at
Fisher Pond in 1962, the embryonic de-
velopment which took place in April with
a monthly mean of 12.8° C resulted in
May in 18°^o (N=:571) of the nymphs
with wing pads; the embryonic develop-
ment which took place in May with a
monthly mean of 23.7° C resulted in June
in 15"; (N = 572) of the nymphs with
wing pads. Therefore, the sensitive stage
appears to take place in the nymph some-
time in the interval beginning with the
first stadium and prior to the late third
stadium.

At Pellston, normally all nymphs re-
gardless of generation pass through the
first critical period when the monthly
means are low. In July, 1955, some nymphs
of a second generation underwent their
first critical period when the monthly
mean was 21.7° C. However, in field
samples the frec^uency of nymphs with
wing pads remained 3°o.

At Lawrence and Carbondale, most
nymphs of the first generation undergo
the first critical period in May when
normally the monthly mean is low. This
was the case in four years (Lawrence,
1957—18.3° C; Carbondale, 1959—20.5° C,
1960—17.6° C and 1961—16.7° C); 1.6%
(N =^ 1,447) of the nymphs had wing
pads. On the other hand, in 1962 at Car-
bondale when the monthly mean of May
was 23.7° C, 18^; (N = 571) of the
nymphs had wing pads in three field
samples: 0°, on May 8 (N = 1), 5.3%
on May 12 (N = 189), 24.1% on May 14
(N=381). For this example. Figure 5
shows the sequence of life history stages
and the ambient air temperatures and
Table 1 records the estimated mean air
and water temperatures.

Comparison of the duration of the
nymphal stage and temperature data from

01 -^

1 1 "I 1 1 1 1

Eggs

First

jnstars

1 L

Second instars

Third instars

Fourth instors

Fifth instars

Adults

Eggs

1 1 1 1 1

10 15

May

25

Fig. 5. The sequence of the life history stages of
the first generation of A/, mulsanti at Fisher Pond,
and the daily fluctuations of air temperatures from
April 23 through May 23, 1962, at Carbondale,
Illinois. The data for the maxima and minima came
from Climatological Data: Illinois (1962). Monthly
means were: 12.8° C, April, and 23.7° C, May.

the field and laboratory (Tables 1 and 8)
suggests that nymphs in the field may
respond to the warmer temperatures of
water rather than to the lower tempera-
tures of air. Indeed, a method of measur-
ing temperatures at the surface film is
needed.

Nymphs with wing pads and tempera-
ture. To relate the percentages of nymphs
with wing pads in field samples to sea-
sonal temperatures, an estimate of the
mean air temperature was made for the
date of each field sample. Study of the

472

The University of Kansas Science Bulletin

Table 2. Numbers of the morphs found as nymphs in field samples during the three tem-
perature intervals. Collections were made at Pellston, Michigan, 1955; Lawrence, Kansas,
1955-1957; and Carbondale, Illinois, 1959-1962; and pooled.

Estimates of mean
air temp, in °C

With wing pads

Winged
morph

Intermediate
morph

Morph

univr

% with
wing pads

Without

wing pads

Wingless

morph

16.7-21.4

23

5

2

1.9

1,589

21.5-28.3-21.5

702

302"

47

14.9

5,994'

21.4-8.3

16

9

9

3.1

1,067

^ Fourth instar with wing pads for which the morph could not be recognized.

" Field count of intermediate and wingless morphs incomplete due to collecting technique.

data suggested three temperature inter-
vals: low temperatures of seasonal warm-
ing, 16.7° to 21.4° C; high temperatures
of summer, 21.5° to 28.3° to 21.5° C; low
temperatures of seasonal cooling, 21.4° to
8.3° C (Table 2). Fourth and fifth instars
of the wingless morph were collected when
the estimated mean air temperature ranged
from 16.7° C in the spring to 28.3° in the
summer to 8.3° C in the fall. When esti-
mates of the mean air temperature were
21.4° C or lower, whether there was sea-
sonal warming or cooling, the frequency
of nymphs with wing pads was consis-
tently low, averaging 2.4°; (N = 2,720).
When estimates of the mean air tempera-
ture were 21.5° C or higher, the frequency
of nymphs with wing pads averaged 15%
(N = 7,045) but in individual samples
ranged from 0 to 45%.

Differences in the percentages of
nymphs with wing pads because of lati-
tudinal differences in temperature were
also observed: 2.9% (N = 445) at Pellston
in contrast to 13.7% (N = 2,654) at Law-
rence and 11. Fo (N = 6,666) at Carbon-
dale.

Nymphs with wing pads and unknown
factors. Much variation (0 to 45%) oc-
curred in the frequency of nymphs with
wing pads when the estimate of mean
air temperature was 21.5° C or higher.
During this temperature interval at Fisher
Pond in 1961, first 0% (N = 92, 10 samples,

June 10 to 18) of the nymphs had wing
pads. Later 5.7%o (N = 743, Table 3, June
19 to July 3) of the nymphs had wing pads;
percentages ranged from 0.0 to 11.5%. At
the same time, at Tar Hill Pond, another
shallow, man-made pond only one quarter
of a mile away, 17%, (N = 795, June 24 to
30) of the nymphs had wing pads. Subse-
quently, at Fisher Pond 44.7% (N = 85,
July 29) of the nymphs had wing pads.
At Tar Hill Pond between July 24 and
30 the population was drastically reduced
so that no accurate frequency of nymphs
with wing pads could be determined on
July 30.

Table 4 records differences in the per-
centages of nymphs with wing pads ob-
served during the temperature interval of
21.5° C or higher. Few nymphs with wing
pads were observed in the populations
studied in 1956, 1959 and 1960. At Law-
rence in 1957 at The Pond, the occurrence
of a high frequency (ll-45%o) of nymphs
with wing pads was the longest on record;
it began when the estimated mean air
temperature was 24.4° C on June 22 and
lasted at least through August 18. At
Carbondale, in 1961 and 1962, high fre-
quencies of nymphs with wing pads were
observed for shorter durations of time:
examples already given in Table 3; Fisher
Pond: 17.7% (N = 656), estimates 23.3°
to 22.8° C, August 30 to September 7, 1961,
17.9% (N = 571), estimates 21.5° to 23.6°

Thoracic Polymorphism in Mesovdia mtdsanti

473

Table 3. Records of nymphs with wing pads at Fisher and Tar Hill Ponds, Carbondale,
Illinois from June 19 to July 30, 1961. Estimated mean air temperatures changed from 22. 8"^

to 25.5° to 24.4° C.

Fisher Pond

Tar Hill Pond

With wing

pads

Without

With wing pads

Without

Date

wing pads
Wing-

wing pads

June-JuK

Winged

Inter-

% with

Winged

Inter-

% with

Wing-

morph

mctiiate

wing

less

morph

mediate

wmg

less

morph

pads

morph

morph

pads

morph

19

1

0

3.7

26

20

0

3

4.5

63

..

,.

....

....

21

1

1

2.4

80

._

_.

....

22

2

2

3.4

113

..

„

....

23

7

13

11.5

154

..

24

2

6

5.1

148

26

33

32.1

125

25

25

30

13.8

344

28

1

3

5.0

76

30

16

7

10.8

189

1

..

„

17

2

....

N.C

3

0

0

0.0

41

__

4-15

..

..

....

....

10

N.C.

....

N.C.

24

„

..

....

1

N.C.

....

N.C.

29

18

20

44.7

47

..

30

-•

0

0

....

2

^ Not collected.

C, May 8 to 14, 1962 and 15.0% (N = 572),
estimates 23.7° to 24.2° C, June 7 to 24,
1962.

When estimates of the mean air tem-
peratures were 21.5° C or higher, there
were years when few nymphs with wing
pads were found as opposed to years when

T.\BLE 4. Percentages of nymphs with wing

pads collected when the estimated mean air

temperature was 21.5° C or higher.

0/

/o

N

Collecting data

with

dates, area

wing

pads

0.5

1,118

July

12^-Sep. 11, 1956:

, Lawrence

36.5=

890

June

5-Aug. 18, 1957=

, Lawrence

6.3

586

May

16-Aug. 18, 1959*,

Carbondale

1.7

479

June

10-Sep. 11, 1960,

Carbondale

14.5=

2.551

June

9-Sep. 16, 1961,

Carbondale

21.4=

K421

May

8-Sep. 6, 1962,

Carbondale

^ First date of field work.

= Field count of intermediate and wingless morphs

incomplete due to collecting technique.

= No collections between Aug. 19 and Sep. 5.

* No collections between Aug. 19 and Sep. 12.

samples contained high percentages of
nymphs with wing pads. Consequently,
one must conclude that unknown factors
plus high temperatures during the first
critical period are necessary for a frequency
of 11 to 45° o of nymphs with wing pads.

The second critical period. The per-
centage of nymphs with wing pads that
are of the winged morph gives a basis for
comparing the influence of the environ-
ment on the developing wings. In differ-
ent samples at Fisher Pond, this percentage
ranged from 33 to 85°'o: 33.3% (N = 42,
June 19 to 28, 1961, Table 3), 47.3%
(N = 38, July 29, 1961, Table 3), 75.0%
(N = 116, August 30 to September 7,
1961), 84.8°; (N = 92, May 14, 1962). The
reason for the differences in percentage is
unknown. Because these data show statis-
tical significance, I think that a second crit-
ical period takes place early to midway in
the fourth stadium of nymphs with wing
pads.

474

The University of Kansas Science Bulletin

^

CO

iH

rt

4->

«

c

• ^

-G

^

c

u:|

o

^-1

r3

1)

_C

bS)

Ui

U

oj

o

'3

-o

— ■'

c

n

OJ

«

03

o

O

rt

•-M

■ — '

OJ

r.

U

-o

rt

c

. —I

C

(— '

<u

c

^

4—1

1/5

r^

u

r^

E

Ui

■A-l

V5

-D

u

-T3

w

O

HJ

ui:

CO

c

J=

Dh

c

r$

(U

W5

f^

-n

cx,

ii

a

Crt

n,

_G

r^

a.

>_

C)

o

O

c

^

t/5

lU

-^

JS

U

4—1

o

bJO

11 1

rt

4~i

1 1

C

iJ

o

Ui

(U

q;

Clh

uS

w

u

«

<

H

c

OJ

c

-a n
u u

U OJ

>^ ,

o

.a c

be
C

^ be i"

f2

J=

a

u

0

c

^

~

CM

r^

t^

o

I^

o

IV)

O

u-\

o

o

C-.

ITS

ON

o

CO
ON

a.

u

O

^ li -

M p.

■^ £

C

> bt '^5

— S "S
o

■- c -c

C 3 C

c

^4-1 *-t

o «

u ^

OJ _'

3
C

r3

u

«;

C

Q>

P

«

«^

u,

u

<

X

OJ

o — -1- r^ O

o '^ C5 r-. o
o Qc' u-N od o

<3 C2 r^i — o

0 0 0^0(0

o c; U-N TT o

O <3 C5 O c>

O p O O c^
O O O O o

CD --H r^l -^ C!

O <^ O wi o

o o6 o -:r o

CO ITN O O VD

O

o

o
o

On

0\

"1

ON

r~4

O u-N O O C5

O C5 o o o

o o o o (^

p
o

o
o

ON

o

ON
ON

U-N

o

o
o

p

o

C3 l/-\ C3

E id

ON

fN

p

Q.

~"

V4-I

X

QC

c

(vj

^

3

3

VJ-I

c

^

>.

>,

_:

CJ

a

o

::

3

^

c/:

^^

-

eC

^^

*^^

c2

OJ

o

>-^

>*

^J

1/5

bJO

o

CO

CO

g

'o

o

u

o

u

Q.

c

h4

c

c

&H

o

^

n
u

o

G

■z) -^

(/)

u

c

C

be

'Sd

<u

_«

3

3

"O

t«

'J

c

aa

Q

Q

V2

>

a r^

l/j

u^

u-N

— '

u-N

, — 1

c^i

r^

TT

u-\

^

in

u

ON,

OJ

Ch

J

— •

u-\

o

QC

0^

c

h

(Z> NO

O '30

O U-N

o (^

C> U-N

O M3

o uK

o r^

O On
o t^

p p
o o

o

o

VD

u-\

ON

TT

U-N

OO

'— *

NO

'-"-^

NO -TT

ON NO

OD

NO

o

_^

-1-

U-N

o

NO

o

r^

o

r<1

o

O

o

^^

o

t-^

p

o

CnI

NO

U-N (V|

OO „
—I NO

s a.

-o' -a

c c

o o

&H CI.

O OJ

J= J3

u-\ r<-) rt-

On

o o

cz> o\

p ^

O ON

o o
o o

p — ;

O ON

o o
o o

— m

ON ON

(Vj NO
00 VD

p p

o o

ON --

-a
c

o

^ ON
u l/^

cj2;

>s

CvJ

C

3

a.

' '

3

-o

nl

c

:^

o

c/5

p-

u

a:

t/i

kJ

ti,

Thoracic Polymorphism in Mcsovtlici midsanti

475

r-4

o
o

so

t^

"^

C\

oo

o
o

OO

as

OS

ro — I

o
o

o — •

o
o

o

o
o

O

o

CM

u-\

VC

tr

o

ON

OO

ON

o

IT)

--r

NO

"7-

ON

r^

o

NO

^^

VO

r^

u-\

iX

1-

•^

in

»1-

r^j

~^

NO

OO

o

O

ON

1^

oc

l/N

OC

o
o

OO
ITS

NO

o

U-N

ON

r^

o

ON

OO

o
o

OO
NO

■vO

o

-^

r^

<=>

o

r--!

o

o
o

OO

O

:^

ro

o

00

^

o

OO

o

1^

o

"^

o

V

w

OO

o

o

o

rg

ro

c

3

•^

r^

OO

d.

ti

rv)

>.

u

<^

3

d.

3

C/3

T3

<

t>

C

OO

c

C

■—

C

c

£

r\

^

«

0

r^

^

u.

X

(^

1-,

U

u

U

O

5J

J5

J=

u.

>>

-C

J=

Oi

^ -«

n

V5

ty;

.« ro

^

£

Uh —

t-'

ix r^

£

o .

^^

NO iv-i

TT

NO

m

NO

r-I

ON

ON

NO
NO

ON

-a
c

y

u

£

4J

U

X

x"

0

o

C

o

B
0

o

w

o
o

C

3

3
CG

ON

I/;

_3

-a

CI
■y.

a

.n

£1,

o

(^1

n

OJ
■y

fNl

c
£

n

"5

3
J3

-o

^

C

t-

c

W

o

L.

rt

H,

c

2

o

S

X

r^

J3

n

c

u,

-a

■y.

U

1^

u

"O

c
c

o
c

>

c

3

o
oc

O

c

o

"0

■y
■y.

"o

r^

rt

3

NO

ir\

oc

■■J

u
c

u

c

o

u

c

«j

bc

o

bi
b*

n

Vh

r^)

3

"s

c

c

CJ

be

C

E

4J

X

'G

u

be
bjj

a

c

_c

• ^

1-

o

C

,

bJL

t*

^^

n

ON

N4-4

>-.

U

~

c

^^

rri

o

c

ra

3

Csl

c

'-4-*

o

3

be

r—

c

o

o

3

3

O

u

<;

s

^

bfj

'■'^

r;

<:n

'w

^

p

3

c

NO

C

—.

tT

_C

d

^

•^

~

'■O

1-

u.

~

x"

oc

a

y

o

C

M-l

'•'-1

'./5

C
3

r^I

X

o

■y

'£

a.

X

O

o

yj

'^

'^

1-

o

o
o

'5

g

C

^o

C

3

o^.

f^

«

rg

CJ

C

be

dJ

o
NO

^

c

•— '

c

E

^

r^

^

^

c

o

o

VJ

Q

* t-

5J

'■A

^

c.

;-5

3

b/.

'J

bjo
c

ij

Ti

X.

X

i/j

c

«

3

M,

c

Si

lU

S

'r£

0

u

g"

C
3

«

^

■y

>

0

3

NO

in

U

E

j:

CI

CI

-O

u

L.

^

3

o

"c^

c

0

'yi

c

3

o.

ra

o:

'C

bo
C

•-J-

C

£

13

E
o

_3

bt

'£

re

NO

a.

«j

5

be

0

'J

CI

"o
o

_3

c

~

^

X

o

i-i

o

u

hJ

o

Wi

c^

^

,

u

3

W5

Ui

o

n

Ci

"O

4-*

X

UJ

o

-5

c

TO

C

3
n
C

be

3
C

CI

CJ
CI

„

r^

u

^

^

c

o

0

a
£

0

CJ

r3
3

o

bi;

3

^
'«

!S

'?

3

(^1

yi

c

12

C
no'

<

o

-r

O
be

O

bJC

r5

3
CI

3
O
CI

J-.

y

3

C

i_

^

<-.

>-■

W
o

1

M-l

o
<u
rt

>

-a
c

o"

M
«

C

V5

n.

X

UJ

u

n

u

«J

Ui

O
O.

C

O

"O

•y
C

_3

"5

u

o

r^

NO

E

E

u

c

o

0
£

CI
'-I

!^

c
£

y
yi

CI

bO
c

'$

c

=i

be

c

n

L^

3

in

,bc

-J5

CJ

2

O

r^

'l-

^

C

C

^

c

-T3

-5

C

.2

c
c

■y
o

c

0

si

3

_3
be

c
o

o

O
O

rg

no'
a.

X

a,

X

o

o

rNj

u

C

a

be

c

CI

£

C

O

>^

<

]^

2

o

tU

o

b£

>

c
15

O

>

C
O

'G

o
c

od

«j

3

(^1

C
_3

y

E

rt

—

,

^

bC

rz

bt

r^

>^

^

J

■-S

2

-7-

(N

c

c

' 'J

Z

£

I. tx:

476

The University of Kansas Science Bulletin

Table 6. Summary of Table 5 showing percentages of morphs in laboratory rearings relative
to the percentages of nymphs with wing pads in parental stocks.

Nymphs
with wing

Laboratory reared progeny

Winged

Inter-

Morph

Total

Wing-

N

Temp.

Area,

pads in

morph

mediate

un-

with

less

in °C

number

parental

morph

known^

wing

morph

of expt.^

stocks

pads

0-10

0.5

7.0

2.6

10.1

89.9

613

23.9

30"0
23.9

27.7

Lawrence
expt. 2
expt. 3

Carbondale
expt. 1
expt. 2
expt. 3
expt. 4
expt. 7

11-45

8.3

30.6

15.6

54.5

45.5

703

26.7

Carbondale
expt. 5

4.9

19.8

4.7

29.4

70.6

676

25.6

Carbondale
expt. 6

0.0

7.9

5.6

13.5

86.5

89

23.9

Lawrence
expt. 4

"^ Nymphs with wing pads for which the morph could not be recognized.
^ Laboratory conditions given in Table 5.

RESULTS OF LABORATORY
STUDIES

Purpose of crossing experiments. The
intent was to learn what kind of morphs
would be produced from different pa-
rental stocks, where field samples indicated
0-10% or 11-45° o of the nymphs with wing
pads. These studies also gave information
on the kinds of crosses which can give
rise to the three morphs. Totals for each
laboratory sample are given; data on the
progeny reared from individual crosses
and from females fertilized in nature are
available from the author.

Parental stoc/{ and filial generation. As
previously mentioned, 2.9% (N ^ 445) of
the nymphs had wing pads in field sam-
ples at Pellston in 1955. However, when
rearing stocks (primarily adults) were col-
lected, the localities contained few or no
fourth and fifth instars. Few progeny
with wing pads were obtained in all but
one reared sample, for which the adult
females were collected on July 2 at Duncan
Bay (Table 5: Pellston expt. 3). This

high percentage (35° o) of nymphs with
wing pads is an unusual occurrence among
mesoveliid nymphs of 45° north latitude.
It was not obtained in the field; a trip to
Duncan Bay at the appropriate time
showed that 9.5°% (N=21) of the
nymphs had wing pads.

The percentage of nymphs with wing
pads in the parental stock was unknown
for the Lawrence expt. 1. Among the
progeny, 1.6% of the nymphs had wing
pads.

In seven studies, the parental stocks
contained 0-10°o nymphs with wing pads
(Table 5: Lawrence expts. 2-3; Carbon-
dale expts. 1-4, 7). In the filial generation,
the percentages of nymphs with wing pads
varied; there were but three winged, a
few intermediate and a preponderance of
wingless young (Tables 5, 6). The winged
progeny were produced by two females
which had been fertilized in nature.

When the parental stocks contained 11-
45/o nymphs with wing pads, the filial
generation reared in the laboratory gave

Thoracic Polymorphism in Mesovclia mulsanti

477

Table 7. Thoracic morphs of the parents of the 109 winged progeny reared in the laboratory.

Area, vear, number of ex

pt.,

Morphs

of parents

Winged

progeny

N

locality, date^

Female

Male

%

(N)

Pellston, 1955

2. Black Lake Pool. June

19

flightless-

unknown

10.0

(1)

10^

3. Duncan Bay. July 2

winged

unknown

33.0

(9)

27

flightless

unknown

23.1

(3)

13

4. Duncan Bay, July 6

winged

unknown

6.2

(1)

16^

6. Virgin cross no. 43

wingless

winged

5.0

(1)

20''

Lawrence, 1956

2. The Pond, July I2-Scp

. 11

winged

unknown

0.8

(1)

127

flighdess

unknown

0.9

(2)

224

Carbondale. 1961

5. Tar Hill Pond,

winged

winged

7.2

(22)

305

June 24-July 12

winged

intermediate

6.8

(5)

73

intermediate

winged

8.6

(20)

233

intermediate

intermediate

12.0

(11)

92

Carbondale, 1961

6. Fisher Pond,

winged

winged

2.6

(5)

190

Aug. 30-Sep. 7

winged

wingless

4.1

(9)

219

intermediate

winged

29.4

(5)

17

intermediate

intermediate

9.3

(6)

64

intermediate

wingless

0.0

(0)

8

wingless

winged

0.0

(0)

38

wingless

intermediate

0.0

(0)

14

wingless

wingless

6.3

(8)

126

' Laboratory conditions given in Table 5.

■ \ flightless adult could be an intermediate or a wingless morph.

^The reason for the difference in this total and the one in Table 5 is because only part of the sample is being

considered here.

high frequencies of nymphs with wing
pads, but no nymphs of the winged morph
were obtained when the fluctuating tem-
peratures averaged 23.9° C (Table 5:
Lawrence expt. 4; Carbondale expts. 5, 6;
Table 6).

Laboratory conditions (Table 5) will
be discussed under "influence of laboratory
conditions."

Morphs of progeny in relation to those
of their parents. Individuals of the wing-
less morph were reared from each of the
nine possible kinds of crosses (three
morphs, sexes taken reciprocally). Among
the samples of laboratory reared progeny
(Table 5), the frequency of the wingless
morph ranged from 46 to 100",. The 46%
was obtained in the Carbondale expt. 5.
In this sample of 76 crosses, 74 yielded 14

to 100% wingless offspring. The two re-
maining crosses lacked wingless progeny
— the only time that this occurred in all
of the research. There was no evidence
that any kind of cross would consistently
produce more wingless progeny than other
kinds of crosses.

The intermediate morph was also
reared from all nine possible kinds of
crosses. Among the samples of laboratory
reared progeny (Table 5), the frequency
of the intermediate morph ranged from 0
to 31°'o, the latter obtained in the Carbon-
dale expt. 5. Among the 76 crosses, 27
had no intermediate offspring; in the
others the numbers of intermediates
ranged from one to 16 and the percentages
from 6 to 71° c. In the Carbondale expt.
6, 20""- of the reared progeny were inter-
mediates; 20 crosses had no intermediate

478

The University of Kansas Science Bulletin

progeny and 36 yielded from one to 11,
equalling 5 to 50° o. There was no evi-
dence that any one kind of cross would
produce more intermediate progeny than
other kinds of crosses.

Winged progeny were obtained from
seven kinds of crosses (Table 7), but not
as yet from crosses of intermediate female
X wingless male and wingless female X
intermediate male. The frequency of the
winged morph ranged from 0 to 30° o
among laboratory reared samples (Table
5). Among the 76 crosses in the Carbon-
dale expt. 5, 26 yielded from one to five
winged progeny, the percentages being 5
to 57%. Among the 56 crosses in the
Carbondale expt. 6, 16 gave one to five
winged progeny, the percentages being
4 to 67%.

In the Carbondale expt. 5 where the
morphs of both parents were known, the
frequencies of the winged morph showed
no statistically significant relation to the
four kinds of crosses (Table 7). In the
Carbondale expt. 6 where the morphs of
both parents were known, the frequencies
of the winged morph likewise showed no
statistically significant relation to four
kinds of crosses (Table 7), those yielding
3 to 9% of the winged morph. The fre-
quency of the winged morph (29° o) for
the two crosses of intermediate female X
winged male, although statistically sig-
nificant, is thought not to be meaningful.
In these two crosses, the percentages of
winged progeny were 21°o (N = 14) and
67%o (N = 3). As mentioned already for
this experiment, 16 out of 56 crosses gave
winged progeny. Study of the 16 per-
centages on winged progeny indicated that
14 crosses for which the percentages
ranged between 4 and 23° o did not show
statistical significance. It is thought to be
a matter of chance that two crosses gave
higher percentages (43%, N = 7, inter-
mediate female x intermediate male, cross
no. 3; 67%, N = 3, intermediate female

X winged male, cross no. 2). In expt. 6
three kinds of crosses lacked winged
progeny. I do not understand these results.
I think that winged progeny may be
equally derived from the nine kinds of
crosses; however, further work is needed
to prove the point. The winged morph
has yet to be reared from two crosses, in-
termediate female X wingless male and its
reciprocal. The results of Carbondale expt.
6 show that additional work with the
cross of intermediate female X winged
male is required.

Influence of laboratory conditions. Ap-
parently laboratory conditions did not in-
fluence the percentages of the different
morphs as these data show good agreement
among different samples in the laboratory
and in the field. The observed limits in
laboratory rearings were 0 to 30° o for the
winged morph, 0 to 31% for the inter-
mediate morph, and 46 to 100° o for the
wingless morph. The observed limits re-
corded for field samples were 0 to 21%
for the winged morph, 0 to 24°% for the
intermediate morph, and 55 to 100% for
the wingless morph.

Proportionately more nymphs had
wing pads in laboratory rearings (observed
upper limit — 54°,; in the Carbondale expt.
5, Table 5) than in field samples (observed
upper limit — 45°.., in the sample of July 29,
1961, Table 3). A suggested reason is the
difference in temperature. For example,
from June 24 to July 31, 1961, in the lab-
oratory the temperature averaged 26.7° C,
range 21.r-32.2°. In July in the field the
monthly mean air temperature was 25.5°
C, extremes 9.4°-35.0°, and the monthly
mean pond water temperature was 26.4°
C, extremes 12.8°-38.3°.

The percentage of nymphs with wing
pads that were the winged morph ranged
from 0 to 86° o in the laboratory (Table 5:
Carbondale: 0%, N = 22, expt. 4; 15.1%,
N = 383, expt. 5; 16.6%, N = 199, expt.
6; Pellston: 85.7%, N = 14, expt. 3).

Thoracic Polymorphism in Mesovelia mulsanti

479

C]um[xiris(Mi of these data on the second
critical period with those obtained in field
samples show that nymphs with wing pads
are more likely to become the winged
morph in the field than in laboratory
rearings. The reason for this ditTerence is
unknown.

In this study there was no evidence that
day length could influence thoracic de-
velopment; see "Discussion."

Because of later work with diapause
(see next side heading), I now think that
long day lengths could have stimulated
the laying of non-diapause eggs in the
following experiments: Table 5, Law-
rence expt. 1 and Carbondale expts. 4, 6, 7.

Diapause. Records of diapause eggs
were obtained in 1960 (Table 5: Carbon-
dale expt. 4). Non-diapause eggs with an
average incubation period of 11.7 days,
range 11 to 12, were recognized from dia-
pause eggs with an incubation period last-
ing from 20 to 227 days. One female, col-
lected in the field as a fourth instar on
September 30, matured in the laboratory,
was mated and 7° ;, (N = 130) of the
eggs were diapause. Six females were col-
lected as adults, four on October 9 and
two on October 13, and in the laboratory
84°o (N = 172) of the eggs were diapause.

In a subsequent study on diapause at
Carbondale, the work started with the
collection of adult females (Galbreath,
1973). In 1971, among females collected
on August 28, 3°o (N = 496) of the eggs
laid in the laboratory between August 28
and September 12 were diapause; among
females collected on September 17, 85%
(N = 748) of the eggs laid between Sep-
tember 17 and 21 were diapause.

Mortality. The mortality from egg to
fourth instar was 91° ^ among progeny in
the Carbondale expt. 5 and 85% in the
Carbondale expt. 6. This agrees with
Hungerford's estimate (1917) of 90/o mor-
tality. I do not think that any significance

should be attached to the high mortality
of nymphs. Everyone who has reared
semiaquatic Hemiptera knows about the
high mortality in the laboratory. I see no
reason to think that a lethal factor exists
for any one kind of individual of M. mttl-
santi as has been postulated for individuals
with the AA genetic constitution in Gerris
(in G. lacitstns, Poisson, 1924; in G. lacus-
tris and G. lateralis, Guthrie, 1959; in
Gerris, Brinkhurst, 1959). Vespsiilainen
(personal communication in 1973) re-
ported that in rearing studies of G. lacustris
he found such an individual to be viable.

Survival. More of the progeny were
reared from eggs laid early in the period
of oviposition because these eggs yielded
a higher proportion of progeny that sur-
vived to a recognizable morph. Records
from the Carbondale expt. 6 showed that
the percentages of survival were 19.5%
(592 progeny from 3,132 eggs) from days
one through four of oviposition and 6.2%
(84 progeny from 1,352 eggs) from days
five through nine. Consequently, in cross-
ing studies it is advantageous to start with
large numbers of crosses and rear only the
eggs laid early in the period of oviposition.

Duration of the tiymphal stage. The
nymphal stage averaged 12 days at 30.0° C
and 20.5 days at 23.9° C (Table 8). These
records agree with my calculations for
Hungerford's data (1917) and the data
for M. mulsanti reported by HofTmann
(1932), respectively.

DISCUSSION

Differing environments are known to
produce different thoracic morphs in some
aquatic Hemiptera {Aphelocheirus, Lar-
sen, 1931; Corixidae, Young, 1965). In the
case of M. mulsanti, field samples of
nymphs and results from laboratory ex-
periments indicate environmental control
of the morphs. In some species of gerrids,
the main influence seems to be environ-

480

The University of Kansas Science Bulletin

Source
of data

Table 8. Duration in days of nymphal stage in the laboratory.

Mean air
temp, in °C

I

II

Stadia
III

IV

Duration of
nymphal stage

Carbondaie

expt. 3'

Hungerford
(1917)

Carbondaie
expt. 4^

Hoffmann
(1932)

30.0

23.9

2.1"
(2-3)=

30*

2.4
(2-5)

9

3.6
(3-4)

23

3.4
(2-6)

71

1.9

2.1

2.7

3.5

12.0

(2-3)

(1-4)

(2-4)

(3-4)

..

30

30

17

2

2

2.1

2.4

2.4

2.8

12.2

(1-3)

(2-3)

(2-3)

(2-3)

(11-14)

9

9

9

9

9

3.1

3.4

4.8

6.5

20.5

(3-4)

(3-6)

(4-7)

(6-7)

(20-22)

23

22

18

4

4

3.2

3.5

4.4

5.4

20.0

(2-5)

(2-5)

(3-6)

(3-9)

(17-24)

71

61

59

47

47

^ Details of laboratory conditions given in Table 5.

" Average in days.

^ Range in days.

* Number of individuals.

mental; in others, both genetic factors and
environmental influences appear to be in-
volved (Vepsaliiinen, 1971a).

In G. odontogaster, rearing data gave
no evidence of genetic control (Ekblom,
1950; Brinkhurst, 1959, 1961; Guthrie,
1959; Vepsiilainen, 1971b). Laboratory ex-
periments suggested that photoperiod di-
rected development toward short-winged,
non-diapause individuals or winged, dia-
pause individuals (Vepsiiliiinen, 1971b).
The critical period appeared to be the
first four stadia. Nymphs from popula-
tions in southern Finland responded to in-
creasing day lengths of 18 hours or longer
to become short-winged adults; nymphs
undergoing the critical point after the
summer solstice became winged adults.
Field samples of morphs supported this
hypothesis. Populations were univoltine
and winged in northern Finland and
Sweden (Vepsaliiinen, 1971b; Ekblom,
1950; respectively). Populations showing
a dimorphic first generation with short-
winged adults emerging first and winged
adults emerging later and a partial second
generation of winged adults were observed
in southern Finland and Denmark (Vep-
salainen, 1971b; Andersen, 1973; respec-

tively). Two studies, a rearing experiment
by Guthrie (1959) and observations on
seasonal differences in ovarian develop-
ment between morphs by Andersen
(1973), substantiated Vepsalainen's obser-
vations that short-winged adults were non-
diapause individuals. Overwintering pop-
ulations in southern Finland consisted of
winged adults from both generations and
only a few short-winged adults, 0.6% in
1970.

The interaction of genetic factors and
environmental influences is not understood
for any gerrid. In G. lateralis, for which
polygenetic control was hypothesized by
Ekblom (1928, 1941, 1949), and in G.
lacustris, for which monogenic control was
hypothesized by Poisson (1924), the ef-
fects of temperature and photoperiod on
nymphs (reported by Guthrie, 1959; dis-
counted by Brinkhurst, 1963) have been
considered. In a preliminary report on
G. lacustris, Vepsiilainen (1971a) thought
that monogenes could be the major control
in univoltine populations but that photo-
period also could be involved in determin-
ing the short-winged morph in the first
generation of bivoltine populations. An-
dersen (1973) pointed out that this hy-

Thoracic Polymorphism in Mesovclia miihanti

48]

pothesis tlid not account ior the increase
of short-winged adtihs that he observed
in autumn.

It should be noted that causation of
morphs in gerroids is not Hkclv to be ex-
pHcable on the basis of any one environ-
mental variable. Undoubtedly there are
a number of interacting intluences that
contribute to the environmental control of
development of morphs. Temperature and
unknown factors acting on the nymphs
were involved in A/, nuilsanti because of
the irregular occurrence of nymphs with
wing pads. If photoperiod were involved
in A/, mulsanti there would have been a
simultaneous occurrence of nymphs with
wing pads among localities within the
same geographical area and a predictable
occurrence of these nymphs from one year
to another.

The possibility of hormonal imbalances
causing flightless morphs in gerroids has
been mentioned (Southwood, 1961; Brink-
hurst, 1963; Wigglesworth, 1964; Novak.
1966; Chapman, 1971; Puchkova, 1971;
Vepsalainen, 1971b), but there are no pub-
lished results to substantiate the hypothe-
sis. In gerroids, the physiological causa-
tion of the developmental differences
between morphs is unknown.

LITERATURE CITED

.A.NDERSEN, N. M. 1973. Seasonal polymorphism and
developmental chantjes in organs of flight
and reproduction in bivoltinc pondskaters
(Hem. Gerridae). Entomoi. Scanilina\ica
4:1-20.

Bkck, S. D. 1968. Insect Photoperiotlisni. .Aca-
demic Press, New York. 288 pp.

Brinkhurst, R. O. 1959. Alary polymorphism in
the Gerroidca (Hemiptera-Heteroptera). J.
.A.nim. Ecol. 28:211-230.

. 1961. Alary polymorphism in the Gerroi-

dea. Verh. Int. Verein. Theor. .\ngc\v.
Limnol. 14:978-982.

. 1963. Observations on wing-polymorphism

in the Heteroptera. Proc. Roy. Entomoi.
Soc. London (A) 38:15-22.

Carlisle, 1). B., .\nd P. E. Ellis. 1967. Abnor-
malities of growth and metamorphosis in
some pyrrhocorid bugs: the paper factor.
Bull. Entomoi. Res. 57:405-418.

Chapnlw, R. F. 1971. The Insects: Structure and
Function, 2nd ed. -American Elsevier Pub-
lishing Co., New York. 819 pp.

Climatolooical Data: Illinois. 1910-1962. U.S.
Dcp. of Com. Weather Bureau, vols. 15-67.

Ci.iMAToLocncAL DArA: Kansas. 1955-1957. U.S.
Dep. of Com. Weather Bureau, vols. 60-62.

Climatoloc.ical Dapa: Michigan. 1955. U.S.
Dep. of Com. Weather Bureau, vol. 60.

Ekblom, T. 1928. Vererbungsbiologische Studicn
iiber Hemiptera-Heteroptera. I. Gcrris as per
Fieb. Hereditas 10:333-359.

. 1930. Morphological and biological studies

of the Swedish families of Hemiptera-Heter-
optera. Part II. The families Mesoveliidae,
Corizidae and Corixidae. Zool. Bidrag Uppsala
12:113-150.

. I 'M 1 . Untersuchungen iiber den Fliigeldi-

morphisiiius bei Gerris asper Fieb. Notulae
Entomoi. Helsinki 21:49-64.

. 1949. Neue Untersuchungen iiber den

Flijgelpolymorphismus bei Gents asper Fieb.
Notulae Entomoi. Helsinki 29:1-15.

. 1950. Uber die Fliigelpolymorphismus bei

Gerris odontogaster Zett. Notulae Entomoi.
Helsinki 30:41-49.

Galbreath, J. E. 1969. Thoracic polymorphism
in Mesorelia tniilsanti (Hemiptera: Meso-
veliidae). Doctoral dissertation. Department
of Entomology, University of Kansas. 96 |ip.

. 1973. Diapause in Mcsorelia miilsauti

(Hemiptera: Mesoveliidae). J. Kansas En-
tomoi. Soc. 46:224-233.

Gilbert, L. 1. and D. S. King. 1973. Physiology
of growth and development: endocrine as-
pects. In Physiology of Insecta (M. Rockstein,
cd.), 2nd ed., \-ol. 1, pp. 249-370. .Academic
Press, New York and London.

Gupta, A. P. 1963. Comparative morphology of
the Saldidae and Mesoveliidae (Heteroptera).
Tiidschr. Entomoi. Amsterdam 106:169-196.

Guthrie, D. M. 1959. Polymorphism in the surface
water bugs (Hemipt.-Heteropt.: Gerroidea).
J. Anim. Ecol. 28:141-152.

Harmeson, R. H., and V. M. Schnepper. 1965.
Temperatures of surface waters in Illinois.
Illinois State Water Survey Rep. of Invest.,
no. 49:1-45.

Hoffmann, C. H. 1932. The biology of three
North American species of Mesorelia (Hemip-
tera-Mesoveliidae). Can. Entomoi. 64:88-95,
113-120, 126-139.

Horvath, G. 1915. Monographie des Mesoveliides.
.Ann. Mus. Nat. Hungary 13:535-556.

Hungereord, H. B. 1917. The life-history of
Mesoielia mulsanti White. Psyche 24:73-84.

. 1953. Concerning Mesovelia cioiiglaseiisis

Hungerford. J. Kansas Entomoi. Soc. 24:
76-77.

HussEY, R. G. 1919. The waterbugs (Hemiptera)
of the Douglas Lake region, Michigan. Occa-
sional Papers Mus. Zool., Unix-. Michigan,
no. 75:1-23.

Jaczewski, T. 1928. Mesoveliidae from the State of
Parana. Ann. Mus. Zool. Polonici 7:75-80.

Jordan, K. H. C. 1931. Zur Biologic von Mesovelia
jttrcata Muls. Re\. Isis Budissina, Bautzen
12:13-19.

. 1936. Die Larve dcr makroptcren Form

von Mesorelia f areata Muls. Rey (Hem.
Het.). Isis Budissina, Bautzen 13:178-179.

. 1951. .Autotomie bei Mesorelia f areata

Mis. R. (Hem. Het. Mesoveliidae). Zool.
Anz. 147:205-209.

482

The University of Kansas Science Bulletin

Larsen, O. 1931. Beitrag zur Kenntnis des Pterygo-

polvmorphismus bei den Wasserhcmipteren.

Acta Univ. Lund (N. F.) Avd. 2 Bd. 27(8):

1-30.
Leth, K. O. 1943. Die Verbeitung der danischen

Wasserwanzen. Entomol. Medd., Copen-
hagen 23:399-419.
Local Climatological Data. Carbondale, 1910

1962. 1964. Illinois State Water Survey

Misc. Pub., no. 19:1-6.
MicHENER, C. D. 1961. Social polymorphism in

Hymenoptera. In Insect Polymorphism ( J. S.

Kennedy, ed.). pp. 42-56. Symp. Roy. Ento-
mol. Soc. London, no. 1.
Miyamoto, S. 1953. Biology of Microrelia dihita

Distant, with descriptions of its brachy-

l^ter()U^ Inrm and larval stages. Sieboldia

1(2):113-133.
. 1964. Semiaquatic Heteroptera of the

South-West Islands, Iving between Kvushu

and Formosa. Sieboldia 3(2) :193-218.
MuLLER, G. 1919. Beitrage zur Rhynchotenfauna

Thiiringens. Entomol. Mitteilungen 8:142-

149.
Neering, T. 1954. Morphological variations in

Mesovelia midsanti (Hemiptera, Mesoveli-

idae). Univ. Kansas Sci. Bull. 36:125-148.
Novak, V. J. A. 1966. Insect Hormones, 3rd (1st

English) ed. Methuen and Co., London.

478 pp.
Poissox, R. 1922. Hemiptcres aquatiques nouveaux

pour la faune normande. Bull. Soc. Entomol.

France 27:75-78.
. 1924. Contribution a IVtude des

Hemiptcres aquatiques. Bull. ImoI. France

Belgique 58:49-305.
. 1933. Note sur les McsorcUa de la faune

francjaise (Hem. Mesoveliiilae). Bull. Soc.
Entomol. France 38:181-187.

PucHKovA, L. V. 1971. The function of the wings
in the Hemiptera and trends in their spe-
cialization. Entomol. Rev. 50:303-309.

Schumacher, F. 1*^*19. Notiz fiber Mesorelia jiircata

Mls.-Rey (Hem.). Entomol. Mitteilungen 8:
195-196.

SouTHWooD, T. R. E. 1961. A hormonal theory
of the mechanism of wing polymorphism in
Heteroptera. Proc. Roy. Entomol. Soc. Lon-
don (A) 36:63-66.

Sprague, I. B. 1956. The biology and morphology
of Hydrotiietra martini Kirkaldy. Univ. Kan-
sas Sci. Bull. 38:579-693.

Thom, H. C. S., and R. H. Shaw. 1958. Climato-
logical analysis of freeze data of Iowa.
Monthly Weather Rev. 86:251-257.

U.S. Dept. Agr., Yearbook of Agr. 1941. Climate
and Man. U.S. Government Printing Office,
Washington, D.C. 1248 pp.

UsixGER, R. L. 1956. Aquatic Hemiptera. In
Aquatic Insects of California (R. L. Usinger,
ed.), pp. 182-228. Univ. California Press,
Berkeley and Los Angeles.

Vepsalainen, K. 1971a. The roles of photoperiod-
ism and genetic switch in alary polymor-
phism in Genis (Het., Gerridae) (a pre-
liminarv report). Acta Entomol. Fennica
28:101-102.

. 1971b. The role of gradually changing

daylength in determination of wing length,
alary dimorphism and diapause in a Genis
odontogaster (Zett.) population (Gerridae,
Heteroptera) in South Finland. Ann. Acad.
Sci. Fennica A, IV, Biol. 183:1-25.

Wesenberg-Lund, C. 1943. Biologie der Siisswas-
serinsekten. Verlag J. Springer, Berlin. 682
pp.

Wigglesworth, v. B. 1961. Insect polymorphism
— a tentative synthesis. In Insect Polymorph-
ism (}. S. Kennedy, ed.), pp. 103-113. Symp.
Roy. Entomol. Soc. London, no. 1.

. 1964. The hormonal regulation of growth

and reproduction in insects. Adv. Insect
Physiol. 2:247-336.

Young, E. C. 1965. Flight muscle polymorphism
in British Corixidae: ecological observations.
J. Anim. Ecol. 34:353-390.

ADDENDUM

Critical point, page 466, column 2, line
5; page 480, column 1, line 13. Vepsaliiinen
(1974c) stated that among individuals o£
this species there could be variation in the
duration of the critical period.

Mortality, page 479, column 2, line 14.
Subsequent studies of G. lacustris (Darn-
hofer-Demar, 1973; Vepsalainen, 1974c)
gave no basis for the hypothesis of a lethal
factor. On the other hand, review of the
available data on G. lateralis by Vep-
salainen (1974c) suggested that the hy-
pothesis could be plausible in this species.

G. odontogaster, page 480, column 1,
line 7. Add Vepsalainen, 1974a.

G. lacustris, page 481, column 1, line 3.
Subsequent studies of G. lacustris have
been published (Darnhofer-Demar, 1973;
Vepsilamen, 1974b, 1974c).

LITERATURE CITED

Darnhofer-Demar, !'). 1^)73 . Zur Populationsdy-
namik cincr unixoltincn Population von
Gerris lacustris (L.) (Heteroptera, Gerridae).
Zool. Anzeiger 190: 1^9-204.

Vepsalainen, K. 1974a. Lengthening of illumina-
tion period is a factor in averting iliapause.
Nature 247: 385-386.

. 1974b. The wing lengths, reproductive

stages and habitats of Hungarian Gerris
Fabr. species (Heteroptera, Gerridae). Ann.
Acad. Sci. Fennica A, IV, Biol. 202: 1-18.

. 1974c. Determination of wing length and

diapause in water-striders (Gerris Fabr.,
Heteroptera). Hercditas 77: 163-174.

}Sf.

i

i

i

s

i

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

MUS. COMP. ZOOL.
LIBRARY

AUG 2 7 W5

HARVARD
U.NJVEIRSrrY.

V

ft:

I

:5

S

s

PHYLOGENY AND CLASSIFICATION

OF THE ACULEATE HYMENOPTERA,

WITH SPECIAL REFERENCE

TO MUTILLIDAE

s:

.V

i

I

s

S

I

By
DENIS J. BROTHERS

i

a

Vol. 50, No. 11, pp. 483-648

August 15, 1975

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out at
the University of Kansas or by University faculty and students. Since its incep-
tion, volumes of the Bulletin have been variously issued as single bound volumes,
as two or three multi-paper parts or as series of individual papers. Issuance is at
irregular intervals, vi^ith each volume prior to volume 50 approximately 1000
pages in length. Effective with volume 50, page size has been enlarged, reducing
the length of each volume to about 750 pages.

The supply of all volumes of the Kansas University Quarterly is now ex-
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased at
$20.00 per volume. Where some of these volumes were issued in parts, individual
parts are priced at the rate of 2 cents per page. Current policy, initiated with
volume 46, is to issue individual papers as published. Such separata may be
purchased individually at the rate of 3 cents per page, with a minimum charge
of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at
the rate of $20.00 per volume. AH communications regarding exchanges, sales
and subscriptions should be addressed to the Exchange Librarian, University
OF Kansas Libraries, Lawrence, Kansas 66045.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

The International Standard Serial Number of this publication is US ISSN
0022-8850.

Editor
Eugene C. Bovee

Editorial Board

William L. Bloom
Philip W. Hedrick

Rudolf Jander

Harvey Lillywhite

Charles D. Michener

Norman A. Slade

Henry D. Stone

George W. Byers, Chairman

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. 50, No. 11, pp. 483-648 August 15, 1975

Phylogeny and Classification of the Aculeate Hymenoptera,
With Special Reference to Mutillidae

Denis J. Brothers

TABLE OF CONTENTS

Abstract 485

Introduction 486

Acknowledgments 488

Methods 489

Investigation of the Aculeata 491

States of the characters considered 491

Character states primitive for the taxa considered 538

Discussion 573

Cladistic relationships 573

Derivation of a classification 578

Investigation of the Myrmosid-Mutillid Complex (Mutillidae) 589

States of the characters considered 589

Characters utilized for derivation of the final cladogram 593

Characters rejected for derivation of the final cladogram 612

Discussion 619

Taxonomic conclusions 619

Life histories and host relationships 626

Geographic distribution 628

General Conclusions 638

Literature Cited 641

Phylogeny and Classihcatiun of the Aculeate Hymcnoptera,
With Special Reference to Mutillidae'

Denis J. Brothers-

AHSTRACT

The phylogeiiv (and classification) of the aculeate Hymcnoptera has not been examined as
a whole since Horner's limited study of the entire Hymcnoptera 55 years ago. The relation-
ships of the members of the "Scolioidea," especially the tiphioid-mutilloid complex, have been
especially confused. This investigation attempts to rectify this situation.

Representatives of 25 taxa of aculeate Hymcnoptera (the taxa varying in categorical level
froni tribe to superfamily in a traditional classihcation, the emphasis being on the "Scolioidea")
were examined with respect to 92 characters. Primitive-derived sequences of states were deter-
mined for these characters and cladograms were constructed by electronic computer ("Wagner
trees") and b\- hand using the principles of Hennig. The most variable characters were elimi-
nated and cladogram construction was repeated until similar cladograms were derived by both
methods. All derived states of all characters were inserted on the cladogram and numerical
measures of distinctness, considering the number of derived states per internode, the number
of species subtended by (i.e., resulting from) each internode and the efficiency (in terms of
unique, parallel or convergent occurrences) of each derived state on each internode, were
calculated. Taxoncmic distinctness of each taxon from every other was calculated, and this
measure was used as a guide in establishing the categorical levels to which the taxa were as-
signed in a classification.

Representatives of 89° ^ of the valid described genera and subgenera of myrmosids and
mutillids were examined and cladograms were derived as for the Aculeata. The final clad-
ogram of the mutillid-myrmosid group was based on 43 characters (involving 61 derived states),
20 of females and 23 of males. Distinctness measures were calculated, based on these charac-
ters only.

These investigations suggest that the aculeate Hymcnoptera comprise three supcrfamilies,
each with numerous families: 1, Bethyloidea, containing Plumariidac, Bcthylidae, Scolebythi-
dac, Cleptidae, Chrysididae, Loboscclidiidae, Dryinidae, Sclerogibbidae and Embolcmidae; 2,
Sphecoidea, containing two informal groups, one (Spheciformes) consisting of Ampulicidae,
Sphecidae, Larridae, Mellinidae, Pemphrcdonidae, Astatidae, Philanthidae and Nyssonidac, and
the other (Apiformes) consisting of CoUetidae, Halictidae, Oxacidae, Andrenidae, Mclittidae,
Fidcliidac, Mcgachilidac, Anthophoridae and Apidae; and 3, Vespoidea, containing two in-
formal groups, one ( Vespiformes) consisting of Tiphiidae, Sapygidae, Mutillidac, Sicrolomor-
phidae, Rhopalosomatidae, Pompilidac, Bradynobacnidac, Scoliidae, Masaridae, Eumenidae and
Vespidae, and the other (Formiciformes) consisting of Formicidae only.

The composition of most families of Vespoidea is unchanged, but the Tiphiidae consists of
only the subfamilies Anthoboscinac, Thynninae, Myzininae, Methochinac, Tiphiinae and
Brachycistidinae. The Bradynobaeninac, with Typhoctinae (including Eotillini), Chyphotinae
and Apterogyninac, is placed in the family Bradynobacnidac. The Mutillidac consists of seven
subfamilies: Myrmosinae, Pseudophotopsidinac, Ticoplinae, Rhopalomutillinae, Sphaeropthalmi-
nac (comprising two tribes, Dasylabrini and Sphacropthalmini, the latter with the subtribes

^Contribution number 1536 from the Department of Philosophy; awarded the Herbert B. Hungerford

Entomology, University of Kansas, Lawrence, Kansas Memorial Prize in Entomology.

66045. Dissertation submitted in partial fulfilment "Present address: Department of Entomology, Uni-

of the requirements for the degree of Doctor of versity of Natal, Pictermaritzburg, South Africa.

485

486

The University of Kansas Science Bulletin

Pseudomethocina and Sphaeropthalmina), Myrmillinae and Mutillinae (comprising two tribes,
Mutillini and Ephutini, the former with the subtribes Mutillina and Smicromyrmina).

The cladogram of the Mutiliidae and the geographic distributions of its component taxa
suggest that the family arose on Laurasia and differentiated in northeastern Africa. The New
World members were possibly derived almost entirely from two independent stocks introduced
to South America from Africa, and the Australasian fauna probably resulted in the main from
a few introductions from South America across Antarctica.

INTRODUCTION

The Hymenoptera is a well-defined
order of endopterygote insects, often con-
sidered to comprise a distinct superorder
(e.g., Mackerras, 1970), and is undoubt-
edly holophyletic (Ashlock, 1971, 1972;
= monophyletic sensii Hennig, 1966a).
Within the Hymenoptera, the suborder
Apocrita (= Clistogastra) is also holo-
phyletic, being characterized by develop-
ment of a constriction between the first
and second abdominal segments and fu-
sion of the first abdominal segment with
the thorax to form the propodeum. This
situation is unique in the Insecta. (Be-
cause of these morphological modifica-
tions, the tagmata of apocritans are re-
ferred to below as "head," "mesosoma,"
and "metasoma," for the head, apparent
thorax and apparent abdomen, following
the suggestion of Michener, 1944.) By con-
trast, the Syinphyta (= Chalastogastra) is
almost certainly a paraphyletic group
{sensii Ashlock, 1971, 1972; not Nelson,
1971) as Rasnitsyn (1969: p. 168, Fig. 273)
has demonstrated, the Apocrita possibly
having been derived from within the
Siricoidea.

Although the Apocrita is often formally
considered to comprise a number of et]uiv-
alent superfamilies (e.g., Riek, 1970), an
informal division into two groups is use-
ful. One of these groups, the Aculeata,
is holophyletic, being characterized mainly
by modification of the ovipositor as a
stinging apparatus (see Oeser, 1961, for
various characters involved). The other
group, the Terebrantia (= Parasitica) is
undoubtedly paraphyletic, comprising the

remaining apocritans, Richards (1956b)
considers these groups more formally as
sections.

The Aculeata is commonly considered
to comprise seven superfamilies ("Beth-
yloidea," "Scolioidea," "Pompiloidea,"
"Formicoidea," "Vespoidea," "Sphecoidea"
and "Apoidea") of which the first two are
judged to be the most primitive in general
(e.g., Evans & West Eberhard, 1970).
(Since this study has led to conclusions
which involve changes in the limits of
taxa previously recognized at the family
and superfamily levels, names used in the
old sense are enclosed in quotation marks
throughout.) Most of these superfamilies
are readily characterized by one or more
unique specializations and thus are clearly
holophyletic. The "Scolioidea" is an ex-
ception, however, since this group appears
to contain those aculeates which do not
clearly fall into any of the other super-
families. It is thus probably paraphyletic
or even polyphyletic {sensii Hennig, 1966a;
Ashlock, 1971, 1972; not Nelson, 1971),
although shown as holophyletic in the
dendrcjgram of Evans & West Eberhard
(1970). The "Scolioidea" further contains
various taxa the placement of which has
been confused. The main purpose of this
investigation is thus the clarification of the
interrelationships of the various taxa com-
prising the "Scolioidea," with special em-
phasis on the "Mutiliidae," and the deriva-
tion of a classification which reflects these
interrelationships adequately. Coinciden-
tally, the study has been extended to cover
the entire Aculeata.

There have been widely differing clas-

PllYLOGENY AND CLASSIFICATION OF THE AcULEATE HyMENOPTERA

487

sifications at the higher levels for the
members of the scolioid complex, notably
the "Tiphiidae" (sensii lato). Thus, de
Saussure (1(S92) considered the entire com-
plex to be a single family (Heterogynes)
and Ashmead (1900, 1903-1904) ^differ-
entiated eight families in this group. Pate
(1947a) provided some clarification of the
genera related to Tiphia, Myzinum, etc.,
and Krombein (1951) included six fam-
ilies in the "ScoHoidea" oi North America.
Tobias (1965) elevated one of these fam-
ilies ("Sapygidae") to the superfamily
level so that he could designate its two
subordinate taxa as families. The "Mutil-
lidae" has consistently been considered an
important member of the "Scolioidea."

The "Mutillidae" is a large, cosmo-
politan group of wasps, the classification
of which has also long been in a state of
confusion. The group included by Lin-
naeus (1758) and other early workers in
the genus Miitilla was much broader than
the present family "Mutillidae." (The fol-
lowing outline excludes those sections
more recently considered to fall in other
famiUes.) Although there were prelimi-
nary attempts by workers such as Latreille
(1809, 1825, 1829), Wesmael (1851), Sichel
& Radoszkovsky (1869), Thomson (1870),
Blake (1871, 1886), Burmeister (1874) and
de Saussure (1892) to develop a higher
level classification within the group (often
by merely describing a few new genera or
subgenera for species which were super-
ficially aberrant), subsequent workers such
as Fox (1899, 1900) again reduced most
genera to synonymy of Mittilla, although
suggesting species groups within the ge-
nus. However, Andre (1899-1903, 1902)
and Ashmead (1900, 1903-4) described
many new genera and each proposed a
different classification for the group. The
differences between the two schemes re-
sulted mainly from tendencies by Andre
to lump many taxa into a single family
(Mutillidae), while Ashmead placed the

equivalent taxa in four families (Cosilidae,
Thynnidae, Myrmosidae and Mutillidae).
Bischoff (1920-21) based his classification
of African species mainly on that of Andre
(1902), with the addition of various tribal
divisions and taking into account Borner's
(1919) phylogeny of the Hymenoptera.
Bradley & Bequaert (1923, 1928) formu-
lated a scheme that combined features of
those of Andre (1902) and Ashmead
(1900, 1903-4). Concurrently they placed
most members of two of Bischoff's tribes
(Trogaspidiini and Smicromyrmini) in a
single genus (Smicromy>'^ne). Skorikov
(1935) proposed a classification which es-
sentially raised the previously recognized
taxa by one categorical step, placing the
two tribes of Bischoff in question as sub-
families. Schuster (1947, 1949) recognized
that previous attempts at a classification of
the group had often been rendered inap-
plicable because they had been based on
too limited material. He proposed a
scheme which attempted to take this into
account even though he apparently saw
few specimens from the Old World. This
scheme was adopted by Krombein (1951)
for Nearctic species. By contrast, Invrea
(1964) used a classification essentially
based only on Old World species for the
Italian fauna, thus perpetuating the type
of arrangement that Schuster had at-
tempted to eliminate.

The present study was initiated as an
attempt to derive a broadly applicable
classification of the "Mutillidae" from con-
sideration of the entire world fauna (speci-
mens were available for approximately 90
percent of the described taxa at the genus
level). After initial investigations it be-
came apparent that the "Mutillidae" as
previously delineated was almost certainly
polyphyletic, as had been suggested by
Schuster (1949). As a result, the investiga-
tion was extended to other members of
the "Scolioidea" in an attempt to find the
smallest holophyletic group upon which a

488

The University of Kansas Science Bulletin

classification could be based. The super-
family "Scolioidea" proved to be at most
paraphyletic (as suggested above), so that
the remaining divisions were added to the
investigation.

Although the study was thus broadened
to include the entire Aculeata, the levels
of accuracy and completeness vary con-
siderably, depending on the problem being
investigated at each stage. Thus the study
of the "Mutillidae" and closely related
groups is the most complete and hopefully
the most accurate. As higher and higher
level taxa were added into the study, time
and kjgistics militated against each being
considered as completely as were the taxa
originally investigated. Further inaccura-
cies may have been caused by the un-
availability of specimens of rare groups,
resulting in the absence of data for some
characters of these. The unavoidable in-
accuracies introduced at the higher levels
as a result of these factors are, however,
of minor importance for the main focus
of the study, viz., the relationships of the
"Mutillidae."

ACKNOWLEDGMENTS

This investigation would not have been
possible without the assistance, encourage-
ment and advice of many individuals, hi
particular, I thank Drs. Charles D. Michener
and Peter D. Ashlock for their unfailing
willingness to discuss many aspects of the
study. Mr. William L. Overal also freely
gave advice. Specimens were generously
placed at my disposal or donated by the
following institutions and individuals: Al-
bany Museum, Grahamstown, South Africa
(through C. F. Jacot-CJuillarmod); Ameri-
can Museum of Natural History, New York
(J. G. Rozen, Jr.); Bernice P. Bishop Mu-
seum, Honolulu, Hawaii (J. L. Grcssitt, C.
M. Yoshimoto, S. Nakata); British Museum
(Natural History), London, England (M. C.
Day, C. R. Vardy); California Academy of
Sciences, San Francisco, California (P. H.
Arnaud, Jr.; H. B. Leech); Departamento de

Zoologia, Sao Paulo, Brazil (N. Papavero);
Florida State Collection of Arthropods,
Gainesville, Florida (H. V. Weenis); Hope
Department, University Museum, Oxford,
England (C. O'Toole); Instituto Miguel
Lillo, Tucuman, Argentina (A. Willink); In-
stituut von Taxononiische Zoologie, Amster-
dam. Netherlands (J. P. Duffels); Kansas
State University, Manhattan, Kansas (C. W.
Rettenmeyer, H. D. Blocker); Koninklijke
Museum voor Midden-Afrika, Tervuren, Bel- ;
giuni (J. Decelle); Museo Argentino de
Ciencias Naturales, Buenos Aires, Argentina;
Museo Nacional de Historia Natural, Santi-
ago, Chile (V. Perez D'A.) ; Museum National
d'Histoire Naturelle, Paris, France (S. Kel-
ner-Pillault); National Museum, Bulawayo,
Rhodesia (E. C. G. Pinhey); National Mu-
seum of Natural History, Washington, D.C.
( K. V. Krombein, A. S. Menke); Rijksmu-
seuni voor Natuurlijke Historic, Leiden,
Netherlands (J. T. Wiebes); Snow Ento-
mological Museum, University of Kansas,
Lawrence, Kansas (C. D. Michener); South
African Museum, Cape Town, South Africa
(A. J. Hesse); Transvaal Museum, Pretoria,
South Africa (J. A. van Reenen); Univer-
sidad de Costa Rica, San Jose, Costa Rica
(A. Wille); University of Arizona, Tucson,
Arizona (C. E. Mickel); University of Cali-
fornia, Berkeley, California (E. I. Schlinger);
University of California, Davis, California;
University of California, Riverside, Cali-
fornia (S. Frommer); University of Minne-
sota, Saint Paul, Minnesota (E. F. Cook, P.
J. Clausen); University of Queensland, Bris-
bane, Australia (T. E. Woodward, E. M.
Exley); F. D. Bennett, Curepe, Trinidad;
W. D. Dyer, Glendora, California; M. A.
Fritz, Buenos Aires, Argentina; A. J. Giai,
S. C. de Bariloche, Argentina; A. R. Hardy,
Sacramento, California; R. C. Heaversedge,
Salisbury, Rhodesia; C. F. Jacot-Guillarmod,
Grahamstown, South Africa; K. Naumann,
S. C. de Bariloche, Argentina; C. W.
O'Brien, Tallahassee, Florida; F. D. Parker,
Logan, Utah; L. E. Peiia G., Santiago, Chile;
T. J. Ridsdill Smith, Armidale, Australia;
C. A. C. Seabra, Rio de Janeiro, Brazil;
F. J. Suarez, Almeria, Spain.

Valuable specimens and data were ob-

Phyloglny and Classification of the Aculeate Hymenoptera

489

rained while I was working on projccrs pri-
marily concerned with studies ol bee biology.
Specimens of immatiires and biological data
on myrmosids and mutillids were obtained
during studies on the origin of sociality
in halictine bees (supported through Na-
tional Science Foundation Grant GB-8588X;
Charles D. Michener, principal investigator).
Manv valuable specimens were also obtained
during a three-month visit to z\rgentina in-
volved with studies on the evolutionary rela-
tionships between bees and Larrea. a sub-
project of the Origin and Structure of
Ecosystems Program. I.B.P. U.S.A. (sup-
ported through National Science Foundation
Grant GB-31I91; Paul D. Hurd, Jr., prin-
cipal investigator). The present study was
thus significantly aided by those grants. In
addition, it was possible to extend the cover-
age of the investigation considerably beyond
that originally intended as a result of finan-
cial support for travel and two one-year
traineeship assistantships awarded me by
the Committee on Systematic and Evolu-
tionary Biology at the University of Kansas
( funded through National Science Founda-
tion Grants GB-4446X and GB-8785). Com-
putation of Wagner trees was done at the
University of Kansas Computation Center,
using the program "WAGNER 702 MOD-
ULE, VERSION 20 /'11/70; HONEYWELL
635 VERSION 4/10/7L"

Finally, I wish to thank numerous friends
and colleagues, too many to list individually,
for their constant interest and encourage-
ment, and their generous hospitality during
my travels in South Africa, the United
States, Argentina and Europe; my parents
for their patience during my protracted ab-
sence in a "foreign land"; and, in particular
Mr. C. F. Jacot-Guillarmod, who first intro-
duced me to the wonderful world of wasps.

METHODS

The methods by which the final cladis-
tic diagrams were derived were essentially
the same no matter which taxonomic
grouping was involved. First, the opera-
tional taxonomic units (OTU's) under
consideration were surveyed in an attempt

to discover as many differentiating char-
acter states as possible. Each character was
re-examined to ensure that the different
states could be unambiguously recognized
and placed in a coded sequence of "primi-
tive" (or ancestral) to "derived." If there
was much uncertainty, the character was
rejected. Then, using the selection of least
equivocal characters remaining, a cladistic
dendrogram was constructed, using two
methods. First, a diagram was constructed
by hand using the principles of Hennig
(1966a) as elaborated by Brundin (1966)
and others, i.e., basing all groups on com-
mon possession of the derived state of one
or more characters (synapomorphies).
Second, a diagram was derived by use of
a computer program for construction of
"Wagner trees" (Farris, 1970; Kluge &
Farris, 1969). The hand- and machine-
derived trees, representing different ap-
proaches to the problem (Moss & Hend-
rickson, 1973), were then compared. All
characters were re-examined in the light
of their distribution on the trees and those
characters which showed many separate
derivations or reversals of the same state
were eliminated because of their instabil-
ity. Certain characters were found to be
in need of re-coding, either because of in-
appropriate delimitation of states or mis-
interpretation of the primitive-derived se-
quence. Judgments were based on the
principle that most parsimonious place-
ment of states on the trees was the most
likely to be correct, unless logically contra-
indicated (e.g., a most parsimonious place-
ment requiring the re-appearance of a
complex structure in an identical form
after being lost). During the process of
consideration and rejection or retention of
characters, due importance was given to
the complexity of the character concerned.
Thus, for example, a character involving
a complex of morphological features was
considered to be less likely to have under-
gone multiple independent changes to a

490

The University of Kansas Science Bulletin

similar derived state than was a simple
character involving a single structure.
During the examination process various
additional characters were discovered and
incorporated into the data matrix.

Once all the characters had been re-
evaluated, cladistic diagrams were again
constructed by hand and using the com-
puter. These were again compared, the
characters re-evaluated and new characters
added, etc. This process was repeated until
highly similar or identical cladograms
were derived using the two methods of
construction, indicating that the characters
used presented a distribution of derived
states that were meaningfully correlated.
This final scheme (or highly similar
schemes) also most likely embodied the
maximal number of "uniquely derived
character [states]" and the lowest "coef-
ficient of character-state randomness" (Le
Quesne, 1969, 1972) possible.

The initial studies were done using
selected genera traditionally included in
the "Mutillidae" {sensii Schuster, 1947,
1949) as well as representatives of the
various subordinate taxa considered to
comprise the "Tiphiidae" {sensit Pate,
1947a), the family thought to be most
closely related to the "Mutillidae." These
investigations led to two independent lines
of study, viz., relationships between the
Mutillidae {sensit stricto) (the character-
istics of which had been established by
the initial study) and other taxa within
the traditional "Scolioidea" and eventually
the Aculeata (taxa being added at higher
and higher levels as the study progressed),
and the relationships within the Mutil-
lidae.

For the studies involving the Aculeata
the characteristics for the various taxa
were derived from relatively few species
chosen to cover the variation within the
taxon as completely as possible, but with
emphasis on the presumably least special-
ized forms. The categorical levels of the

taxa involved varied from tribe (in the
"Mutillidae") to family (for most other
"Scolioidea") or superfamily (for the non-
scolioid aculeates). The material exam-
ined in each instance is listed in the ac-
count of the primitive characters of each
taxon, below.

The general usefulness of any classifi-
cation depends on the purposes for which
it was drawn up. In the present study the
attempt has been made to formulate a
classification of as general applicability as
possible. A classification appears to be of
broadest use when it is based neither ex-
clusively on raw phenetic data divorced
from any consideration of evolutionary
pathways ("pure phenetics," as advocated
by Sneath & Sokal, 1973, and others) nor
derived from a rigid insistence upon
strictly holophyletic taxa, paraphyletic
groups being inadmissible ("pure cladis-
tics," as proposed by Hennig, 1966a, 1969,
necessitating a complex numbering sys-
tem; and others), but considers both these
aspects, as well as chronistic information
(which is present in a relative sense in any
cladogram), and may be termed "natural."

Such a classification may be based on
a cladogram, with the limits and ranks of
the included taxa being delineated by
phenetic considerations. Sneath & Sokal
(1973) consider that "basing taxonomy on
all three approaches [i.e., phenetics, cladis-
tics and chronistics] requires art or com-
promise, both of which are inadmissible as
bases for a precise science." The present
study is an attempt to achieve the ideal
by minimizing this "requirement," as has
been called for by Hull (1970).

As Tuomikoski (1967) and Ashlock
(1971, 1972) have pointed out, paraphy-
letic groups fulfill the basic requirements
of monophyly which have always been re-
garded as essential properties of good taxa.
There is thus no reason to discard taxa
which are found to be paraphyletic if they
are phenetically about as homogeneous as

Phylogeny and Classification oi- the Aculeate Hymenoptera

491

arc holopliyletic taxa at the same cate-
gorical level. On the other hand, a classifi-
cation incliiciiiig paraphyletic taxa should
not be used as the basis for studies (e.g., zoo-
geography) which require accurate knowl-
edge of branching patterns displayed in
the cladoirram from which the classifica-
tion was derived. (The ntimerical meth-
ods used to obtain indices of phenetic
distinctness for the various groups in a
cladogram are discussed below, in the sec-
tion on "derivation of a classification" for
the Aculeata.)

Previous classifications and evolution-
ary schemes of the higher Hymenoptera
have generally not been based on strict
consideration of synapomorphies. This de-
fect was encouraged by the presence of
trends which appear in various lines ap-
parently independently but are character-
istic of many of the more derived groups.
For example, there is a general tendency
toward reduction of the decree of articu-
lation between the various mesosomal
sclerites, especially in the pleural region,
expressed in various ways in the higher
Aculeata. The consideration of such
trends as being characteristic of a single
line has been responsible for the conven-
tional (inadequate) view involving se-
quential divergence of various taxa from
a single evolutionary Hne (e.g., the scheme
provided by Evans & West Eberhard,
1970).

A detailed study of the hymenopteran
ovipositor by Oeser (1961) has, however,
used Hennigian principles in the deriva-
tion of a very limited cladogram which
demonstrates that the Aculeata in the
broad sense are holophyletic [females have
lost the cerci, section 1 of the gonocoxite
IX is dorsoventrally constricted, and the
basal portion of the "notum" of gonapo-
physis IX is detached (terminology of
Smith, 1970a & b) ; all three are derived
states occurring in the Aculeata only], as
are its two component sister groups, the

"Bethyloidea" and the remaining Aculeata
(sensti stricto). This study, admittedly
based on a limited suite of characters albeit
ones involved in a structural system of
great complexity, has provided the ration-
ale for restricting the present study to the
Aculeata. Comparisons were also made
to the Trigonalidae (Terebrantia) since
these have been considered representative
of the possible ancestors of the Aculeata
(Lanham, 1951; Riek, 1970). Decisions as
to the direction and course of evolution
of some characters were also aided by
reference to a few members of the Sym-
phyta (Argidae, Cimbicidae, Diprionidae,
Siricidae, Tenthredinidae, Xyelidae) and
Terebrantia (Braconidae, Chalcididae,
Gasteruptiidae, Ibaliidae, Ichneumonidae).

INVESTIGATION
OF THE ACULEATA

States of the Characters Considered

The characters used in the analysis of
the Aculeata vary greatly in scope and
plasticity. Some characters are uniform in
all taxa examined except for a single taxon,
these representing unique evolutionary in-
novations serving to differentiate only the
taxon bearing the derived state. Such
autapomorphies do not associate taxa and
are of minor phylogenetic significance.
They do, however, contribute to the phe-
netic component of the degree of differ-
entiation of the taxon in question from its
sister group. In such cases it is usually
relatively easy to decide which state is
primitive and which is derived; the more
common state is primitive. Other charac-
ters present synapomorphies which serve
to associate two or more taxa on a com-
mon branch of the cladogram bv virtue
of their sharing the uniquely derived state
of the character. The direction of evolu-
tion in such characters may sometimes not
be obvious and must then be determined
with reference to the states in forms pre-

492

The University of Kansas Science Bulletin

sumed to be generally more primitive than
the Aculeata. The degree of congruence
with the patterns of derivations presented
by other characters is also significant.

Some states that are seemingly synapo-
morphous may not represent unique deri-
vations; a superficially similar derived
state may occur in more than one location
on the cladogram but in remote sections
of it (resulting in convergences). Such a
situation reduces the usefulness of the
character to some extent. Other characters
show tendencies toward the evolution of
similar derived states numerous times,
often in the same general section of the
cladogram, expressed as parallelisms. Al-
though such plastic characters are mostly
of minor use in associating taxa on par-
ticular branches, they are often useful in
indicating which of two taxa associated
on the basis of other characters may be
considered the more highly derived. In
such cases the character involves the "ten-
dency toward" a particular derived state.
If formulated in this way the derived state
may be envisioned as having evolved (in
some preliminary but unexpressed fash-
ion) some time before its appearance on
any of the branches of the cladogram, per-
haps as the result of accumulated muta-

21

.1.1

21.3

A

21.1.1

A

21.1

21.2.1

21.2

\ t /^

21

Fig. 1. Diagram of evolution of character 21, show-
ing scheme for numbering the states.

tions in a non-functional set of gene loci
producing a "frozen accident" (see Ohno,
1973).

The numbering system for character
states reflects the pattern of evolution for
the character involved. Each subsequent
digit, reading from the left, refers to a
state derived from that expressed by the
digits preceding it. This is illustrated in
Fig. 1, which shows the pattern of evolu-
tion of character 21. Where a particular
derived state is characteristic of an entire
taxon it is referred to below as being pres-
ent "in" the group concerned. If a derived
state is found in only some members of
a taxon and is not characteristic of the
entire taxon, it is referred to as occurring
"within" the group involved. Placement
of the various derived states on "inter-
nodes" refers to the final cladogram (Fig.
2). Where examples are given below, the
use of a genus name does not necessarily
imply that all members of the genus show
that character state, but that one or more
species do; nor are examples of groups
containing a few members with a par-
ticular state meant to encompass all such
groups. The ending "-id" also does not
imply family status.

Although it is inappropriate to refer to
taxa as being "primitive" or "advanced,"
since each taxon exhibits characters in
both the "primitive" and "advanced" (or
"derived") states, it is convenient to refer
to taxa in this wav on occasion. When a
taxon is referred to as being "primitive"
this generally means that it originates low
on the cladogram and bears relatively few
characters in derived states.

Those characters considered to be most
useful in the final construction of the
cladogram are marked with an asterisk
(*). Even in these, all derived states are
usually not equally useful and in some
instances one state is practically of no sig-
nificance whereas another is of great im-
portance.

Phylogeny and Classification of the Aculeate Hymenoptera

493

scolid.

vespid

formic.

myrmos.
9T apids

sphecd.

Fig. 2. Cladogram of 25 taxa of Acultata, based on 92 characters. Lengths of branches (measured from the
circumference of each circle) are proportional to the distinctness measures (DC) derived below.

494

The University of Kansas Science Bulletin

Certain characters were confirmed or
derived from studies by Reid (1941) on the
mesosoma, Snodgrass (1941) on the male
genitaUa and Oeser (1961) on the female
genitalia. Genitalic terminology is that of
Smith (1969, 1970a & b). Wing venation
is based on the terminology of Hamilton
(1971, 1972a & b), derived from that of
Ross (1936). The number of closed cells
includes the costal cell if that cell is pres-
ent and closed. The symbols used for the
behavioral formulae are based on those of
Iwata (1942, 1950) as used by Evans
(1966b) but with some modifications,
mainly those necessary to include the bees
in the same scheme. The symbols are as
follows: Preparation of next = I (in-
struere); Searching for larval provisions
= V (venari); Transport of provisions
= T (transjerre); Preparation of provi-
sions— paralysis of animal prey = P
(pungere), preparation (often "mixing")
of pollen-nectar mass == M (rniscere); Ovi-
position = O (ovum parere); Final clo-
sure ^ C (clandere); Provisions — of ani-
mal origin ^ subscript a. of vegetable
origin := subscript v. Other terminology is
explained where necessary in the section
on the character involved, below.

*1. Sexual dimorphism, general form.
Primitively, sexual dimorphism in general
body form is minimal except that the
male may be slightly smaller than the fe-
male and sometimes members of one sex
may be apterous. 1.1 — The male is con-
siderably more slender than the female
and is very elongate, thus being of quite
different form.

No or only slight sexual dimorphism
is considered primitive because this is the
condition in most non-aculeate Hymenop-
tera and also in most members of the
Aculeata.

State 1.1 is unicjuely derived, occurring
on internode 12-13 and forming a synapo-
morphy which links the myzinid and
methochid groups. However, Pterombrus

(myzinid) has the general sexual dimor-
phism only slight, indicating that state 1.1
may have been reversed within the my-
zinid group, or it may have arisen within
the myzinid group. There is actually a
tendency toward increased slenderness in
the males within the entire branch sub-
tended by internode 7-10 except for the
tiphiid and brachycistidid groups, but its
development to a very similar extreme in
the myzinid and methochid groups leads
to male slenderness being placed as above.
Further, some members of the scoliid
(e.g., Campsomen's) and apid (e.g., Cory-
nura; Eickwort, 1969) groups have mod-
erately slender males, but these are not of
the same form as those included in the
groups possessing state 1.1.

2. Sexual dimorphism, aptery. Primi-
tively, both sexes are fully winged, or,
rarely, both sexes have the wings equally
reduced. 2.1 — Sexual dimorphism is con-
siderable, with the male winged and the
female apterous.

No sexual dimorphism in wing devel-
opment is considered primitive because
this is the condition in most non-aculeate
Hymenoptera and also in most aculeates.

Aptery in the female only has occurred
on numerous occasions independently
throughout the Aculeata. State 2.1 has
developed on internodes 8-9 (associating
the myrmosid and mutillid groups) and
18-21 (where it links five taxa), and in
the plumariid, thynnid, methochid and
brachycistidid groups. In view of its
sporadic occurrence, little weight should
be attached to it even where it does indi-
cate synapomorphy. Furthermore, female
aptery (or brachyptery) has arisen within
the bethylid (e.g., Deinodryinus, Pristo-
cera; Evans, 1964a), myzinid (e.g., Braun-
someria), tiphiid (e.g., Pseudotiphia;
Nagy, 1969b) and formicid (all workers,
some queens, e.g., Dorylus, Eciton; Wil-
son, 1971) groups. [Aptery or brachyptery
in both sexes has arisen within the

Phvlochny and Classiiication of the Aculeate Hymenoptera

495

rhopalosomatid (e.g., Olixon) group;
brachypter\ in males only occurs within
the apids (e.g., Lasioglossiim; Houston,
1970: Perdita; Rozen, pers. comm.).J

3. Sterile caste. Primitively, all female
individuals arc capable of reproduction.
3.1 — There is a specialized caste of females
with reduced or no reproductive capability.

The lack of a sterile female caste is
considered primitive because this is the
condition in all non-aculeates and the vast
majority of aculeate Hymenoptera.

The developinent of a sterile worker
caste (3.1) is characteristic of the entire
formicid group only, and thus provides no
information on relationships at the higher
levels. However, a similar derived state
has evolved independently on various oc-
casions within the apid (e.g., Lasioglos-
siim, Bombiis, Apis, Trigona; Michener,
1974) and vespid (e.g., Vespa, Vespiila;
Wilson, 1971) groups.

4. Pubescence (Fig. 3). Primitively, all
pubescence is simple and unbranched. 4.1
— Some of the body setae, especially those
dorsally at the base of the metasoma, are
branched to some extent, being sub-
plumose or plumose.

Simple pubescence is considered primi-
tive because this is the condition in most
Hymenoptera, both non-aculeates and
aculeates.

The presence of plumose pubescence is
characteristic of the apid group and is one
of the classical characters for distinguish-
ing this group from the sphecids. State
4.1 also occurs on internode 21-22, how-
ever, associating the eotillid and typhoctid
groups. A similar state has developed
within the anthoboscid (e.g., Lalapa) and
mutillid (e.g., Sphaeropthahna, Pseiido-
methoca) groups at least, so that the im-
portance of this character and its useful-
ness for establishing higher groupings is
somewhat diminished.

5. Ch'peiis (Fig. 4). Primitively, the cly-
peus is rather short and transverse so that

Figs. 3-4. Characters of Aculeata. 3, body seta,
showing primitive and derived states of pubescence;
4, head, anterior view, showing primitive and derived
states of clypeus (5 based on Anthobosca, $ ; 5.1 on
Clystopsenella, $ ; 5.2 on Apoica, $ ).

the antennae are inserted below the mid-
dle of the face but not immediately above
the oral cavity, i.e., the clypeus is termed
"moderate." 5.1— The clypeus is extremely
short and reduced to a transverse strip so
that the antennae appear to be inserted
just above the oral cavity. 5.2 — The cly-
peus is somewhat dorsally produced and
increased in height so that the antennae
are inserted at or slightly above the middle
of the face.

A moderate clypeus is considered prim-
itive because this is the condition in most
Aculeata and in particular in those con-
sidered primitive on the basis of other,
stronger characters. Most non-Aculeata
appear to have a slightly larger clypeus,
although the Trigonalidae seem to fall

496

The University of Kansas Science Bulletin

more or less midway between clear expres-
sion o£ either the primitive state or state
5.2.

The clypeal form is difficult to interpret
and is somewhat variable so that the states
are fairly equivocal. However, state 5.1
represents an extreme and is characteristic
of the scolebythid group as well as the
females of the brachycistidid group. It
also occurs within at least the myzinid
group (e.g., female of Braiinsomeria) and
there is a tendency toward this state in the
females of the bradynobaenid group. State
5.2 (representing the opposite trend) is
present in the scoliid group (although
somewhat equivocally so) and more clearly
in the vespids (thus arising on internode
19-20), and in the males of the apterogynid
group. It has also arisen within various
other taxa such as the apid (practically all
members except e.g., Brachyhesma), sphe-
cid (e.g., Sphex, Zyzzyx) and formicid
(especially males, e.g., Paltothyreiis, Caw-
ponotus) groups. In view of the sporadic
occurrence of both derived states and their
equivocal nature, the clypeal form is ac-
corded little weight in the analysis of
relationships.

*6. Antennul soc/{et (Fig. 5). Primitively,
the antennal socket is an approximately
circular foramen bordered evenly by a
slightly raised rim. 6.1 — The dorsomesal
region of the rim is much produced, form-
ing a projecting "tubercle" which is semi-
circular in cross-section and well-differ-
entiated from the interantennal frontal
region. 6.2 — The interantennal front dor-
somesal to the antennal socket is produced
into a "frontal ledge" which is not highly
differentiated from the front dorsally and
is thus distinguishable from state 6.1.

A simple antennal socket is considered
primitive because this is the condition in
most non-aculeate Hymenoptera as well
as most Aculeata.

State 6.1 occurs twice on the cladogram,
on internodes 7-8 (where it associates the

Figs. 5-6. Characters of Aculeata. 5, head, anterior
\ievv, showing derived states of antennal socket (6.1
based on Aiireotilla, $ ; 6.2 on Meria, $ ) ; 6, head,
anterior view, showing primitive and derived states
of eye form (7 based on Anthobosca. S ; 7.1 on
Apterogyna, $ \ 7.1.1 on Aptciogyna, 9 ; 7.2 on
Apoica, $ ; 7.3 on Methocha, $).

sapygid, myrmosid and mutillid groups)
and 21-23 (linking the chyphotid and
apterogynid groups), being subsequently
modified into state 6.2 in the bradynobae-
nid group. State 6.2 occurs elsewhere on
internode 12-13 thus linking the myzinid
and methochid groups although it is not
as well-developed in many of the metho-
chids as in most myzinids. Although there
is a tendency toward a state like 6.2 in
female Plumariiis, this state is not con-
sidered to be fully developed in the plu-
mariid group. States 6.1 and 6.2 are not
as distinct as the notation would indicate,
as witness the apparent derivation of 6.2
directly from 6.1 in the bradynobaenid
group. A state similar to 6.1 also occurs
within the bethylid group (e.g., Apenesia,
Pristocera), so that this character is ac-

Phylogenv and Classification of the Aculeate Hymenoptera

497

tuallv not an extremely efficient indicator
of groupings.

7. Eye form (Fig. ()). Primitively, each
compound eye is large, oval and with the
inner margin (seen from directly an-
teriorly) shallowly sinuate. 7.1— The eye
is somewhat rounded but retains the sinu-
ate inner margin. 7.1.1 — The eye is
rounded and the inner margin is convex.
7.2 — The eye is essentially oval but the
inner margin is deeply incised so that the
shape is reniform. 7.3 — The eye is oval
but the inner margin is convex.

An oval eye with sinuate inner margin
is considered primitive because this is the
condition in most non-Aculeata and also
in those aculeates considered to be the
most primitive on the basis of other
characters.

The differences between the various
states of eye form are very subtle and often
equivocal. Nevertheless, they do seem to
provide useful information on higher
groupings. Although state 7.1 has appar-
ently evolved in parallel in both the chy-
photid and apterogynid groups and its
derivative occurs in the females of both
these groups, there is a trend toward this
state in the bradynobaenid group also, so
that state 7.1 may logically be placed on
internode 21-23 as another synapomorphy
shared by these three taxa. A limitation
is, however, inherent in the usefulness of
this state since similar states have occurred
within the plumariid (e.g., Pliimarius male
— 7.1), brachycistidid (e.g., Brachycistis —
7.1), mutillid (e.g., Chrestomutilla — 7.1;
Sphaeropthabna — 7.1.1), formicid (most
females — 7.1.1) and other groups. State
7.2 is unic|uely derived on the tree and
links the scoliid and vespid groups, falling
on internode 19-20. The strength of this
is somewhat lessened by the occurrence of
a similar condition within the sphecid
(e.g., Trypargiliim), anthoboscid (e.g.,
Lalapa), sapygid (e.g., Sapyga), mutillid

{Rhopalomutilla, Mutilla males), myzinid
(e.g., Myzinum male) and rhopalosomatid
(e.g., Rhopalosoma) groups also. State
7.3 has occurred on at least four separate
occasions, in the plumariid (female),
scolebythid, methochid and brachycistidid
(female) groups. It also has arisen within
various taxa such as the bethylid (e.g.,
Parnopes), sphecid (e.g., Cerceris) and
rhopalosomatid (e.g., Olixon) groups so
that it is of little significance in delineating
relationships.

8. Eye contour. Primitively, the com-
pound eye more or less follows the general
contours of the head. 8.1— The eye is pro-
tuberant and very prominent, being highly
differentiated from the surrounding
cuticle.

A non-protuberant eye is considered
primitive because this is the condition in
most non-aculeates and also in most acu-
leate Hymenoptera.

Although state 8.1 has apparently oc-
curred as a parallelism in both the chy-
photid and apterogynid groups, it may
actually be that there has been a reversal
in the bradynobaenid group to a situation
similar to the primitive state, a hypothesis
possibly supported by the fact that the
eyes do not merge smoothly into the sur-
rounding cuticle but present a disconti-
nuity at their margins in the males of the
bradynobaenids. In this case state 8.1
would have arisen on internode 21-23. In
the absence of any further indications,
parallel derivations are considered more
likely than is the reversal. Elsewhere,
state 8.1 occurs in the brachycistidid group
and also within the plumariid (e.g.,
Pliimarius male), mutillid (e.g., Sphaerop-
thalma) and formicid (e.g., Paraponera
male) groups, especially in those species
of nocturnal or crepuscular habit. This
character is probably thus of minimal sig-
nificance in establishing relationships be-
tween the taxa investigated.

498

The University of Kansas Science Bulletin

9. Eye pores and setae. Primitively, the
compound eye has scattered pores which
penetrate the cuticle between the omma-
tidia and which bear minute setae which
are not readily distinguishable except un-
der extreme magnification (referred to as
"not evident"). 9.1— The setae are readily
visible at magnifications of approximately
lOOX, and are referred to as "short." 9.2—
The sensory setae are obvious at magnifi-
cations of 20X or less and are referred to
as "moderately long." 9.3— The pores and
setae are apparently completely absent and
are not visible under magnifications of
lOOX.

An eye with pores and minute setae is
considered primitive because this is the
condition in most non-aculeates (in par-
ticular the Terebrantia, including Tri-
gonalidae) and also in most Aculeata.

It is often extremely difficult to dis-
tinguish the various states involved in this
character and each has apparendy arisen
independently on several occasions, so that
they are of minimal significance in the
delineation of the cladogram. State 9.1 is
characteristic of the scolebythid, brachy-
cistidid (male), sierolomorphid, rhopalo-
somatid and formicid groups and also
links the eotillid, typhoctid and males of
the chyphotid groups (on internode 18-
21). It has apparently been modified in
the females of the chyphotids and both
sexes of the apterogynid and bradynobae-
nid groups, the last three groups being
associated by possession of state 9.3 (on
internodes 21-23 for the female and 23-24
for the male). State 9.1 has also evolved
independently within the bethylids (e.g.,
Pristocera). State 9.2 appears in the myr-
mosids and methochids and also within
the apids (e.g., Apis, Coelioxys) and mu-
tillids (e.g., Areotilla). Apart from its
occurrences in the branch subtended by
internode 21-23, state 9.3 appears in the
plumariid (female), brachycistidid (fe-
male) and scoliid groups and also within

the plumariid (e.g., Phtmariiis male) and
mutillid (e.g., Pseudophotopsis) groups at
least.

10. Ocelli. Primitively, three ocelli are
present on the vertex. 10.1 — The ocelli are
completely absent, not even traces being
retained.

The presence of ocelli is considered
primitive because this is the condition in
most Hymenoptera (both Aculeata and
non-aculeates) and also in most insects.

The complete loss of ocelli is almost
invariably associated with aptery although
the inverse relationship does not always
hold. Thus state 10.1 has appeared in the
females of the plumariid and brachycis-
tidid groups and also in the females of the
entire branch subtended by internode 18-
21. It is also present within the bethylid
(e.g., Pristocera female; Evans, 1964a),
mutillid (all females except some Pseudo-
photopsis), myrmosid (e.g., Myrmosula
female), rhopalosomatid (e.g., Olixon)
and formicid (most workers) groups. In
addition, some or all of the ocelli are
greatly modified or reduced (although
traces generally remain) in various mem-
bers of the sphecid group (e.g., some
Bembix; Evans, 1966a). State 10.1 is thus
of little or no cladistic significance.
*11. Genal organ (Fig. 7). Primitively,
the gena is simple, without any specialized
organ. 11.1— The gena bears an appar-
ently secretory organ opening externally
via an invaginated line or scattered pores.

A simple gena is considered primitive
because this is the condition in most non-
Aculeata (if not all of them) and also in
most aculeates.

The development of a genal secretory
organ in the apterogynid and bradynobae-
nid groups is apparently unique in the
Aculeata and thus provides a very good
synapomorphic character associating these
two groups (on internode 23-24). The
organ has a slightly different appearance
in the two groups, forming a slight pro-

Phyloghny and Classification of the Aculeate Hymenoptera

499

14.1

Figs. 7-8. Characters of Aculeata. 7, head, lateral
view, showing derived state of genal organ (11.1
based on Apterogyna, 9 , and Brady nobaeims, $ ,
left to right); 8, antennal scape, anteromesal view,
showing primitive and derived states of radicle axis
and radicle-scape insertion (13, H based on Zyzzyx,
9; 13.1, 14 on Chyplwtes, i ; 13.1, 14.1 on
Atticotilla, (? ).

tuberance in the female of apterogynids
and not being raised in the female brady-
nobaenids. Nevertheless it seems clear that
the two forms are homologous. In male
apterogynids it forms a patch of scattered
pores and is not clearly distinguishable in
bradynobaenid males. The organ may be
analogous to the "felt line" on the second
metasomal tergum (character 70),

*12. Antennal dimorphism. Primitively,
the antenna comprises the same number
of segments (usually 13) in both sexes.
12.1 — The number of antennal segments
is 12 in the female and 13 in the male.

A condition with the antenna 13-seg-
mented in the male and 12-segmented in
the female is unique in the Hymenoptera
to most groups of Aculeata, those aculeates
in which it does not occur being con-
sidered primitive on the basis of other
characters also. This particular se.xual di-
morphism is thus considered derived.

Almost all groups of Aculeata show
state 12.1 with great consistency, this ap-
parently being a condition which evolved
early and which serves to link almost all

the aculeates into a holophyletic group (on
internode 1-4). Only the plumariid, beth-
ylid and scolebythid groups typically show
no sexual dimorphism in the number of
antennal segments. Within the bethylid
group there is much variation in the num-
ber of segments (from 10 to 40; Riek,
1970) but the number is usually the same
in both sexes or fewer in the males. The
importance of state 12.1 is not much di-
minished by the fact that it has apparently
been reversed on rare occasions within
some taxa, such as in the pompilid group
(e.g., Cteniziplwntes; Evans, 1972) and
the mutillid group {Atilliim, Hoplocrates)
where both sexes have the antenna 13-
segmented. In addition, the number of
segments has been reduced to 12 in both
sexes in a few instances, for example
within the vespids (e.g., Belonogaster)
and apids (e.g., Neopasites; Linsley &
Michener, 1939), and some dacetine ants
have as few as four antennal segments in
the female (Brown & Taylor, 1970).

13. Radicle axis (Fig. 8). Primitively, the
axis of the radicle does not deviate much
from that of the remainder of the scape,
so that the scape merely has a somewhat
differentiated annulus basally. 13.1 — The
axis of the radicle forms a marked angle
with that of the remainder of the scape.

A scape with a simple radicle is con-
sidered primitive because this is the con-
dition in most non-aculeates as well as
most Aculeata.

An angulate radicle (13.1) appears
three times on the cladogram, in the
plumariids (female) and on internodes
8-9 (linking the myrmosids and mutil-
lids) and 21-23 (associating the chyphotid,
apterogynid and bradynobaenid groups).
Nevertheless, similar states are present
within various other taxa such as the beth-
ylids (e.g., Bethyhis, Anisepyris), myzinids
(e.g., Meria) and thynnids (e.g., Elaph-
roptera) so that this state is actually not
very useful in delineating relationships.

500

The University of Kansas Science Bulletin

*14. Radicle-scape insertion (Fig. 8).
Primitively, the radicle is demarcated by
a simple annular constriction. 14.1 — The
scape is produced externally into a flange
that forms a cup-shaped depression or in-
vagination into which the radicle is set.

A simple radicle-scape insertion is con-
sidered primitive because this is the con-
dition in most Hymenoptera, both non-
aculeates and aculeates.

The invagination of the radicle appears
once on the tree, on internode 8-9, thus
associating the myrmosid and mutillid
groups strongly. Although a similar de-
velopment has occurred within the bethy-
lid group (e.g., Anisepyris), such species
are almost certainly only remotely related
to the myrmosid-mutillid group, so that
this affects the usefulness of this character
minimally.

Fig. 9. Labio-maxillary complex, ventral view, some-
what diagrammatic, showing primitive and derived
states (15 based on Anthobosca, i ; 15.1 on
Callomelitta, $; 15.2 on Monohia, $, modified;
15.3 on Vedtschenhja, 9 ; 15.4 on Bradynohaentis,
9, cardines may difler from form shown).

15. Labio-maxillary complex (Fig. 9).
Primitively, the labio-maxillary complex
is well-developed but relatively short and
adapted for lapping. 15.1 — The labio-max-
illary complex is elongated by production
of the prementum and stipes only. 15.2 —
The labio-maxillary complex is elongated
by production of the glossa and paraglossa
only. 15.3 — The labio-maxillary complex
is elongated by production of the glossa
only. 15.4 — The labio-maxillary complex
is much reduced in size.

A well-developed but relatively short
labio-maxillary complex is considered
primitive because this is the condition in
most non-Aculeata and also in most
aculeates.

Derived states 15.1, 15.3 and 15.4 are
each present in only one taxon, the apid,
sapygid and bradynobaenid groups re-
spectively, thus not contributing any in-
formation on higher groupings. State 15.2
is apparently possessed in common by the
scoliid and vespid groups, seemingly hav-
ing arisen on internode 19-20. There is
some uncertainty about this, however,
since the labio-maxillary complex is some-
what different in form in the two taxa,
although its slight elongation in both has
involved the same structures. In addition,
many bees have a modification similar to
state 15.2 superimposed on state 15.1.

16. Maxillary palpus. Primitively, the
maxillary palpus is six-segmented. 16.1 —
The maxillary palpus is five-segmented.
16.2 — The maxillary palpus is two-seg-
mented.

A six-segmented maxillary palpus is
considered primitive because this is the
condition in most aculeates and many
non-aculeates, this being the maximal
number of segments in the Hymenoptera.
This is also probably the primitive con-
dition for the Insecta (Matsuda, 1965).

A five-segmented palpus (16.1) is char-
acteristic of the female of the plumariid
group. Nevertheless, similar reductions in

Phvlogeny and Classification of the Aculeate Hymenoptera

501

the number of segments have occurred
within many of the other taxa, such as
the hethyhtl group (e.g., Dissomphahis
male; Evans, l%4a), the male of Plitma-
roidc's in the plumariid group, and the
apid group (where the number of seg-
ments ranges from 6 to 0). This state is
thus not of great significance, especially
since it involves the loss of only one seg-
ment. State 16.2 involves the loss of four
segments and has occurred in the brady-
nobaenid group and elsewhere such as
within the bethylid (e.g., Dissomphahis
female; Evans, 1964a), mutillid {Rhopalo-
miitilla female) and apid (e.g., Apis)
groups. It is thus of no use in establishing
higher groupings.

17. Labial palpus. Primitively, the labial
palpus is four-segmented. 17.1 — The labial
palpus is three-segmented.

A four-segmented labial palpus is con-
sidered primitive because this is the con-
dition in many non-Aculeata and in most
aculeates; it also represents the maximal
number of segments in the Hymenoptera.
The primitive condition for the Insecta is,
however, probably three segments (Mat-
suda, 1965).

The loss of one segment in the labial
palpus has apparently occurred on the tree
at least three times independently, in the
plumariid (female), bethylid and brady-
nobaenid groups. It is also found within
the plumariid group (e.g., males of Myr-
mecopterina, Pliimarius). It is thus of no
use in delimiting higher groupings.

*1S. Hind margin of pronotitm (Fig. 10).
Primitively, the pronotum is large and
somewhat saddle-shaped with its hind
margin nearly straight and only very
slightly anteriorly arcuate. 18.1 — The pro-
notum is shortened with its hind margin
strongly concave in a fairly regular and
somewhat acute parabolic curve (roughly
V-shaped). 18.2 — The pronotum is short-
ened with its hind margin shifted anteri-

18.2

n

Figs. 10-11. Characters of Aculeata. 10, anterior
region of mesosoma, dorsal view, showing primitive
and derived states of him! margin of pronotum (18
based on Anthobosca, $ ; 18.1 on Trielis, 9 ; 18.2
on Cerceris, 9 ) ; 1 1 , anterior region of mesosoma,
lateral view, showing primitive and derived states of
pronotal articulation (19 based on Anthobosca, 9;
19.1 on ScoUa, 9).

orly over almost its entire width (broadly
U-shaped).

A large pronotum with approximately
straight hind margin is considered primi-
tive because this is the condition in those
groups of Aculeata considered primitive
on the basis of other characters, and also
in various of the most primitive non-
aculeates (e.g., xyelids). Furthermore,
elsewhere in the insects reduction in rela-
tive size of the pronotum is generally
derived.

State 18.1 has arisen independently on
at least six occasions, on internodes 18-19
and 21-23 (males), and in the plumariids
(male), mutillids, brachycistidids (male)
and rhopalosomatids. It is thus of little
use in establishing relationships. By con-

502

The University of Kansas Science Bulletin

trast, state 18.2 has apparently arisen only
once, on internode 4-5, thus associating
the sphecid and apid groups. A vaguely
similar condition is present in some mem-
bers of the bethylid group (e.g., Deino-
dryiniis, Euchroeiis) causing some authors
to consider Anipiilex (definitely a sphecid
although it has the pronotum larger than
most other sphecids; see Evans, 1959a;
Leclercq, 1954) closer to (or even a mem-
ber of) the bethylid group (e.g., Nagy,
1969a). The condition in the Trigonalidae
is also vaguely similar to state 18.2. How-
ever, in all three of these instances there
are marked differences in detail (espe-
cially in the form of the posterolateral
angle and the spiracular lobe) so that these
superficially similar states appear definitely
to have had independent origins.

*19. Pronotal articulation (Fig. 11).
Primitively, the attachment of the prono-
tum to the mesothorax is loose and freely
articulating. 19.1 — The pronotum is very
closely attached and coadapted to the
mesothorax so that no or extremely little
movement is possible between them, al-
though the sclerites are not actually fused.

A freely movable connection between
the pronotum and mesothorax is con-
sidered primitive because this is the con-
dition in most non-aculeates as well as
in most Aculeata.

Apparently immovable association of
the pronotum and mesothorax has occur-
red in only two taxa, the scoliid and vespid
groups, and is so similar in both that it
is considered as very good evidence for
their association, this state thus having
arisen on internode 19-20.

20. Pronotal collar (Fig. 12) . Primitively,
the pronotum is expanded anteriorly as a
well-developed flange or "collar" which
aids in the support of the head and pro-
vides protection to the neck region dor-
sally. 20.1 — The pronotum is somewhat
flattened anteriorly and not at all ex-

12

J

Figs. 12-13. Characters of Aculeata. 12, prothorax
and base of head, lateral view, showing primitive
and derived states of pronotal collar (20 composite;
20.1 based on ClystopseneUa); 13, anterodorsal re-
gion of mesosoma, lateral view, showing primitive
and derived states of posterolateral angle of pronotum
(21 based on Anthobosca, 5; 21.1 on Pseudophotop-
sis. $\ 21.1.1 on Scolia, 9: 21.1.1.1 on Rygc/ilnm,
9 ; 21.2 on Callomelitta, 9 ; 21.2.1 on Trypargilnm,
9 ; 21.3 on Brady nohaenus, $ ).

panded, so that the neck region is dorsally
exposed.

A well-developed pronotal "collar" is
considered primitive because this is the
condition in most non-Aculeata (although
generally not well-developed in Sym-
phyta) as well as in practically all acu-
leates.

Although the derived state is ap-
proached in various taxa (e.g., the ves-
pids), it is fully expressed only in the
plumariid (female) and scolebythid
groups, and in a different fashion in each.
State 20.1 is thus of no use in estabHshing
relationships between taxa.

*21. Posterolateral angle of pronotum
(Fig. 13). Primitively, the posterolateral
angle of the pronotum is evenly rounded,
reaching the tegula but not exceeding its
anterior margin, 21.1 — The posterolateral
angle is very slightly dorsally produced so

as to appear truncate anterior

to the

Phvloceny and Classification of the Aculeate Hymenoptera

503

tegula; it attains the tegula but does not
exceed its anterior margin. 21.1.1 — The
posterolateral angle is dorsally produced
and exceeds the anterior margin of the
tegula very slightly so that the lobe is
notched. 21.1.1.1 — The posterolateral angle
is markedly produced dorsally and pos-
teriorly so that it exceeds the level of the
tegula and forms an acute lobe above it.
21.2 — The posterolateral angle is reduced
dorsally above and slightly anterior to the
spiracular operculum; the operculum
forms a highly dilTerentiated lobe reach-
ing the tegula. 21.2.1 — The posterolateral
angle is reduced anterodorsal to the spi-
racular operculum and is somewhat re-
tracted anteriorly so that the highly dif-
ferentiated operculum does not reach the
level of the tegula. 21.3 — The posterolat-
eral angle is slightly posteriorly produced
below the tegula and thus exceeds its
anterior margin slightly, forming a fairly
acute lobe.

A simple posterolateral angle which
reaches the tegula is considered primitive
because this is the condition in many non-
aculeates (especially Symphyta; many
Terebrantia show differing modifications)
and also in many Aculeata, in those taxa
which are generally considered most prim-
itive on the basis of other characters.

A slight dorsal production of the pro-
notal angle (22.1) has apparently occurred
on at least four occasions, on internodes
12-14 and 18-19 and in the mutillid and
rhopalosomatid groups, and is thus not of
much use in establishing relationships.
However, further development of this
structure has occurred in both taxa derived
from internode 19-20; state 21.1.1 arose on
internode 19-20, being present in the sco-
liid group, and its further elaboration
(21.1.1.1) is present in the vespid group.
This derivation of state 21.1.1 is unique
and thus provides good evidence of the
close relationship of the scoliid and vespid
groups. State 21.2 arose uniquely on inter-

node 4-5 and thus is a very good indicator
of the holophyly of the sphecid and apid
groups. Its elaboration (21.2.1) has oc-
curred in the entire sphecid group, al-
though it is not very well-developed in
some (such as Astata) and a similar state
is present in most of the apids (but 21.2
in Callomelitta and Megachile, e.g.). A
superficially similar condition is also pres-
ent within the bethylid group (e.g., Chry-
sis, Eiichroeus) but this anterior retraction
of the spiracular operculum differs in de-
tail from the state in the sphecids, so that
it does not diminish the significance of
state 21.2.1 or its antecedent. State 21.3 is
apparently unique to the bradynobaenid
group although a slight tendency toward
this state is distinguishable in the aptero-
gynids, so that its importance in delimit-
ing higher groupings is minimal.

22. Fosteroventral margin of pronottim.
Primitively, the posteroventral margin of
the pronotum is approximately straight.
22.1 — The posteroventral margin is dis-
tinctly concave.

An approximately straight posteroven-
tral margin is considered primitive because
this is the condition in many non-aculeates
(in particular Symphyta) and also in
many of the Aculeata which are consid-
ered to be the most primitive on the basis
of other characters.

The derived state represents a wide-
spread tendency that has arisen on at least
four occasions and has been reversed at
least once. Thus, state 22.1 appears on
internodes 4-5, 12-14 and 6-15, and in the
methochid group. The apparently primi-
tive state has been regained on internode
21-22. Because of its plasticity and the
somewhat equivocal nature of this char-
acter, it is not of much utilitv in demon-
strating relationships, although it does
seem to have some value in indicating
which taxa are relativelv more advanced
than others.

*23. Ventral angle of pronotum (Fig. 14) .

504

The University of Kansas Science Bulletin

15

24.1

24.2

Figs. 14-15. Characters of Aculcata. 14, pronotum,
anterior view, showing primitive and derived states
of its ventral angle (23 based on Anthobosca, $ ;
23.1 on Synoeca, 9; 23.2 on Ceneris, ?); 15, pro-
thorax, ventral view, showing primitive and derived
states of propleural separation (24 based on Clysto-
psenella, $ , and Eiichroeiis, i , left to right; 24.1 on
Eticlavelia, 9 ; 24.2 on Plum art m, $ ).

Primitively, the ventral angle of the pro-
notum is rounded and does not much (or
at all) exceed the level of the base of the
forecoxa ventrally. 23.1— The ventral an-
gle is acute and produced ventrally be-
yond the forecoxal base. 23.2 — The ven-
tral angle is greatly produced so that it
almost contacts its counterpart midven-
trally.

A rounded ventral angle is considered
primitive because this is the condition in
most non-aculeates as well as in most
Aculeata.

State 23.1 has apparently arisen twice
on the cladogram, on internodes 18-19 (as-

sociating the formicid, vespid and scoliid
groups) and 21-23 (associating the chy-
photid, apterogynid and bradynobaenid
groups). Although it is also weakly pres-
ent within the plumariid group (e.g., male
of Pliinuvius), it seems to be fairly useful
and may actually represent a trend origi-
nating on internode 16-18. The extreme
development of the ventral angle (23.2)
is uniquely present in the sphecid and
apid groups and apparently arose on inter-
node 4-5, thus forming a strong indicator
of the holophyly of this grouping.

*24. Propleural separation (Fig. 15).
Primitively, the propleura are separated
posteriorly, diverging at an angle and thus
exposing the prosternum anterior to the
forecoxae. 24.1— The propleura are mes-
ally contiguous posteriorly and do not di-
verge at an angle; their posterior margins
form a more or less straight line, so that
the prosternum is not visible between the
propleura. 24.2— The propleura are mod-
ified similarly as in 24.1 but they are fused
along the midline both dorsally and ven-
trally.

Posteriorly diverging propleura are
considered primitive because this is the
condition in many non- Aculeata (except
that various taxa highly specialized in
other respects, as well as the Trigonalidae,
show a tendency toward a state approach-
ing 24.1) and in those aculeates which
possess many other characters in their
primitive states.

The primitive extreme is apparently
present in the scolebythid group where
the prosternum is remarkably well-devel-
oped anteriorly and is extensively exposed
between the widely diverging propleura.
Since this condition is more extreme than
in any other taxon, it may be a secondary
development, however. State 24.1 is
uniquely developed in all the taxa sub-
tended by internode 4-6, although the
superficially similar 24.2 is present in the

Phylogeny and Classification of the Aculeate Hymenoptera

505

female plumariids where the propleura
form a rii;icl tube. This character is prob-
ably not as reliable as mi^ht be supposed,
however, since some members of the sphe-
cid (e.g., Crabro) and apid (e.g., Serico-
gaster) groups show conditions approach-
ing state 24.1.

*25. Prosterniim (Fig. 16). Primitively,
the presternum forms an approximately
uniform plane and is not sunken except
perhaps for a very short posterior section.
25.1 — The prosternum is sunken over most
of its surface so that only a short anterior
section (shorter than the sunken section)
is visible ventrally, this section being in a
different plane from the remainder of the

25

25.1

25.1.1

16

Figs. 16-17. Characters of Aculeata. 16, posterior
region of propleura and prosternum, oblique ventro-
lateral view, showing primitive and derived states of
prosternum (25 based on cleptid sp.: 25.1 on
Anthobosca, ?; 25.1.1 on Polybia, ?); 17, posterior
region of prothorax, ventral view, showing primitive
and derived states of forecoxal contiguity (26 based
on cleptid sp.: 26.1 on Anthobosca, $ ; 26.2 on
Clystopsenella, 2 ) .

Sternum. 25.1.1— The prosternum is en-
tirely sunken and not visible ventrally.

A prosternum with most of its surface
in a single plane is considered primitive
because this is the condition in most non-
aculeates. (Even though the prosternum
may be somewhat hidden in many, e.g.,
the Trigonalidae, even in these it gen-
erally forms a single plane.)

State 25.1 has apparently arisen on at
least two occasions in the aculeates, in the
plumariid group and on internode 1-4. It
is additionally present within the bethylid
group (e.g., Pristocera). The second de-
rived state (25.1.1) apparently has a
unique origin (on internode 19-20) and
associates the vespid and scoliid groups.
Since the difference between 25.1 and
25.1.1 is not very great, however, this state
may not be as valuable as its apparent
unit^ue derivation would indicate.

*26. Forecoxal contiguity (Fig. 17). Prim-
itively, the forecoxae are somewhat sepa-
rated basally by the width of the relatively
well-developed prosternum. 26.1 — The fore-
coxae are contiguous basally due to reduc-
tion in the posterior width of the sternal
region. 26.2 — The forecoxae are basally
separated but are posteriorly produced be-
yond the trochanteral insertions so as to
become contiguous apically only.

Separated forecoxae are considered
primitive because this is the condition in
most non-Aculeata, although there is
much variation and the forecoxae ap-
proach contiguity in various groups (in-
cluding the Trigonalidae).

Although there is variation in the in-
tercoxal distance in the bethylid and
plumariid groups so that some members
have the forecoxae almost contiguous,
actual contiguity (26.1) apparently arose
uniquely on internode 1-4. State 26.2 is
uniquely derived and occurs only in the
scolebythid group, thus not contributing
information on relationships.

506

The University of Kansas Science Bulletin

27. Mesonotitm. Primitively, the meso-
notum does not extend anteriorly much
beyond the level of the tegulae. 27.1 — The
mesonotum is mesally anteriorly produced
so that its anterior margin extends much
anterior to the level of the tegulae.

A short mesonotum is considered prim-
itive because this is the condition in those
Aculeata considered to be generally the
most primitive on the basis of other char-
acters, and is also the condition in the
most primitive non-aculeates (e.g., Xyeli-
dae).

State 27.1 has arisen on at least six
separate occasions, on internodes 18-19 and
21-23 (males), in the plumariid (male),
brachycistidid (male) and rhopalosomatid
groups, and somewhat less obviously on
internode 4-5. It is thus essentially useless
for associating groups. This state is logi-
cally linked with the derived states of char-
acter 18, but does not present an identical
distribution on the cladogram and is thus
not considered a duplication.

*28. Scutelluin (Fig. IS). Primitively, the
mesoscutellum is simple, more or less flat-
tened and not highly differentiated from
the remaininLr notal area. 28.1 — The scu-
tellum is enlarged, being posterodorsally
swollen and evenly protuberant. 28.1.1 —
The scutellum is extremely enlarged and
produced so that it overhangs the meta-
notum.

A simple scutellum is considered prim-
itive because this is the condition in most
non-aculeates as well as in most members
of the Aculeata.

Although there are scattered instances
of enlargement of the scutellum (e.g.,
apids, Eiiglossa; mutillids, Trispilotilla;
formicids, Atta) these are of varying types,
and a swollen scutellum is characteristic
of only two of the taxa considered, the
apterogynid and bradynobaenid groups.
State 28.1 has thus apparently arisen on
internode 23-24 and is considered to be
c|uite good evidence of the grouping of

Figs. 18-19. Characters of Aculeata. 18, dorsal region
of mesosoma, lateral view, showing primitive and
derived states of scutellum (28 based on Anthohosca,
9 ; 28.1 on Aplerogyna, c5 ; 28.1.1 on Bradynohae-
ntis, $); 19, anterior region of mesopleurosternum,
ventral and lateral view (as if flattened), showing
primitive and derived states of prepectus (29 based
on Clcptcs, 9 ; 29.1 on Sierolomorpha, 5 ; 29.1.1
on Chyphotes. S • 29.1.1.1 on Polistes, 9; 29.1.1.2
on Tmlis. $■ 29.1.1.3 on Paraponera, $ \ 29.1.2
on Fedtsclicnkia, 9 ; 29.1.2.1 on Psettdopliotopsis,
S ; 29.2 on Chloiion, 9 ).

these two taxa although it has been taken
a step further in the bradynobaenids where
state 28.1.1 is present.

*29. Prepectus (Fig. 19). Primitively, the
prepectus (= epicnemium) is a transverse
sclerite divided midventrally but with the
halves contiguous (or almost so), extend-
ing across the anterior margin of the meso-
pleurosternum and articulating freely with
this margin. 29.1 — Each half of the pre-
pectus is narrowed so that the two sections
become widely separated and do not ex-
tend to the ventral surface of the meso-
soma but do extend the entire height of
the pleural region laterally. 29.1.1 — Each
section of the prepectus is widely separated
from its counterpart and is shortened so
that each half forms a very short but

Phylogeny and Classification of the Aculeate Hymenoptera

507

elongate strip at the anterior margin of
the mesepisternum; the articulation is re-
tained and the prepectal sclerite is hidden
under the posteroventral margin ot the
pronotum. 29.1.1.1— Each prepectal scler-
ite is verv narrow and short, extending
over only the dorsal half or less of the
mesepisternum and articulating with it;
the sclerite is hidden under the postero-
lateral angle of the pronotum. 29.1.1.2 —
Each prepectal sclerite is extremely short
I and narrow, extending over less than the
! dorsal half of the mesepisternum and
fused to it with almost no trace of difTer-
entiation; the sclerite is hidden under the
posterolateral angle of the pronotum.
29.1.1.3 — Each prepectal sclerite is short-
ened but extends over most of the height
of the mesepisternum and is fused to it;
the sclerite is hidden under the postero-
ventral margin of the pronotum. 29.1.2 —
Each prepectal sclerite is not shortened
I and extends the height of the mesepister-
num, being fused to it with the line of
fusion forming a sulcus. 29.1.2.1 — Each
prepectal sclerite is not shortened and ex-
tends the height of the mesepisternum,
being fused to it but with the line of
fusion obliterated except for a pair of pits
ventrally. 29.2 — The prepectus extends
completely across the anterior margin of
the mesopleurosternum, is fused in the
midline and is also fused to the pleuro-
sternum, forming a depressed anterior
margin to it.

The allocation of states of the prepectus
is that which shows greatest correlation
with the groupings made on the basis of
other characters. A state similar to 29.2
is present in Trigonalidae and some chal-
cidoids, however. Although the sequence
and pattern of modification forms a logical
scheme, the homologies of the sclerites in-
volved are somewhat uncertain despite the
studies by Snodgrass (1910, 1935), Mat-
suda (1970) and others. In particular, the
possibility of confusion of the "postspirac-

ular sclerite" (probably the anepisternum)
and the "epicnemium" (prepectus) has
been noted by Richards (1956a; see also
1956b, 1971). Additional investigations,
especially of groups not critically studied
hitherto, should clarify the situation.

The primitive state is present in
the bethylid, scolebythid and plumariid
groups, although some members of the
last (e.g., Plumaroides) have the prepectus
tending towards state 29.1. Apart from
this, state 29.1 has apparently been derived
only on internode 4-6, thus associating the
remainder of the aculeates except for the
sphecids and apids. State 29.1.1 has ap-
parently arisen on three occasions, on
internode 16-18 (where it groups eight
taxa) and in the brachycistidid and rhopa-
losomatid groups. The next modification

(29.1.1.1) has occurred only in the vespid
group. Even greater fusion and reduction

(29.1.1.2) has occurred in the scoUid group
and state 29.1.1.3 is characteristic only of
the formicid group. Modification in a
different direction (29.1.2) has apparently
occurred twice, on internode 7-8 (asso-
ciating the sapygid, myrmosid and mutillid
groups) and in the pompilid group. The
next step in this sequence (29.1.2.1) has
occurred once, on internode 8-9, linking
the myrmosid and mutillid groups even
more strongly. State 29.2 is characteristic
of the sphecid and apid groups and serves
to emphasize their relatively basal position,
this state occurring on internode 4-5. A
few members of the bethylid group (e.g.,
Pseudisobrachium coxalis) show a similar
condition, however, this fact somewhat
reducing the strength of this state in asso-
ciating the sphecids and apids. Although
a more detailed investigation of these
structures is necessary, the pattern of evo-
lution derived here seems to provide use-
ful phyletic information. Although Reid
(1941) misinterpreted the situation in
some taxa (e.g., Mutillidae, sensii stncio),
he also suggested that the prepectus pro-

508

The University of Kansas Science Bulletin

vided useful information for establishing
relationships.

*30. Mesepimeron (Fig. 20). Primitively,
the mesepimeron is differentiated from the
mesepisternum by a distinct pleural sulcus
which is continuously distinguishable from
the pleurointersegmental suture, and ex-
tends from the pleural wing process to a
point just dorsal to the mesocoxa. 30.1 —
The pleural sulcus is indistinct or invisible
over its ventral half so that the mese-
pimeron is not differentiated ventrally al-
though it is apparently not reduced in size.
30.2 — -The pleural sulcus is coincident with
the pleurointersegmental suture over its
ventral half or more, so that the mese-

21

V-^^ty-^

Figs. 20-22. Characters of Aculeata. 20, mesopleu-
ron, lateral view, showing primitive and derived
states of mesepimeron (30 based on Phimarius, $ ;
30.1 on Areotilla, i ; 30.2 on Cliyphotes. $)\ 21,
posterior region of mesosternum, ventral view, show-
ing derived states (31.1 based on Fedtschenkia, 9 :
31.1.1 on Anthobosca, ^\ 31.2 on Apoica, c5 ) : 22,
metanotum, dorsal view, sliowlng primitive and
derived states (34 based on Anthobosca, $, ; 34.1 on
Pseudophotopsis , $ ; 34.2 on Clystopsenella).

pimeron is reduced to a small sclerite at
the posterodorsal angle of the meso-
pleuron.

A complete and distinct mesopleural
sulcus is considered primitive because this
is the condition in most non-aculeates as
well as in those Aculeata considered the
most primitive on the basis of other
characters.

Reduction of the mesepimeron is ap-
parently a general tendency in the more
advanced Hymenoptera and is thus useful
only in indicating relative position on the
tree. Such a general trend cannot, how-
ever, be adequately divided into discrete
states. State 30.1 has apparently been at-
tained twice, in the mutillid and scoliid
groups, and is thus of no importance in
associating taxa. State 30.2 is character-
istic of the apid and formicid groups and
also occurs within the sphecid (e.g.,
Gorytes) and vespid (e.g., Belonogaster)
groups. It has also arisen on internode
18-21 where it serves to associate five taxa.
Its apparent weakness because of multiple
origins is somewhat offset in this last in-
stance in that some fine details of the
modification are constant in these five taxa
but differ from those in other groups.
State 30.2 is thus considered quite good
evidence for the association of the eotillid,
typhoctid, chyphotid, apterogynid and
bradynobaenid groups.

*31. Mesosternum (Fig. 21). Primitively,
the mesosternum is moderately convex
and smoothly truncate posteriorly without
any protuberances or carinae. 31.1 — The
mesosternum has a short transverse carina
or dentate projection anteromesal to each
coxal cavity. 31.1.1 — The mesosternum has
a platelike projection originating antero-
mesal to each coxal cavity and projecting
posteriorly over it. 31.2 — The mesoster-
num is posteriorly produced mesally, car-
rying the mesal points of articulation
(condyles) of the mid-coxae posteriorly

Phylogeny and Classification of the Aculeate Hymenoptera

509

and with the antcromesal maitrins of the
coxal cavities thus somewhat produced.

A simple mesosternum is considered
primitive because this is the condition in
most non-aculeate and in most aculeate
Hymenoptera.

The development of a transverse carina
or small tooth anterior to each mid-coxa
(state 31.1) has apparently occurred on at
least three occasions, on internodes 6-7 and
16-17 and in the sierolomorphid group.
This is thus a rather weak state for demon-
strating relationships. The further devel-
opment of lamellae overlying the coxae
(31.1.1) has occurred twice, on internode
7-10 (where it links six taxa) and in the
rhopalosomatid group. Within the branch
stibtended by internode 7-10 the lamellae
have been reduced to small teeth (an ap-
parent reversal to state 31.1) in the metho-
chid group. Nevertheless, 31.1.1 appears
to provide rather good evidence for the
grouping of the anthoboscid, thynnid,
myzinid, tiphiid, brachycistidid (and
methochid) groups. The posteromesal
production of the mesosternum (31.2) has
apparently occurred twice, on internodes
4-5 (linking the sphecids and apids) and
19-20 (associating the scoliid and vespid
groups), although the presence of state
32.2 in the scoliids obscures this rela-
tionship.

32. Mesocoxal contiguity. Primitively, the
mid-coxae are slightly separated basally.
32.1 — The mid-coxae are contiguous as a
result of a reduction in the intercoxal re-
gion of the mesosternum. 32.2 — The mid-
coxae are very widely separated as a result
of considerable lateral expansion of the
intercoxal region of the mesosternum.

Slightly separated mid-coxae are con-
sidered primitive because this is the condi-
tion in most groups of Aculeata and in
[^articular those taxa considered most prim-
itive on the basis of other characters.

Contiguity of the mid-coxae (32.1) has
apparently arisen on at least three occa-

sions, on internodes 8-9 (in the myrmosids
and mutillids) and 16-17 (linking the
pompilid and rhopalosomatid groups), and
in the vespid group. It thus does not pro-
vide very strong evidence of relationships.
Broad separation of the mid-coxae (32.2)
has occurred twice, on internode 23-24
(female), thus associating the apterogynid
and bradynobaenid groups, and in both
sexes of the scoliid group. Details diflfer
in these two occurrences of the derived
state, so that the presence of 32.2 on inter-
node 23-24 is considered strong evidence
of relationship.

a. Meso-metapleitral suture. Primitively,
the meso- and metapleura articulate freely
with each other. 33.1 — The meso- and
metapleura are closely associated and not
mutually movable although not fused.
33.1.1 — The meso- and metapleura are
fused, at least over the dorsal half, al-
thoutrh the suture is distinct.

Mutually movable meso- and meta-
pleura are considered primitive because
this is the condition in most non-Aculeata
and most insects in general.

Immobility of the ineso-metapleural
suture has occurred on at least four occa-
sions, and is an expression of the general
trend toward consolidation of the meso-
soma in the aculeates. State 33.1 has arisen
on internodes 4-5, 12-14 and 16-18 and in
the pompilid group, and is thus of little
utility except in indicating approximate
relative advance on the tree. The further
fusion of the meso- and metapleura
(33.1.1) has occurred independently in the
formicid and chyphotid groups and is thus
of no use in indicating relationships.

34. Metanotiim (Fig. 22). Primitively,
the metanotum is a transverse sclerite of
approximately the same length mesally as
laterally. 34.1 — The metanotum is slightly
shortened mesally so that it is only a little
more than half as long mesally as laterally.
34.2 — The metanotum is extremely con-

510

The University of Kansas Science Bulletin

stricted niesally and reduced to a minute
strip connecting the lateral areas.

A well-developed metanotum of even
length is considered primitive because this
is the condition in most non-aculeates (in-
cluding Trigonalidae) as well as in most
Aculeata.

Mesal shortening of the metanotum
(34.1) has occurred at least twice, on inter-
node 23-24 (linking the apterogynid and
bradynobaenid groups, in which this state
is at least partly correlated with the meso-
scutellar enlargement) and in the mutillid
group. It is thus not of great utility in
establishing relationships. It has also oc-
curred elsewhere, such as within the myr-
mosid group (e.g., Myrmosa). Extreme
reduction of the metanotum so that the
scutellum and propodeum are almost in
contact (34.2) has occurred on the tree
only in the scolebythid group. A similar
state is present within the bethylid group
(e.g., Epyrinae; Evans, 1964), however,
and this may be of significance in asso-
ciating these groups.

*35. Metapostnotitm (Fig. 23). Primi-
tively, the metapostnotum forms a trans-
verse groove at the anterior margin of the
propodeum, being fused to the propodeum
and slightly depressed. 35.1— The meta-
postnotum is very considerably shortened
and invaginated so that it is barely visible
mesally between the propodeum and the
metanotum. 35.1.1 — The metapostnotum
is completely invaginated and not visible
mesally, the propodeum being in contact
with the metanotum, but the metapost-
notum is still visibly continuous with the
metepimeron laterally. 35.2 — The meta-
postnotum is obscurely distinguishable lat-
erally but merges completely with the pro-
podeum mesally although it is apparently
not invaginated. 35.3 — The metapostno-
tum is greatly enlarged and posteriorly
produced mesally, forming a "triangular
area" which occupies most of the apparent
disc of the propodeum. It carries the mus-

FiG. 23. Anterotlorsal region of metathorax and
propodeum, dorsal and lateral view (as if flattened)
on left and profile on right (dorsal to right), showing
primitive and derived states of metapostnotum (35
based on Endavclia, 9 ; 35.1 on Anthobosca, ^ ;
35.1.1 on Methocha, {, \ 35.2 on Clystopsenella, $;
35.3 on Cerceris, 9).

cles between the second and third phrag-
mata to their insertion apparently far pos-
teriorly on the propodeum.

A distinct and depressed metapostno-
tum of approximately uniform length is
considered primitive because this is the
condition in many non-aculeates, espe-
cially Terebrantia (including Trigonali-
dae).

There is a general tendency above in-
ternode 4-6 toward shortening and invagi-
nation of the metapostnotum as part of
the trend toward mesosomal consolidation.
Although state 35.1 appears only once on
the tree (on internode 6-7), its derivative
(35.1.1) appears at least three times, in the
tiphiid and myzinid groups, and appar-

Phylocuny and Classification of the Aculhatl Hymenoptera

511

enth' de novo on internode 16-18. This
last derivation thus weakens the unique
placement of state 35.1 somewhat, although
35.1 is logically not absolutely necessary
as an antecedent to 35.1.1. State 35.2 is
present in the bethylids and scolebythids,
thus having arisen once (on internode
2-3) and providing useful information as-
sociating these taxa. State 35.3 is unique
in the Aculeata and apparently in the
Hymenoptera as a whole. This remark-
able modification of the metapostnotum
which is present in the sphecid and apid
groups provides extremely strong evidence
of the holophyletic association of these
groups, having arisen on internode 4-5.

Although the various modifications of
the metapostnotum need further investiga-
tion and clarification, the scheme outlined
here provides useful information. In par-
ticular, the above interpretation of the
origin of the "triangular area'' in bees and
sphecids (35.3) seems to be upheld by the
arrangement of sutures (or sulci) in vari-
ous sphecids (e.g., Cerceris) and by the
placement of the muscles between the sec-
ond and third phragmata (2ph-3p/i). Daly
(1964) concluded that in Apis the trans-
propodeal lines (defining the propodeal
triangle) do not result from the migration
of muscles 2ph-3p/i from the sides of the
propodeum to the median pit, but he ap-
parently did not consider that the propo-
deal triangle might represent the meta-
postnotum, a condition which is actually
obscure in this hiiihlv evolved bee.

36. Metapleuron (Fig. 24). Primitively,
the metepimeron is dorsally quite long,
with the pleural sulcus almost or quite
coincident with the meso-metapleural su-
ture dorsal to the endophragmal pit. Be-
low the pit the pleural sulcus is coincident
with the metapleural-propodeal suture.
The pit is very close to the anterior mar-
gin of the metapleuron which is con-
stricted at this point. 36.1 — The metepi-
sternum and to a slightly lesser extent the

36.3

25

37.1

Figs. 24-25. Characters of Aculeata. 24, metapleu-
ron and propodeum, lateral view, showing primitive
and derived states of metapleuron (36 based on
Anthobosca, S ; 36.1 on Psetidophotopsis, $ ; 36.1.1
on Colocistis, $ : 36.1.2 on Trielis, 5; 36.2 on
Euclavelia, 9; 36.3 on Cerceris, 5); 25, metapleu-
ron and propodeum, lateral view, showing derived
state of metapleural gland (37.1 composite).

metepimeron are expanded anteroventral
to the endophragmal pit which is thus
some distance posterior to the meso-meta-
pleural suture. The pleural sulcus dis-
tinctly curves anterodorsally from the pit
and the transepisternal groove may be
visible ventrally. 36.1.1 — The metepister-
num and metepimeron are expanded an-
terior to the endophragmal pit so that the
pit is some distance posterior to the meso-
metapleural suture. The pleural sulcus
and transepisternal suture are very indis-

512

The University of Kansas Science Bulletin

tinct and barely distinguishable if distin-
guishable at all. 36.1.2 — The metepister-
num is posteriorly produced behind the
endophragmal pit, in addition to slight
anterior expansion of the metepisternum
and metepimeron. The pleural sulcus is
angulate behind the pit which is a short
distance from the anterior margin of the
metapleuron. 36.2 — The metepimeron is
expanded and somewhat produced antero-
ventrally so that the endophragmal pit is
shifted posteriorly and the pleural sulcus
is angulate. 36.3 — The metepimeron is
greatly expanded anteroventrally so that
the pleural sulcus issues from a pit just
above the mesocoxa and passes postero-
dorsally to the endophragmal pit.

The allocation of states for the meta-
pleuron is that which is best correlated
with the branching pattern derived from
consideration of other characters, the prim-
itive state in particular being that present
in the taxa which are considered to be the
most primitive on other grounds. Further
investigations are needed to clarify the
situation, however.

The various modifications of the meta-
pleuron, based mainly on the position of
the endophragmal pit as a marker, are
somewhat equivocal and most have oc-
curred more than once, so that they are
not as strong in indicating relationships
as might be expected. State 36.1 has had
at least three origins, on internodes 8-9,
18-19 and 21-23. Since state 36.1.1 occurs
on internode 12-14, a modification similar
to 36.1 perhaps also occurred here. State
36.1.1 is also present in the bradynobaenid
group. State 36.1.2 is uniquely present in
the scoliid group. State 36.2 is present in
the sierolomorphid, pompilid, rhopalo-
somatid, eotillid and typhoctid groups and
thus probably arose three times (on inter-
nodes 16-17 and 21-22 and in the sierolo-
morphids) ; however, if 36.2 could have
been a precursor to 36.1, then 36.2 probably
arose only once, on internode 6-15. A de-

cision is, however, not possible at this
stage so that 36.2 has been placed three
times on the tree. State 36.3 has arisen
once, on internode 4-5, and serves to asso-
ciate the sphecid and apid groups strongly.

37. Metapleitral gland (Fig. 25). Primi-
tively, there is no gland opening to the
exterior on the metapleuron. 37.1 — There
is a gland developed on the metapleuron
and opening via a bulla and meatus just
above the hind coxa.

Absence of a metapleural gland is con-
sidered primitive because this is the con-
dition in the non-aculeate Hymenoptera
and in most Aculeata.

The metapleural gland is a unique
feature of the formicid group and thus
does not serve to indicate any groupings
of the taxa considered here (see Wilson,
Carpenter & Brown, 1967).

38. Metastenium (Fig. 26). Primitively,
the mesal section of the metasternum is
in approximately the same plane as the
mesosternum. The metasternal area is
depressed laterally to accommodate the
mid-coxae, the metasternum being diflfer-
entiated into approximate thirds. 38.1 —
The metasternum is depressed anteriorly
and laterally but not completely so postero-
mesally, the mid-coxae being contiguous.
38.1.1 — The metasternum is entirely de-
pressed, the mid-coxae being contiguous.
38.1.1.1 — The metasternum is entirely de-
pressed but small teeth are developed
just anterior to the metacoxal cavities, the
mid-coxae being contiguous. 38.1.1.2 — The
metasternum is entirely flat and broad but
not depressed, being at the same level as
the mesosternum, with the mid-coxae
widely separated.

A metasternum with only the lateral
thirds depressed is considered primitive
because this is the condition in various
aculeates that are considered to be the most
primitive on the basis of other characters.

Although state 38.1 has apparently

Phvlogeny and Classii-ication of the Aculeate Hymenoptera

513

-"^Es. ^ ~ 38.1.1

'Of "

38.1.1.1

38.1

26

\

38.1.1.2

27

40

40.1.1

Figs. 26-27. Characters of Aculeata. 26, posterior
region of mesosoma, ventral view, showing primitive
and derived states of metasternum (38 based on
PrUtocera, S '■ 38.1 on Aiithohosca, $ ; 38.1.1 on
Tedtschenhja, $; 38.1.1.1 on Pseudophotopsis, $\
38.1.1.2 on Trielis, $); 27, posterior region of
mesosoma, ventral view, showing primitive and
derived states of metasternal anterior production
(40 based on Pristocera, $ ; 40.1 on Typhoctoides,
5; 40.1.1 on Apterogyna, $, and Bradynohaenits,
$ , left to right) .

arisen only twice (on internode 4-6 and in
the plumariid group), its presumed utility
is diminished by the frequent additional
modifications of the metasternum. Thus,
38.1.1 has arisen three times, on internodes
7-8 and 6-15, and in the plumariid group
(male), and has apparently been reversed
on internode 18-21. State 38.1.1.1 is
uniquely characteristic of the mutillid
group and 38.1.1.2 is present in the scoliids
only, so that these two states are of no
use in grouping the taxa.

39. Metasternal differentiation. Primi-
tively, the meso- and metasterna are clearly
differentiated by a definite discontinuity
in the form of a deep sulcus or difference
in level. 39.1 — The meso- and metasterna
are barely differentiated because of fusion
and loss of any definite sulcus, especially
mesallv.

Well-differentiated meso- and meta-
sterna are considered primitive because
this is the condition in most non-aculeates
as well as in most Aculeata.

Loss of differentiation of the metaster-
num is most probably another expression
of the general tendency toward consolida-
tion of the mesosoma. It has occurred at
least three times, on internodes 4-5 and
18-21 and in the scoliid group, so that it
is of little use in establishing relationships.

*40. Metasternal anterior production (Fig.
27). Primitively, the metasternum has the
anterior margin approximately straight.
40.1 — The metasternum is slightly anteri-
orly produced mesally between the mid-
coxae so that its anterior margin attains
the level of the anterior extremities of the
mid-coxae. 40.1.1 — The anterior margin
of the metasternum is anteriorly produced
mesally and reaches a point anterior to the
level of the anterior extremities of the
mid-coxae.

A metasternum with approximately
straight anterior margin is considered
primitive because this is the condition in
most non-Aculeata as well as in most
aculeates.

Anterior production of the metaster-
num has taken place in only one line,
providing good evidence of the holophy-
letic relationship of the eotillid, typhoctid,
chyphotid, apterogynid, and bradynobae-
nid groups. State 40.1 has arisen on inter-
node 18-21, and its derivative (40.1.1) re-
inforces the association of the apterogynids
and bradynobaenids, appearing on inter-
node 23-24.

514

The University of Kansas Science Bulletin

41. Metacoxal contiguity. Primitively, the
metacoxae are nearly or actually contig-
uous. 41.1 — The metacoxae are broadly
separated as a result of lateral expansion
of the intercoxal region of the metaster-

num.

Contiguous metacoxae are considered
primitive because this is the condition in
most non-aculeate and in most aculeate
Hymenoptera.

The derived state is characteristic of
the scoliid group only, and is thus of no
importance in the derivation of higher
groupings.

42. Metathoracic-propodeal pleural suture.
Primitively, the metathoracic-propodeal in-
tersegmental pleural suture is clearly dis-
cernible over its entire length, both dorsal
and ventral to the endophragmal pit. 42.1
—The metathoracic-propodeal pleural su-
ture is completely obliterated ventral to
the endophragmal pit but is distinct dor-
sally. 42.1.1 — The metathoracic-propodeal
pleural suture is completely obliterated,
both dorsal and ventral to the endophrag-
mal pit.

A completely discernible metathoracic-
propodeal pleural suture is considered
primitive because this is the condition in
most non-aculeates.

The obliteration of the metathoracic-
propodeal pleural suture has occurred on
numerous occasions and this character is
thus almost valueless for determining
higher groupings. State 42.1 has appar-
ently arisen on six occasions, on internodes
1-2, 12-14 and 18-21, and in the apid, meth-
ochid and sierolomorphid groups. State
42.1.1 has occurred twice, in the tiphiid
group and (apparently de novo) on inter-
node 8-9. The placement of this state on
internode 8-9 is especially interesting in
the light of Krombein's (1940) statement
that a major difference between the "Myr-
mosinae" and "Mutillidae" lay in the com-
plete absence of this intersegmental suture
in the myrmosids and its presence in the

mutillids. It actually appears that the su-
ture is completely obliterated in both, but
in some of the more highly developed
mutillids there is a secondary development
of a weak carina along the apparent line
of this suture. State 42.1.1 is thus a condi-
tion linking the myrmosid and mutillid
groups, rather than one differentiating
them.

43. Propodeal length (Fig. 28). Primi-
tively, the propodeum is of moderate
length, being at least as long as high, 43.1
— The propodeum is much shortened in
its entirety but especially dorsally, so that
the metanotum extends posteriorly to a
point almost perpendicularly above the
base of the metasoma.

A propodeum of moderate dorsal
length is considered primitive because this
is the condition in most non-aculeates as
well as in most aculeates.

The shortening of the propodeum may

28

Figs. 28-29. Characters of Aculeata. 28, dorsal re-
gion of mesosoma, lateral view, showing primitive
and derived states of propodeal length (43 based on
Anthobosca, 9; 43.1 on Ceramius, $, modified);
29, metapleuron and propodeum, lateral view, show-
ing primitive and derived states of discal distinction
(44 based on Anthobosca, $, \ 44.1 on Colocistis, <$).

Phylogeny and Classimcation oi- the Aculeate Hymenoptera

515

be another expression of the general ten-
dency toward consolidation of the meso-
soma. State 43.1 has occurred twice, the
mechanism dilTering slightly in detail, in
the vespid and hradynobaenid groups. It
is thus of no use in the derivation of
higher groupings.

44. Discal distinction (Fig. 29). Primi-
tively, the disc and declivity of the pro-
[lodeum are not distinct but merge grad-
ually into each other. 44.1— The disc and
declivity form distinct dorsal and posterior
surfaces, often separated by a carina.

A propodeum with merging disc and
declivity is considered primitive because
this is the condition in most non-aculeates
as well as in most members of the
Aculeata.

The degree of distinctness between the
disc and declivity is difficult to quantify.
State 44.1 is considered to be fully de-
veloped when the two surfaces involved
are almost perpendicular to one another.
Such a condition has arisen on at least
three occasions, on internode 12-14 (link-
ing the tiphiid and brachycistidid groups)
and in the hradynobaenid and scoliid
groups. Since this state is approached in
many other instances, its presence is not
a good indicator of higher groupings.

*45. Extent of joreiving venation. Prim-
itively, the longitudinal veins of the fore-
wing attain the apical margin of the wing
membrane. 45.1 — The venation of the
fore wing is reduced or retracted so that
the veins extend into the apical half of the
wing membrane but do not reach the
margin. 45.1.1 — The venation of the fore-
wing is extremely reduced so that it does
not extend beyond the basal half of the
wing membrane.

Venation w^hich attains the apical mar-
gin of the wing membrane is considered
primitive because this is the condition in
most Symphyta and many Terebrantia
(including Trigonalidae).

Retraction of the venation from the
apex of the wing membrane is extremely
common, and slate 45.1 has apparently
occurred on at least nine occasions, on
internodes 1-2, 12-14 and 18-21, and in
the apid, mutillid, anthoboscid (female),
sierolomorphid, rhopalcjsomatid and scoliid
groups. This state is thus of little utility
in determining relationships. The extreme
reduction of venation in relatively large
individuals without an associated tendency
toward brachyptery is apparently present
in only the apterogynid and hradynobae-
nid groups (on internode 23-24, state
45.1.1) and is thus considered quite good
evidence of their relationship. There are,
however, some members of the myzinid
group which show a somewhat similar
reduction in venation but with an accom-
panying decrease in wing or body size
(e.g., "Meria" injradentata; Myzinella
patrizii, Guiglia, 1968).

*46. Cells of forewing (Fig. 30). Primi-
tively, there are ten closed cells in the
forewing, viz., C, SC4-R4-S, SC4-R, R,
IS, 2S, S+M, IM, M+Cu, ICu. 46.1—
There are eight closed cells in the fore-
wing, viz., C, SC+R+S, SC+R, R, IS,
S+M, M+Cu, ICu. 46.1.1— There are
seven closed cells in the forewing, viz., C,
SC+R+S, SC+R, R, S+M, M+Cu, ICu.
46.1.1.1— There are six closed cells in the
forewing, viz., C, SC+R+S, SC+R, R,
S+M, M+Cu. 46.2 — There are seven
closed cells in the forewing, viz., C,
SC+R+S, (SC+R) + 1S, R, S+M, M+Cu,
ICu. 46.3 — There are nine closed cells in
the forewing, viz., C, SC+R+S, R,
(SC+R) + 1S, 2S, S+M, IM, M+Cu, ICu;
vein S is obliterated proximal to its fusion
with r-s. 46.4 — There are nine closed cells
in the forewing, viz., C, SC+R+S, R,
(SC+R) + 1S, 2S, S+M, IM, M+Cu, ICu;
vein S is obliterated just distal to its sepa-
ration from vein M. 46.5 — There are five
closed cells in the forewing, viz., C,
SC+R+S, SC+R, M+Cu, ICu. 46.5.1—

516

The University of Kansas Science Bulletin

31

47

47.1

47.1.1

Figs. 30-31. Characters of Aculeata. 30, forewing, showing primitive state of cells; veins (smaller letters) are:
C = costa, Cu = cubitus, E =: empusal, M = media, R ^ radius, S = sector, SC = subcosta (46 based on
Anthobosca, ^)\ 31, pterostigma, showing primitive and derived states of its size (47 based on Anthobosca,

2 ; 47.1 on Chimdamnus, 9 ; 47.1.1 on Aptcrogyini. $ ).

There are three closed cells in the fore-
wing, viz., C, SC+R+S, M+Cu.

Ten closed cells is considered the prim-
itive condition because this is the state in
the Trigonalidae and also in many aculeate
taxa which are considered relatively prim-
itive on the basis of other characters. The
trend has apparently been toward reduc-
tion in cell number in the Terebrantia,
and ten is the ma.ximal number of cells
found in the aculeates.

Losses of various veins and thus reduc-
tions in the number of cells have been
quite common although most appear to
be characteristic of single taxa only and
are thus of no use in establishing group-
ings. State 46.1 has arisen at least three
times, on internode 1-2 and in the rhopalo-
somatid and formicid groups, as well as
within many of the other groups. State
46.1.1 is present on internode 2-3, and
46.1.1.1 is characteristic of the scolebythid
group and also some species within the
bethylid group (e.g., Lytopsenella; Evans,
1964). State 46.2 appears in the sierolomor-

phids; 46.3 is developed in the methochid
group; 46.4 is present in the tiphiids.
State 46.5 has apparently arisen on inter-
node 23-24, linking the apterogynid and
bradynobaenid groups, with its derivative
(46.5.1) being present in the bradynobae-
nids; these last states are very character-
istic and considered good evidence of this
relationship.

47. Pterostigma! size (Fig. 31). Primi-
tively, the pterostigma is large and prom-
inent. 47.1 — The pterostigma is reduced
although nevertheless distinct and is mod-
erate to small in size. 47.1.1 — The ptero-
stigma appears as a mere swelling in the
venation, and is thus very small.

A large pterostigma is considered
primitive because this is the condition in
most non-aculeates and in many aculeates
which are considered relatively primitive
on the basis of other characters.

There appears to be a tendency towards
reduction in the size of the pterostigma
in various of the more highly developed

Phylogeny and Classification of the Aculeate Hymenoptera

517

groups. Thus, state 47.1 has apjxircntly
occurred on at least two occasions, on in-
ternodes 4-5 and 15-16. This trend has
hecn reversed, however, in the eotillid and
chyphotid groups, where the apparent
primitive state is present. Extreme reduc-
tion (47.1.1) is characteristic of the aptero-
gynid and bradynobaenid groups, and has
thus developed on internode 23-24; it also
occurs within the sphecid (e.g., Zyzzyx)
and apid (e.g.. Apis) groups. These states
are thus not considered to provide good
evidence of relationships.

48. Pterostigmal sclerotization. Primi-
tively, the pterostigma is heavily sclerotized
and uniformly thickened. 4S.1 — The scler-
otization of the pterostigma is reduced
so that the pterostigma appears to be
bounded by a distinct vein, although some
sclerotization of the cell so formed is gen-
erally retained.

A heavily sclerotized pterostigma is
considered primitive because this is the
condition in manv non-aculeates as well
as in most Aculeata.

Reduction of pterostigmal sclerotiza-
tion has occurred at least twice, on inter-
nodes 16-17 (associating the pompilid and
rhopalosomatid groups) and 19-20 (link-
ing the vespids and scoliids) as well as
within various other groups such as the
mutillids (e.g., Mutilla) and sphecids (e.g.,
Psetidoplisits). This character is thus only
of use in confirming the indications of re-
lationships provided by other characters.

49. Extent of hind wing venation. Prim-
itively, the longitudinal veins of the hind
wing attain the apical margin of the wing
membrane. 49.1 — The venation of the
hind wing is retracted or reduced so that
the veins extend into the apical half of
the wing membrane but do not attain the
margin. 49.1.1 — The venation of the hind
wing is greatly reduced and restricted to
the basal half of the wing membrane.

Venation which attains the apical mar-

gin of the wing membrane is considered
primitive because this is the condition in
most Symphyta and many Terebrantia
(including Trigonalidae).

Retraction of the veins away from the
apical margin of the hind wing has oc-
curred on numerous occasions, as with the
fore wing, although the pattern of deriva-
tions is not identical for both wings. State
49.1 is present on internodes 2-3, 12-14 and
18-21, and in the mutillid, sierolomorphid,
rhopalosomatid and scoliid groups, as well
as within various other groups. State 49.1.1
has occurred on internode 23-24 and in the
scolebythid group, as well as within the
bethylid group (e.g., Pristocera). This
character is thus of minimal utility in
determining relationships.

50. Cells of hind wing (Fig. 32). Primi-
tively, there are three closed cells in the
hind wing, viz., C, SC+R+S, M4-Cu.
50.1 — There are two closed cells in the
hind wing, viz., SC+R+S, M+Cu; vein
C is reduced distally. 50.2 — There are two
closed cells in the hind wing, viz., C,

Figs. 32-33. Characters of Aculeata. 32, hind wing,
showing primitive state of cells and veins; veins as
in forewing, plus A ^ anal, J =^ jugal bar, P =;
plical fold (50, 51 based on Piionyx, modified);
33, hind wing, showing derived state of crossvein
cu-e (52.1 based on Anthobosca, $ ).

518

The University of Kansas Science Bulletin

(SC+R+S) + (M+Cu); vein M+Cu is
obliterated. 50.3 — There is one closed cell
in the hind wing, viz., C; vein M+Cu
and all cross-veins are obliterated. 50.3.1
— There are no closed cells in the hind
wing.

Three closed cells is considered the
primitive condition because this is the
maximum number of cells found in the
Aculeata. In the Terebrantia the trend
has apparently been toward reduction in
cell number.

The distal reduction of the costal vein
(state 50.1) has occurred on numerous
occasions, on internodes 7-<S and 16-17, as
well as in the apid, methochid, sierolomor-
phid, formicid, eotillid and chyphotid
groups. This state is thus of no use as
an indicator of higher groupings. State
50.2 is uniquely present in the bradyno-
baenid group. State 50.3 is present in the
bethylid group, and its derivative (50.3.1)
is present in the scolebythids as well as
within the bethylid group (e.g., Pristo-
ceru), so that this modification is probably
significant in associating these groups.

*51. Hind wing anal and jiigal veins
(Fig. 32). Primitively, one anal vein and
a veinlike jugal bar are present in the
hind wing, i.e., veins A and J are present.
51.1 — An anal vein is present in the hind
wing but vein J is obliterated. 51.2 — Both
veins A and J are obliterated in the hind
wing.

The presence of both an anal vein and
a jugal bar is considered primitive because
this is the maximal number of such veins
found in the Aculeata. In the non-acu-
leates the trend has been toward loss of
these veins; they are generally both absent
in Terebrantia, including Trigonalidae,
but are present in the most primitive
Symphyta (e.g., Xyelidae). Although the
jugal "vein" is not a true vein, but rather
a development from the jugal bar (Hamil-
ton, 1972a), as shown by its position in
the jugal lobe, it is most convenient to

treat it in the same way as the true veins
for reference purposes.

Both the anal and jugal veins are re-
tained in only the sphecid group and the
anal vein only is retained in the apid
group, which is thus characterized by state
51.1. Both veins have apparently been lost
in all other groups, so that state 51.2 has
arisen on internodes 1-2 and 4-6, as well
as within the sphecid (most species, but
both veins present in Sphex, Prionyx,
e.g.) and apid (e.g., Megachile, Meso-
cheira) groups. Vein A has apparently
reappeared as a minute spur within the
anthoboscid (e.g., Anthobosca sp. male)
and thynnid (e.g., Heinithynnits) groups,
although its origin on vein E is shown as
a basal thickening of that vein in many
other groups. This character is thus only
useful in so far as it emphasizes the pres-
ence of a very primitive state in a taxon
(the sphecids) otherwise considered highly
derived, thus indicating a remote separa-
tion of the sphecid (and apid) line from
the other aculeates.

52. Hind wing cross-vein cti-e (Fig, 33).
Primitively, in the hind wing cross-vein
cu-e originates basal to the point of separa-
tion of veins M and Cu. 52.1 — Cross-vein
cu-e originates distal to the point of sepa-
ration of veins M and Cu.

A basal position for cross-vein cu-e is
considered primitive because this is the
condition in most taxa of Aculeata. In
taxa showing both states, state 52.1 gen-
erally occurs in those members considered
most derived on the basis of other char-
acters.

Although state 52.1 appears on the tree
only once (on internode 6-7) and thus
seems to provide good evidence of the rela-
tionship of nine mutilloid and tiphioid
taxa, its strength is considerably dimin-
ished by its presence within various other
taxa such as the sphecids (e.g., Gorytes)
and vespids (e.g., Synoeca). There have
also rarely been apparent reversals to the

Phylogeny and Classification of the Aculeate Hymenoptera

519

35

54.1

55.1.1

Figs. 34-37. Characters of Aculeata. 34, hind wing,
showing derived state of vein Cu (53.1 based on
Myrmosa, $)\ 35, anterior basal region of hind
wing, showing primitive and derived states of basal
hamuli (54 based on sphecid; 54.1 on pompilid);
36, hind wing, showing primitive and derived states
of pHcal lobe (all composite) ; 37, hind wing, show-
ing derived states of jugal lobe (all composite).

primitive state, within the myzinid and
tiphiid groups (some females, e.g., Myzi-
num, Tiphid).

*53. Hind wing vein Cii (Fig. 34). Prim-
itively, vein Cu is distinctly present in the
hind wing distal to its point of separation
from vein M. 53.1 — Vein Cu is obliterated
distal to its separation from vein M.

Presence of vein Cu distally is con-
sidered primitive because this is the con-
dition in most non-aculeate and in most
aculeate Hymenoptera.

The loss of the free section of vein Cu
has apparently occurred on a single occa-
sion, on internode 7-8, thus strongly asso-

ciating the sapygid, myrmosid and mutil-
lid groups. There is some confusion
regarding the mutillids, however. The
mutillid genus with the greatest propor-
tion of primitive character states (Pseu-
doplwtopsis) has a condition like state
53.1, but the more derived members often
seem to show the primitive state (e.g.,
Tric/iolabiodes). Such a direct reversal is
logically unlikely, since it involves the re-
gaining of a lost structure. The mutillids
commonly have the venation of the hind
wing considerably modified, however, var-
ious veins or sections of them being lost
or suppressed so that there are no closed
cells. Such modification could conceivably
have caused the apparent redevelopment
of a free spur in the position of vein Cu,
perhaps actually as a result of a break in
and displacement of vein E.

54. Basal hamuli (Fig. 35). Primitively,
the basal hamuli are dispersed along the
costal margin (on vein C) of the hind
wing distal to the point of separation of
veins C and SC+R+S. 54.1— The basal
hamuli are concentrated into a basal
cluster approximately at the point of sepa-
ration of veins C and SC4-R+S. 54.2 —
The basal hamuli are completely absent.

A dispersed series of basal hamuli is
considered primitive because this is the
condition in the Trigonalidae as well as
in those Aculeata judged to be most prim-
itive on the basis of other characters.
Lanham (1951) considered the Braconidae
(Terebrantia) to have a very primitive
condition, merely bearing a dispersed
series of strong, straight bristles. By con-
trast, Rasnitsyn (1969) showed both a
basal cluster and a dispersed series in his
putative generalized symphytan.

State 54.1 has occurred on two occa-
sions, apparently, on internodes 8-9 and
15-16. State 54.2 is present on internodes
18-19 and 23-24 and in the brachycistidid
group, as well as within others such as the
mutillid group (most species, except Pseu-

520

The University of Kansas Science Bulletin

dophotopsis) so that these states are use-
ful only as confirmatory indicators rather
than prime establishers of relationships.

55. Plical lobe (Fig. 36). Primitively, the
plical lobe is indicated by a moderate in-
cision of the hind margin of the hind
wing. 55.1— The plical lobe is indicated
by a shallow notch in the hind margin of
the wing. 55.1.1 — The plical lobe is not
indicated by any modification of the mar-
gin of the wing although the plical furrow
indicates its extent within the wing mem-
brane.

A plical lobe indicated by a moderate
incision is considered primitive because
this is the condition in those Aculeata
judged to be most primitive on the basis
of other characters. As Hamilton (1971)
has pointed out, the jugal and vannal
folds are alternative mechanisms of wing
folding and thus are mutually exclusive;
the presence of both jugal and vannal
lobes in the same wing is thus impossible.
In Hymenoptera there is no true vannal
fold, the anterior margin of the supposed
"vannal" lobe being formed by the plical
furrow. This lobe may thus conveniently
be called the "plical lobe" to maintain
morphological consistency.

Modification of the depth of the notch
indicating the plical lobe has apparently
occurred a number of times. Thus state
55.1 has arisen on at least three occasions,
on internodes 7-8 and 6-15 and in the
thynnid group; 55.1.1 has developed twice,
on internode 18-21 and in the scoliid
group. An apparently primitive state has,
however, been redeveloped in both the
rhopalosomatid and eotillid groups inde-
pendently, and the condition varies within
many of the taxa, so that this character is
not of fundamental value in indicating
groupings.

*56. ]iigal lobe (Fig. 37). Primitively, the
jugal lobe is long, being indicated by a
slight notch in the margin of the hind

wing. 56.1 — The jugal lobe is moderately
long and indicated by an axillary incision
which extends about half the anterior
length of the lobe. 56.1.1 — The jugal lobe
is small and indicated by a well-developed
axillary incision which extends almost to
the base of the hind wing. 56.2 — The
jugal lobe is absent.

A long jugal lobe indicated by a notch
is considered primitive because this is the
condition in most Symphyta and also in
the more generalized members of a few
taxa of Aculeata judged to be relatively
primitive on the basis of other characters
(especially of the wings). The Tere-
brantia have generally lost this lobe and
so do not provide any useful information
relative to this character.

The development of a marked incision
(56.1) has apparently occurred only once
on the tree, on internode 4-6, although it
is also present within the sphecid (e.g.,
Philanthiis) and apid (e.g., Megachile)
groups. State 56.1.1 has also arisen on the
tree only once, on internode 7-8, asso-
ciating the sapygids, myrmosids and mu-
tillids, although a similar tendency is
rarely present within the sphecids (e.g.,
Trypargihim). By contrast, complete ab-
sence of the jugal lobe (56.2) has occurred
at least three times, on internode 1-2 (as-
sociating the plumariids, bethylids and
scolebythids) and in the sierolomorphid
and typhoctid groups. This state is also
present within the mutillid group (all
members except Pseitdophotopsis). De-
spite the fact that these states are some-
what equivocal, this character seems to
provide useful information on groupings.

57. Leg form (Fig. 38). Primitively, the
legs are relatively slender and of general-
ized form, all three pairs being similar.
57.1 — The lesis have the femora and tibiae
expanded to some degree so that they are
rather stout, especially the middle and
hind pairs. 57.2— The legs have only the
femora much inflated. 57.3— The legs have

Phylogeny and Classification of the Aculeate Hymenoptera

521

Fig. 38. Leg, showing primitive and derived states
of leg form (57 based on Anthohosca, ^, hind leg;
57.1 on Anthohosca, 9. hind leg; 57.2 on Clysto-
pscncUa, $, hind leg: 57.3 on Olixon, 9. front leg).

the tarsi expanded and flattened and the
fore-tibiae swollen.

Slender, unspecialized legs are con-
sidered primitive because this is the condi-
tion in most non-aculeates as well as in
most Aculeata.

Modifications of the legs have occurred
numerous times, often being correlated
with burrowing or prey manipulation.
State 57.1 has arisen on at least four occa-
sions, always only in females, on inter-
nodes 7-10 and 21-23, and in the plumariid
and scoliid groups. The legs have reverted
to a slender condition in the methochid
group and within the myzinid (e.g.,
Pterombnts) and thynnid (e.g., Diamma)
groups, so that this state is additionally
weakened as an indicator of relationships.
State 57.2 is uniquely characteristic of
both sexes of the scolebythid group, and
57.3 is present in only the females of the
rhopalosomatid group (and is of some
importance here since it associates Rhopa-

losoma and Olixon, the latter genus hav-
ing been placed elsewhere on occasion —
e.g., in the Pompilidae, by Reid, 1939).

58. Aroliiim. Primitively, each arolium
forms a well-developed pad between the
tarsal claws. 58.1 — The arolium is com-
pletely reduced and absent, or at least not
distinguishable under a magnification of
lOOX^

A well-developed arolium is considered
primitive because this is the condition in
most non-aculeate and aculeate Hymenop-
tera as well as the Insecta in general.

The arolium has been lost on a number
of occasions, in the females of the myr-
mosid group, both sexes of the eotillids
and in the female on internode 21-23 and
the male on internode 23-24, so that the
chyphotids possess arolia in males only
(although much reduced in some males
in the subgenus Pitanta, e.g., Chyphotes
mojave; Mickel, 1967), and the apterogy-
nids and bradynobaenids lack them in
both sexes. Loss or extreme reduction of
the arolium has also occurred within the
apids (e.g., Megachile) at least. Because
of its multiple occurrences this state is
not a good indicator of groupings.

59. Claws (Fig. 39). Primitively, each
claw bears one or more processes on the
ventral margin, these forming teeth or
producing a cleft appearance; in either
case the claw is referred to as "toothed."
59.1 — Each claw is simple, with a smooth
ventral margin and no trace of any tooth.

Toothed claws are considered primi-
tive because this is the condition in many
non-aculeates (including Trigonalidae)
and in many Aculeata considered relatively
primitive on the basis of other characters.

The loss of teeth on the tarsal claws
has occurred on numerous occasions.
State 59.1 is characteristic of the entire
brachycistidid, scoliid and bradynobaenid
groups and the females of the plumariid
and myrmosid groups. It has also occurred
within the bethylids (e.g., C/irysis), mutil-

522

The University of Kansas Science Bulletin

Figs. 39-40. Characters of Aculeata. 39, claw, show-
ing primitive and derived states (59 based on
Anthohosca, $ , Apterogyna, i , and Sphcx, $ , left
to right then below; 59.1 on Bradynobaeniis, 9);
40, foretibial calcar, posterior view, showing primi-
tive and derived states (60 based on Anthohosca, ^ ;
60.1 on Apterogyna, 5", 60.1.1 on Bradynnbaeniis ,
9 ; 60.2 on Trielis, $. ) .

lids (all except Pseitdophotopsis), eotillids
(e.g., Eotilla). typhoctids (e.g., Typhoc-
toides), sphecids (e.g., Bembix) and apids
(e.g., Megachile) at least. The primitive
state also covers a variety of forms of teeth
which cannot readily be separated into dis-
tinct states, some of which may even be
secondary derivations (see the section on
the mutillids below for a discussion of
Rhopalomutilla). This character is thus
very weak and of essentially no signifi-
cance in determining relationships.

*60. Foretibial calcar (Fig. 40). Primi-
tively, the single calcar of the foretibia is
approximately straight with an elongate
inner lamina which may be somewhat
shortened. 60.1 — The foretibial calcar is
strongly curved inward and is more or
less even in width but with a small outer

spine present at the apex. 60.1.1 — The
foretibial calcar is strongly curved inward
and is more or less even in width with the
apex obtuse. 60.2 — The foretibial calcar
is inwardly curved, spatulate at the apex
and hollowed along the posterior surface.

An approximately straight calcar is
considered primitive because this is the
condition in most non-Aculeata as well as
in most aculeates.

There is actually greater variation in
the foretibial calcar than would appear
from the above. The primitive state covers
many rather subtle variations which could
not unequivocally be divided into separate
states. Further, 60.1, 60.1.1 and 60.2 are
actually quite similar and it is thus pos-
sible that they should really be considered
as expressions of the same series. Never-
theless, 60.1 (apterogynids) and its deriva-
tive, 60.1.1 (bradynobaenids) do serve to
associate the groups subtended by inter-
node 23-24, as has previously been noted
by Reid (1941). State 60.2 characterizes
the scoliid group.

61. Middle tibial spines (Fig. 41). Prim-
itively, each middle tibia bears many scat-
tered setae, a few of which are slightly
stronger than the remainder and are
termed "spiniform." 61.1 — Each middle
tibia bears scattered spiniform setae which
are readily identifiable as weak spines.
61.1.1 — Each middle tibia bears scattered
spines which are fairly strong and are
termed "moderate." 61.1.1.1 — Each middle
tibia bears moderate spines which are ab-
sent basally and present only apically.
61.1.1.1.1— Each middle tibia bears very
strong spines apically only. 61.1.2 — Each
middle tibia bears scattered spines which
are extremely strong and prominent. 61.2
— Each middle tibia is completely spine-
less and bears only a few weak setae.

An essentially spineless but setose mid-
dle tibia is considered primitive because
this is the condition in most non-aculeates.

The tibial spines are very variable

Phylogeny and Classimcation of the Aculeate Hymenoptera

523

*' *';' *l;_'-l 61'-'1 61.1.1.I.I 61.1.2 61.2

^h A 1

'.»

m

[ik

IR

vk

¥a

M

M

i'\*i'

I'H

i

¥

■< ' ' 64.1.2

64.1.2.1 64.1.2.1.1

Fi(,s. 41-44. Characters of Aculeata. 41, middle
tibia, showing primitive and derived states of spines
(61 based on Pristocera, c5 ; 61.1 on Plnmaiiiis, S ;

61.1.1 on Anoplius, 9; 61.1.1.1 on Paraponaa, $;
61.1.1.1.1 on Bradynobaentis, $ ; 61.1.2 on Trielis,
9; 61.2 on ClystopsencUa, 9); 42, middle tibial

spur, cross section and dorsal view, showing primi-
tive and derived states (64, 64.1, 64.1.1 composite;

64.1.2 based on Apterogyna, $ ; 64.1.2.1 on Brady-
nobaentis, $, hind spur; 64.1.2.1.1 on Bradynobae-
nus, 9 , hind spur) ; 43, middle or hind tibial spur,
showing derived states of calcaria (65.1, 68.3 based
on Paraponera, 9 ; 65.2, 68.4 on Methocha, 9 ; 68.1
on Pepsis, 9 ; 68.2 on Etiparagia, 9 ) ; 44, hind
coxa, showing primitive and derived states (66 based
on Anthobosca, $ \ 66.1 on Sierolomorp/ia, $ ;

66.1.1 on Myrmosa, $ ).

within and among taxa and the states
described are somewhat equivocal so that
assignment of the various states to the
different taxa is sometimes almost arbi-
trary. This character is thus of minimal
use in determining relationships. State
61.1 has apparently arisen on internode
1-4 and in the plumariid group, some de-
velopment of tibial spines being charac-

teristic of almost all aculeates. State 61.1.1
has developed on internodes 6-7 and 21-23,
and in the sphecid, pompilid and typhoctid
groups as well as female plumariids. The
spines have been somewhat reduced in the
methochid group, however, so that it is
assigned state 61.1. State 61.1.1.1 has de-
veloped in the formicid group and also
on internode 23-24, while its derivative,
61.1.1.1.1, is present in the bradynobaenids.
State 61.1.2 is present only in the scoliids.
Complete loss of spines (61.2) has taken
place only in the scolebythid group.

62. Hind tibial spines (Fig. 41). The
allocation of states for the hind tibial
spines is as for the middle tibia (charac-
ter 61).

The same comments apply to the hind
tibial spines as to those of the middle legs.
The allocation of states is similar except
that 62.1.1 is not present in female plu-
mariids and 62.1.1.1 is developed on inter-
node 21-23, being present in the chyphotid
group. Because of these dififerences, this
character is not considered to duplicate
the previous one.

63. Middle tibial spur number. Primi-
tively, each mid-tibia bears two spurs
apically. 63.1 — Each mid-tibia bears one
spur apically. 63.2 — Each mid-tibia bears
no spurs apically.

A middle tibia with two spurs is con-
sidered primitive because this is the con-
dition in most non-Aculeata and also in
most aculeates.

The loss of middle tibial spurs has oc-
curred on a number of occasions, each
time in a single taxon, so that this charac-
ter is of no use in establishing relation-
ships. State 63.1 is present in the apid,
methochid (female only), brachycistidid
and scoliid groups, and 63.2 has arisen in
the bradynobaenids. One spur has also
been lost within various other taxa such
as the bethylid (e.g., Deinodryinus), mu-
tillid (e.g., Acanthophotopsis) and sphecid

524

The University of Kansas Science Bulletin

(e.g., Larra, Zyzzyx) groups, and both
spurs have been lost within the formicid
group (e.g., Crematogaster).

64. Basic form of middle and hind tibial
spurs (Fig. 42). Primitively, each spur is
simple and approximately circular in cross-
section, forming a very narrow cone. 64.1
— Each tibial spur is slightly flattened
dorsally but retains simple margins. 64.1.1
— Each tibial spur is dorsally flattened and
has serrate margins. 64.1.2 — Each tibial
spur is dorsally flattened and with deeply
dentate margins. 64.1.2.1 — Each tibial spur
is dorsally flattened and somewhat elon-
gated, with a few teeth on the margins.
64.1.2.1.1 — Each tibial spur is dorsally flat-
tened and threadlike (nematiform) with
simple margins.

Conical tibial spurs are considered
primitive because this is the condition in
many non-Aculeata, particularly Tere-
brantia, and in those aculeates considered
relatively primitive on the basis of other
characters.

Although the modifications of spur
form have occurred on various occasions,
there does seem to be a logical progres-
sion in the pattern shown in at least the
chyphotid, apterogynid and bradynobaenid
groups which serves to associate them.
State 64.1 is present in females only on
internodes 6-7 and 21-23, and in both sexes
of the pompilid and scoliid groups. 64.1.1
has arisen in females on internode 7-8 and
in both sexes on internode 4-5. State
64.1.2 is present in both sexes on internode
23-24, and its modification (64.1.2.1) oc-
curs in the males of bradynobaenids while
the females of that taxon are even more
modified and exhibit state 64.1.2.1.1.

65. Middle tibial calcar (Fig. 43). Primi-
tively, the mid-tibial spurs are similar,
showing no signs of modification as cal-
caria. 65.1 — The inner spur of the mid-
tibia is modified as a calcar by formation
of a dorsal pectinate carina. 65.2 — The

(inner?) spur of the mid-tibia is modified
as a calcar by the development of dorsal
pectination but without carina formation.

Unmodified spurs are considered prim-
itive because this is the condition in most
Hymenoptera, both non-aculeates and
aculeates.

The formation of a calcar on the mid-
dle tibia is uncommon and has occurred
twice, each time in a slightly different
fashion. State 65.2 is present only in the
females of the methochid group and 65.1
occurs in both sexes of the formicids, al-
though the modification is minimal in
some (e.g., Cawponotus). This character
is therefore of no use in establishing
higher groups.

66. Form of bind coxa (Fig. 44). Primi-
tively, each hind coxa is smoothly rounded
dorsally. 66.1 — Each hind coxa bears a
definite longitudinal carina along the dor-
sal surface, especially basally. 66.1.1 — Each
hind coxa bears a dorsal carina which is
expanded and lamellate.

A simple hind coxa is considered prim-
itive because this is the condition in most
non-Aculeata (although the coxa is cari-
nate in Trigonalidae) and in those acu-
leates judged to be the most primitive on
the basis of other characters.

The development of a dorsal carina or
tooth on the hind coxa has occurred on
various occasions, sometimes rather equiv-
ocally. Such projections have also been
lost again within some groups such as the
mutillids (see section below for discus-
sion) and brachycistidids (present in
males of most species but absent in
Brachycistis). This character is thus of
little importance in establishing groupings.
State 66.1 has developed on internodes 7-8,
12-14 and 21-22, in the thynnid, sierolo-
morphid and apterogynid groups, and in
the males of the methochid and chyphotid
groups. The expansion of the tooth into
a lamella (66.1.1) is characteristic of the
myrmosid group only, although it may

PlIYLOGENY AND ClASSII-ICATION OI- THE AcULEATE HyMENOPTERA

525

be approached within the brachycistidid
group {Colocistis male).

67. Hind tibial spar uiimbcr. Primitively,
each hind tibia bears two s[nirs apically.
67.1 — Each hind tibia bears a single spur
apically.

A hind tibia with two spurs is con-
sidered primitive because this is the con-
dition in most Hvmcnoptera. both non-
aculeates and aculeates.

The loss of one of the hind tibial spurs
has occurred on the tree only in the
methochid group, thus not providing any
information on higher relationships. A
similar state is, however, found within the
bethylid group (e.g., cleptid sp.) and a
more extreme state, involving the loss of
both spurs, has occurred within the apid
(e.g., Apis) and formicid (e.g., Creinato-
gaster) groups.

*68. Hind tibial calcar (Fig. 43). Primi-
tively, the hind tibial spurs are similar,
showing no evidence of modification as
calcaria. 68.1 — The inner spur of the hind
tibia is modified as a calcar by the forma-
tion of a basal dorsal tuft of bristles with
little modification of the cuticular portion
of the spur. 68.2— The inner spur of the
hind tibia is modified as a calcar by a
dorsal carinate expansion of the cuticle
over a considerable length. 68.3 — The in-
ner spur of the hind tibia is modified as a
calcar by pectinate elaboration of a dorsal
carina. 68.4 — The inner spur of the hind
tibia is modified as a calcar by develop-
ment of dorsal pectination but without
carina formation.

Unmodified spurs are considered prim-
itive because this is the condition in non-
aculeates as well as in most Aculeata.

The formation of a calcar on the hind
tibia has occurred at least four times but
in different ways on each occasion. State
68.1 has developed on internode 16-17, thus
associating the pompilid and rhopaloso-
matid groups quite strongly, as has been

noted by Rick (1970). States 68.2, 68.3
and 68.4 are characteristic of the vespid,
formicid and methochid groups respec-
tively (the last two are similar to the
condition of the middle tibial spurs in
these taxa). Different forms of hind tibial
calcaria have also been developed within
the sphecid (present in most but absent
in a few such as Astata) and myzinid
(e.g., Myzininn female) groups. Although
some members of the apid group (e.g.,
Halictits) have the inner hind tibial spur
with large teeth, this state does not appear
to be comparable to calcar formation in
the other groups.

*69. Modified mesosoma of apterous je-
male (Fig. 45). Primitively, the mesosoma
of the female is essentially unmodified and
similar to that of the male. 69.1 — The
female is apterous with the mesosoma
modified so that there are movable sutures
between the pro- and meso- and the meso-
and metathorax, the mesonotal subdivi-
sions are distinguishable, the prepectal
sclerite is free and the mesepimeron is
distinct. 69.1.1 — The female has the meso-
soma modified as in 69.1 except that the
mesepimeron is not distinguishable exter-
nally. 69.1.1.1 — The female has the meso-
soma modified as in 69.1.1 except that the
mesonotal subdivisions are not distinguish-
able and the prepectal sclerite is reduced
although free. 69.2 — -The female is ap-
terous with the mesosoma modified so that
the pleura are flattened, there is a func-
tional suture between the pro- and meso-
thorax, the metathoracic-propodeal suture
is distinct but not functional dorsally, and
the prepectal sclerite is fused to the mes-
episternum. 69.2.1 — The female has the
mesosoma modified as in 69.2 except that
the metathoracic-propodeal suture is oblit-
erated dorsally. 69.2.2 — The female has the
mesosoma modified as in 69.2 except that
the suture between the pronotum and
mesothorax is non-functional but distinct,
and the metathoracic-propodeal suture is

526

The University of Kansas Science Bulletin

69.4

Fig. 45. Mcsosoma, lateral view, showint; derived states in apterous females (69.1 based nn Zaspilothynntis;
69.1.1 on Methocha; 69.1.1.1 on Bniesiella: 60.2.1 on Myrmosa; 69.2.2 on Pscudophotopsis; 69.3.1 on

Typhoctoides; 69.3.2 on Apterogyna; 69.4 on Phtmariiis) .

Phylogeny and Classification of the Aculeate Hymenoptera

527

indistinct dorsally. 69.3 — The female is
apterous with the mesosoma modified so
that the pleura (especially the mesopleu-
ron) are somewhat protuberant, the suture
between the pronotum and the mesothorax
is functional, the meso-metathoracic su-
ture is visible but non-functional, the mes-
onotum is neither reduced nor enlarged,
and the prepectal sclerite is fused to the
mesepisternum. 69.3.1 — The female has
the mesosoma modified as in 69.3 except
that the mesonotum is very short and
transverse. 69.3.2 — The female has the
mesosoma modified as in 69.3 except that
the meso-metathoracic suture is onlv in-
distinctly visible and the mesonotum is
somewhat posteriorly produced. 69.4 — The
female is apterous with the mesosoma
modified so that the pro-meso- and meso-
metathoracic articulations are retained, the
propleura are fused to form a rigid tube,
and the meso- and metathoraces are sepa-
rated by a deep lateral and ventral con-
striction, causing the metacoxal cavities to
be considerably separated from the meso-
coxal cavities.

An unmodified mesosoma in the fe-
male is considered primitive because this
is the condition in most Hymenoptera,
both non-aculeates and aculeates.

Associated with aptery in the females
(state 2.1), the mesosoma has been vari-
ously modified, as was shown by Reid
(1941). (Although worker ants are prim-
itively apterous and thus show mesosomal
modifications, they are not included here
because the queens are winged with no
such mesosomal changes, except in a few
highly specialized forms. The primitive
condition in the formicids is thus not con-
sidered to involve this character.) There
are at least four different basic patterns
along which such modification has oc-
curred, three of them linking various taxa
into higher groupings. Because of the
complexity and extent of the various
modifications involved, each line of de-

velopment is considered to provide very
good evidence of the relationships demon-
strated.

The line based on 69.1 associates vari-
ous members of the tiphioid complex. In
this line it is apparent that each of the
derived states is expressed only when ac-
companied by state 2.1, but that the po-
tential for such expression exists even in
the absence of winglessness in the female.
Although state 69.1 is displayed in the
thynnid group only, it has logically arisen
on internode 10-11 since its derivatives are
expressed in other taxa derived from that
internode. State 69.1.1 is expressed in the
entire methochid group and also in some
members of the myzinid group (e.g.,
Braiinso7neria), so that it must have arisen
below node 13. In addition, its derivative
state (69.1.1.1) is present in the brachy-
cistidid group, so that state 69.1.1 is log-
ically placed on internode 11-12. The line
based on state 69.2 associates the myrmosid
and mutillid groups. State 69.2 is logically
derived on internode 8-9 and has appar-
ently subsequently been modified in dif-
ferent directions in its two derivative
groups, so that 69.2.1 appears in the myr-
mosids and 69.2.2 in the mutillids. The
line based on 69.3 associates five taxa in
the typhoctoid and bradynobaenoid lines,
state 69.3 having been derived on inter-
node 18-21. Its two derivative states are
expressed in difTerent lines; 69.3.1 appears
on internode 21-22, strongly associating
the typhoctid and eotillid groups, and
69.3.2 is placed on internode 21-23, strongly
linking the chyphotid, apterogynid and
bradynobaenid groups. State 69.4 has oc-
curred in only the plumariid group. Apart
from these states, aptery and various meso-
somal modifications have occurred within
certain other taxa, as is indicated in the
discussion of character 2.

*70. "Felt lines" (Fig. 46). Primitively,
there are no modifications of the second
metasomal segment in the form of differ-

528

The University of Kansas Science Bulletin

47

71.2

Figs. 46-47. Characters of Aculeata. 46, second
metasomal segment, lateral view, showing derived
states of "felt lines" (70.1 based on Pseudophotopsis,
$ ■ 70.2 on Typhoctes, S ; 70.2.1 on Apterogyna,
9, and Bnidyiiobaeiuis. 9, left to right); 47, second
to fourth metasomal segments, dorsal view, somewhat
expanded, showing derived states of stridulitra
(both composite).

entiated areas for the openings of sub-
cuticular glandular organs. 70.1 — There
are longitudinally elongate and narrow
depressed areas ("felt lines") on each side
of the second metasomal segment on both
tergum and sternum, each bearing dense
short pubescence and the openings of
secretory organs. 70.2 — There is a longi-
tudinally oval area ("felt line") developed
on each side of the second metasomal seg-
ment on the tergum only, this bearing
dense moderately long pubescence and the
openings of secretory organs. 70.2.1 —
There is a longitudinal area on each side
of the second metasomal tergum consist-
ing of a line of cuticular depression or in-
vagination but without dense pubescence.
A second metasomal segment without
"felt lines" is considered primitive because

this is the condition in most Hymenop-
tera, both non-aculeates and aculeates.

Although it would at first seem likely
that a structure as complex as a "felt line"
would most probably have a single deriva-
tion, this is apparently not so. Despite the
similarities between states 70.1 and 70.2,
these do exhibit slight morphological dif-
ferences, as was found by Debolt (1973).
In addition, similar structures, in gross
appearance at least, are present on the
first metasomal tergum within the tiphiid
group {Paratiphia) and ventrally on the
pronotum within the mutillids (Pseiido-
pJwtopsis). Some primitive members of
the apid group (e.g., Bmchyhesma, Serico-
gaster) bear a lateral "fovea" on the sec-
ond metasomal tergum, which is appar-
ently secretory although not as densely
pubescent as in state 70.2; this is probably
analogous (or even homologous) to a "felt
line." Structures which are probably
analogous are also found on the sixth
metasomal sternum in some vespids (e.g.,
Paravespiila; Spradbery, 1973).

State 70.1 is characteristic of the mutil-
lid group only, within which it has under-
gone various modifications involving loss
of the sternal or tergal elements or both.
State 70.2 has occurred on internode 18-21
and links five taxa. Loss of the pubescence
in the "felt line" resulting in a mere cu-
ticular depression or invagination (70.2.1)
has apparently taken place in the female
on internode 23-24 so that male apterogy-
nids have pubescent lines whereas the fe-
male has cuticular grooves. In the brady-
nobaenids the loss of pubescence has
occurred in the male also and both sexes
have lines along which the cuticle is
apparently slightly invaginated. These
structures have not previously been re-
corded in this group, and their presence
is strong evidence of the relationship be-
tween the bradynobaenid, apterogynid and
chyphotid groups and of these to the
typhcKtids and eotillids.

Phyloceny and Classification of thf Aculeate Hymenoptera

529

*71. Stridiilitru (Fig. 47). Primitively,
there are no specialized stridulatory organs
on the metasoma. 71.1 — There is a single
striclulitrum developed as a small shield-
shajKtl, liiich' transversely striated area
mesalK at the base of the third metasomal
terguni. This rubs against a plectrum
(ridge) on the underside of the posterior
margin of the second tergum. 71.2 — There
is a pair of stridulitra developed as a fairly
narrow, finely transversely striated, shield-
shaped area on each side at the base of the
fourth metasomal tergum. This rubs
against a plectrum on the underside of the
posterior margin of the third metasomal
tergum on each side.

The absence of metasomal stridulitra
is considered primitive because this is the
condition in most Hymenoptera, both
non-aculeates and Aculeata. The termi-
nologv is derived from that of Ashlock
k Lattin (1963), although the "instru-
ment" is moved against the plectrum in
the Hymenoptera rather than the plectrum
being the movable part. The fine struc-
ture of the mesal stridulitrum has been
investigated by Hinton, Gibbs & Silber-
glied (1%9). '

The development of a mesal stridu-
litrum (71.1) has occurred on internode
S-9 and thus strongly associates the myr-
mosid and mutillid groups. The strength
of this state is, however, slightly reduced
by the fact that a similar development has
taken place within the formicid group
(present in the poneroid and pseudomyr-
mecoid complexes but absent in the myr-
mecioid line; see Haskins & Ensman,
1938; Markl, 1973) and within the rho-
palosomatids (e.g., Olixon), where the
form is somewhat different, however. Lat-
eral stridulitra (71.2) are present in the
typhoctid, eotillid and chyphotid groups
and apparently arose on internode 18-21;
they were subsequently lost on internode
23-24, being absent in the apterogynids
and bradynobaenids. It seems more likely

that such a structure, apparently unique
in the aculeates (although a very few
members of the sphecid group, e.g., Pseiid-
oplisus. possess lateral stridulitra on the
fourth and fifth terga) should have arisen
once only, than that it had two inde-
pendent origins resulting in identical ex-
pressions. Similar loss of stridulitra has
apparently occurred in the Ligyrocoris
group of Lygaeidae (Hemiptera) (Ash-
lock & Lattin, 1963), so that this possibility
should not be rejected out of hand.

72. Constriction of first metasomal tergum
(Fig. 48). Primitively, especially as viewed
from above, the first metasomal tergum
overlies the second and its surface contour
merges smoothly with it in a continuous
arc. 72.1 — The posterior margin of the
first metasomal tergum is slightly con-
stricted so that there is a definite discon-
tinuity between the contours of the first
and second terga, but the first segment is
not nodose. 72.1.1 — The posterior margin
of the first metasomal tergum is markedly
constricted and differentiated from the
second so that the first segment is def-
initely nodose.

An unconstricted first tergum is con-
sidered primitive because this is the condi-
tion in most Symphyta and many Tere-
brantia as well as in those Aculeata
considered primitive on the basis of other
characters.

Slight posterior constriction of the first
metasomal tergum has apparently occurred
on numerous occasions so that this state
is of minimal use in establishing groups.
State 72.1 has arisen on at least seven
occasions, on internodes 8-9 and 11-12 in
males only, 12-14 in females also, and
18-21 in both sexes, and in addition in
the rhopalosomatid, formicid and scoliid
groups. It is also present within various
other taxa such as the sphecids (e.g., Try-
par gihun), apids (e.g., Osiris female),
sierolomorphids (e.g., Sierolomorpha can-
adensis) and vespids (e.g., Monobia).

530

The University of Kansas Science Bulletin

Figs. 48-52. Characters of Aculeata. 48, base of
metasoma, dorsal view, showing primitive and de-
rived states of constriction of first tergum (72 based
on Anthobosca, ^ ; 72.1 on Typhoctes, 9 , modified;
72.1.1 on Apterogyiia, $); 49, base of metasoma,
dorsal view, showing derived state of petiole (73.1
based on Brady nobaentis, 5); 50, first metasomal
segment, lateral view, showing primitive and derived
states of lateral margin of tergum (74 based on
Sierolomorpha, 9; 74.1 on Typhoctes, 9; 74.1.1 on
Chypbotes, $,)\ 51, first metasomal segment, lateral
view, showing derived state of width of tergum
(75.1.1 based on Chyphotes, 9; Fig. 50, 74.1.1
represents 75.1 also); 52, base of metasoma, lateral
view, showing primitive and derived states of differ-
entiation of first sternum (76 based on Anthobosca,
(5 ; 76.1 on Tiphia, $ , modified).

State 72.1.1 occurs only in the apterogynid
group and also within the formicid group
(all members except Amblyoponini; Wil-
son, Carpenter & Brown, 1967).

73. Metasofnal petiole (Fig. 49). Primi-
tively, the first metasomal segment is

evenly constricted anteriorly toward its
articulation with the mesosoma. 73.1 —
The anterior extremity of the first meta-
somal segment is highly differentiated
from the main body of the segment as a
short petiole which is more or less cy-
lindrical.

A non-petiolate first metasomal seg-
ment is considered primitive because this
is the condition in the Symphyta and most
other non-aculeates as well as in most
Aculeata.

The formation of a definite petiole has
occurred on internode 18-21 where a short,
cylindrical petiole is characteristic of five
taxa, and also in the formicid group. The
strength of this character is, however,
weakened by the fact that various petiolar
modifications of the first metasomal seg-
ment have occurred within many other
taxa such as the sphecids (e.g., Ammo-
phila), myzinids (e.g., Meria, Mesa) and
vespids (e.g., Belonogaster, Eiimenes).

*74. Lateral margin of first metasomal
tergum (Fig. 50). Primitively, the first
metasomal tergum broadly overlies the
sternum laterally and is freely movable
against it. 74.1 — The first metasomal ter-
gum overlies the sternum only posteriorly
and narrowly, being closely associated
with it but not fused except perhaps at
the extreme base. 74.1.1 — The first meta-
somal tergum overlies the sternum only
posteriorly and narrowly and is fused to
it along the petiole.

A first tergum that broadly overlies
the sternum is considered to be primitive
because this is the condition in most non-
aculeates and in most members of the
Aculeata.

The tendency toward fusion of the
first tergum and sternum has apparently
been established on internode 18-21 since
state 74.1 associates the typhoctid and eotil-
lid groups, and its derivative (74.1.1) has
arisen on internode 21-23, linking the chy-
photid, apterogynid and bradynobaenid

Phylogeny and Classification of the Aculeate Hymenoptera

531

groups. This trend is almost unique and
good evidence of these relationships al-
though a rather similar condition is shown
within the myzinid group (e.g., Meria).
DifTerent forms of iusion have occurred
within other taxa such as the sphecid (e.g.,
Chlorion) and vcspid (e.g., Zetluis)
groups.

*75. Width of first tergum (Fig. 51).
Primitively, the first metasomal tergum is
hroader than the sternum over its entire
length due to its overlap of the sternum.
75.1 — The first tergum is very narrow an-
teriorly, its lateral margin running along
the dorsal surface of the petiole on each
side. 75.1.1 — The first tergum is com-
pletely absent anteriorly, the sternum ap-
parently forming the petiole in its entirety.

A broad first tergum is considered
primitive because this is the condition in
most non-aculeates and in most members
of the Aculeata.

Extreme reduction in width of the first
tergum anteriorly (75.1) is characteristic
only of the chyphotid, apterogynid and
bradynobaenid groups and has occurred
on internode 21-23. Complete absence of
the tergum along the petiole (75.1.1) has
apparently occurred only in the females
of the chyphotid group, traces of the ter-
gum being distinguishable to the base of
the petiole in the males of that taxon. A
state similar to 75.1 has, however, occurred
within the myzinid group (e.g., Meria)
and the absence of the tergum along the
petiole has taken place within the sphecid
group (e.g., Prionyx), although in a
slightly different fashion from state 75.1.1.
Despite these separate origins, this char-
acter prcjvides cjuite good evidence of the
relationship of the taxa subtended by in-
ternode 21-23.

76. Differentiation of first sternum (Fig.
52). Primitively, the first metasomal ster-
num overlies the second posteriorly and is
not differentiated from it by anv marked

discontinuities. 76.1 — The first sternum is
depressed and differentiated from the sec-
ond by a deep constriction.

An undifferentiated first sternum is
considered primitive because this is the
condition in most non-aculeates and also
in those members of the Aculeata con-
sidered to be primitive on the basis of
other characters.

Posterior differentiation of the first
sternum has apparently occurred at least
three times, on internodes 8-9, 11-12 and
16-18 and also within the thynnid (e.g.,
Dianuna female) and sphecid (e.g., Cer-
ceris) groups at least, so that it is of little
importance in establishing relationships,
althoutrh there are differences in detail in
the various occurrences.

77. Constriction of second metasomal seg-
ment. Primitively, the second metasomal
segment is not at all constricted apically
but its contours merge evenly with those
of the third. 77.1 — The second metasomal
segment is strongly constricted apically,
dorsally and ventrally as well as laterally,
and is strongly differentiated from the
third.

An unconstricted second metasomal
segment is considered primitive because
this is the condition in most Hymenoptera,
both non-aculeates and aculeates.

A nodose second metasomal segment
is characteristic only of the apterogynid
group and is thus not of any importance
in determining higher groupings, A simi-
lar state is also present within the formicid
group (e.g., Atta).

*78. Seventh tergum of female (Fig. 53).
Primitively, the seventh metasomal ter-
gum (eighth abdominal tergum) is some-
what exposed. It is evenly sclerotized over
its entire length, forming a fairly long
transverse sclerite. 78.1 — The seventh met-
asomal tergum is retracted and entirely
hidden. Its sclerotization is reduced to a
short strip across the anterior margin con-

532

The University of Kansas Science Bulletij

Figs. 53-55. Characters of Aculeata. 53, seventh
tergum of female, dorsal view, showing primitive
and derived states (all composite) ; 54, gonocoxite IX
of female, lateral view, showing primitive and derived
states (both composite); 55, gonapophysis VIII of
female, lateral view, showing primitive and derived
states (both composite).

necting the lateral plates which bear the
spiracles. 78.1.1 — The seventh metasomal
tergum is retracted and entirely hidden.
Its sclerotization is entirely reduced
mesally so that the lateral spiracular plates
are linked by membrane only.

An exposed and fully sclerotized
seventh tergum is considered primitive be-
cause this is the condition in the non-
aculeate Hymenoptera.

Concealment and partial desclerotiza-
tion of the seventh tergum in the female
is characteristic of the entire group com-
prising the Aculeata (sensii stricto) as de-
fined by Oeser (1961). Since this state
(78.1; internode 1-4) has been attained
only once in the entire Hymenoptera, it
provides extremely good evidence of the
holophyletic nature of this grouping. Al-
though the seventh tergum is concealed

in the plumariids, this has occurred by
enlargement of the sixth sternum which
enfolds the seventh tergum, and not by
anterior retraction of the seventh tergum
under the sixth. State 78.1.1 is uniquely
characteristic of the apid group and so
does not serve to indicate any further
higher groups.

*79. Gonocoxite IX of female (Fig. 54).
Primitively, section 1 of gonocoxite IX
(Smith, 1970a) is dorsoventrally con-
stricted but is not completely divided. 79.1
— Section 1 of gonocoxite IX is dorsoven-
trally constricted and completely divided
by an internal articulation.

The absence of an articulation within
section 1 of the gonocoxite is considered
primitive because this is the condition in
all non-aculeate Hymenoptera as well as
in most Aculeata.

The development of an articulation
within section 1 of the gonocoxite is
uniquely characteristic of the bethylid,
scolebythid and plumariid groups (it even
occurs in the highly modified genitalia of
the Chrysididae; Oeser, 1961). This is
thus a very strong character uniting these
taxa into a holophyletic group, and has
arisen on internode 1-2.

80. Gonapophysis VIII of female (Fig.
55). Primitively, each gonapophysis VIII
bears a lamellate valve dorsally near the
base. 80.1 — Each gonapophysis VIII is
simple and smooth dorsally with no trace
of a valve.

A valve on the gonapophysis VIII is
considered primitive because this is the
condition in most Aculeata including
those which are judged to be most primi-
tive on the basis of other characters. A
valve is also present in at least some non-
aculeates (e.g., Ichneumonidae; Oeser,
1961).

The loss of the gonapophyseal valve
has occurred on at least seven occasions,
on internode 21-22 (in the typhoctids and

Phyloghny and Classification of the Aculeate Hymenoptera

533

most likely the colillid i;i'oup althoUi;h the
actual condition in the eotillids is un-
known) and in the plimiariid, sapygid,
I pompilid, vespid. chyphotid and bradyno-
haenid groups. Alternatively the valve
could have been lost on internode 18-21
and then been redeveloped in the aptero-
gynids. Although this would be more
parsimonious placement on the cladogram,
it seems intuitively less likely than mul-
tiple losses. In view of its large number
of independent derivations, this state is
not of any use in establishing higher
groupings.

81. Gonupophysis IX of female (Fig. 56).
Primitively, the gonapophysis IX (fused
to its counterpart) is gently arcuate dor-
sally so that the apex is directed obliquely
ventrally. 81.1 — The gonapophysis IX is

81.2

82.1.)

Figs. 56-57. Characters of Aculeata. 56, gonapophy-
sis IX of female, showing primitive and derived
states (all composite): 57, posterior region of meta-
soma of male, ventral view, showing primitive and
derived states of seventh sternum, sterna numbered

(all composite).

Strongly curved dorsally, especially basally,
being somewhat elongated and with the
apex directed downward. 81.2 — The gona-
pophysis IX is almost straight or even
slightly ventrally arcuate so that the apex
is directed slightly upward or directly pos-
teriorly.

A gently arcuate gonapophysis IX with
the apex oblitjuely ventrally-directed is
considered primitive because this is the
condition in many Aculeata which are
judged to be relatively primitive on the
basis of other characters.

The degree and type of curvature of
the gonapophysis IX is difficult to assign
unequivocally to one of the designated
states, and even when such decisions have
been made, both derived states have ap-
parently originated on numerous occa-
sions. Nevertheless, the states do give
some general indications of relationship
and are somewhat useful. State 81.1 has
apparently been derived on internodes 8-9
and 12-14 and in the methochid group.
These taxa are all on the branch subtended
by internode 6-7, the other members of
which also show tendencies toward posses-
sion of this state, especially in elongation
of the gonapophysis. State 81.1 is also
present in the bradynobaenid group. State
81.2 has apparently been derived on inter-
nodes 4-5 and 18-21 (modified to 81.1 in
the bradynobaenids) and in the rhopalo-
somatids and formicids. These are mostly
groups on the branch subtended by inter-
node 6-15, the other members of which
also show a tendency toward a less elon-
gate and less flexible gonapophysis.

82. Seventh sternum of male (Fig. 57).
Primitively, the seventh metasomal ster-
num (eighth abdominal sternum) is well-
developed, not much smaller than the
sixth and clearly visible and exposed. 82.1
— The seventh metasomal sternum is re-
duced and much smaller than the sixth
although it is partly exposed. 82.1.1 — The
seventh metasomal sternum is greatly re-

534

The University of Kansas Science Bulletin

duced, much smaller than the sixth and
completely hidden by it.

A well-developed seventh sternum is
considered primitive because this is the
condition in most Hymenoptera, both
non-aculeates and aculeates.

Reduction of the seventh sternum has
apparently occurred on numerous occa-
sions. State 82.1 has arisen on internodes
6-7, 16-17 and 18-21, as well as in the
plumariid and vespid groups. Further-
more, re-enlargement of the seventh ster-
num has apparently occurred on internode
12-13 and in the myrmosid group. State
82.1.1 has been derived on internode 12-14
and in the apid group. As in all charac-
ters involving relative sizes, it is difficult to
assign the various states unequivocally, and
it appears that because of multiple deriva-
tions this character cannot be used to
provide reliable information on higher
groupings,

*83. Form of wale hypopygium (Fig. 58).
Primitively, the hypopygium (eighth met-
asomal or ninth abdominal sternum) is
simple, unmodified and apically rounded
although it may be reduced in size and
concealed. 83.1 — The hypopygium is much
narrowed, forming a peglike structure,
but is not acute apically. 83.2 — The hypo-
pygium bears three subequal spines
apically, the spines about equal to the base
in length (excluding the anterior process).
83.3 — The hypopygium is modified to
form a single upcurved spine or aculeus.
83.4 — The hypopygium is apically pro-
duced as three spines, the middle one up-
curved and much longer than the laterals
which are considerably shorter than the
base (excluding the anterior process).
83.4.1 — The hypopygium is modified as in
83.4 except that the middle spine is only
very slightly longer than the laterals and
not upcurved.

A simple hypopygium is considered
primitive because this is the condition in

83.1

Fig. 58. Hypopygium of male, dorsal view and
media! longitudinal section (approximate), showing
derived states (83.1 based on Sierolotnorpha; 83.2
on Trielis; 83.3 on Paratiphia; 83.4 on Apterogyna;
83.4.1 on Brady nobaentis).

most Hymenoptera, both non-Aculeata
and Aculeata.

Each modification of the hypopygium
has apparently only occurred once and
where any one modification is present in
more than one taxon it thus provides very
good evidence of relationships. State 83.1
is characteristic of the sierolomorphids
only and 83.2 of the scoliids. State 83.3
has been derived on internode 11-12 and
associates four taxa in the tiphioid group,
although this condition is approached
within the thynnids (Salter, 1957). State
83.4 has originated on internode 21-23 and

PnYLOGliNY AND CLASSIFICATION OF THE AcULEATE HyMENOPTERA

535

links three taxa in ihe bradynobaenoid
group; the derived form of this state
((S3.4.1) is present in the bradynobaenid
group only. States 83.2 and (S3.4.1 are
superficially similar but apparently have
different origins and do appear somewhat
different, especially when considered in
conjunction with the following character
(84). Within some of the other groups
the hypopygium has often been modified
in various ways but apparently never
like any of the above states (except per-
haps for 83.1).

84. Concealment of jriale hypopygium.
Primitively, the hypopygium (eighth met-
asomal or ninth abdominal sternum) is
not concealed by the preceding sterna.
84.1^The hypopygium is partially (up to
the basal half) concealed by the preceding
sterna. 84.2 — The hypopygium is com-
pletely or almost completely concealed by
the preceding sterna.

An unconcealed hypopygium is con-
sidered primitive because this is the con-
dition in most Hymenoptera, both non-
aculeates and aculeates.

Partial or complete concealment of the
hypopygium has apparently occurred on
various occasions, mostly in single taxa,
and is thus of minimal use in establishing
relationships. State 84.1 has arisen in the
anthoboscid and chyphotid groups and
state 84.2 is present on internode 12-14
and in the apid, myrmosid, sierolomorphid
and scoliid groups.

85. Cerciis of triale. Primitively, a cercus
is present on each side of the composite
tenth and eleventh abdominal tergum of
the male. 85.1 — The cercus is absent.

Presence of a cercus in the male is con-
sidered primitive because this is the condi-
tion in most Hymenoptera, both non-
aculeate and aculeate, and also in insects
in general. The cercus is absent in the fe-
male of all members of the Aculeata,
however.

Loss of the cercus in the male has ap-
parently occurred on at least four occasions
and is thus not a strong character for indi-
cating higher groups. State 85.1 has ap-
parently arisen (jn internodes 12-14, 19-20
and 23-24 (in each case associating two
taxa) and in the apid group.

*86. Gonapophysis IX of male (Fig. 59).
Primitively, the gonapophyses IX are
fused dorsally to form a chitinous "notum"
(Smith, 1970a) over much of their length.
86.1 — The notal fusion between the gona-
pophyses is desclerotized so that the gona-
pophyses are linked by a membrane along
their entire length dorsally. 86.2 — The
notal fusion between the gonapophyses is

59

88.1

61

Figs. 59-61. Characters of Aculeata. 59, gonapophy-
ses IX of male, ventral view, showing primitive and
derived states (86 based on Sierolomorpha; 86.1 on
Trielis; 86.2 on Fedtschenk.ia); 60, mandible of
larva, showing primitive and derived states of teeth
(all composite); 61, first and second spiracles of
final instar larva, showing derived state (88.1 based
on Pseudomethoca).

536

The University of Kansas Science Bulletin

retracted basally and desclerotized so that
the gonapophyses are linked only basally
by a membrane, and are completely free
apically, over much more than half their
length.

Dorsally fused gonapophyses are con-
sidered to be primitive because this is the
condition in most Hymenoptera, both
non-Aculeata and Aculeata, as well as in
insects in general (Smith, 1969, 1970a).

The particular modifications of the ae-
deagal notum have each apparently arisen
only once in the taxa considered. State
86.1 is characteristic of the scoliid group
only. State 86.2 has apparently arisen on
internode 7-8 and associates the sapygid,
myrmosid and mutillid groups strongly.
The form in Fedtschen\ia (sapygid) for
example, is remarkably similar to that in
Pseiidophotopsis (mutillid), and the myr-
mosid type is easily derived by increased
production of the dorsal area of the apex
of each gonapophysis, a tendency which is
present in the sapygids also. There are
various other modifications involving de-
sclerotization of the notum within other
taxa such as the apids (Snodgrass, 1941),
but generally not in detail like the de-
scribed states.

87. Larval mandibular teeth (Fig. 60).
Primitively, the final instar larva has
strongly sclerotized mandibles which are
quadridentate. 87.1 — The final instar larva
has tridentate mandibles. 87.2 — The final
instar larva has bidentate mandibles. 87.3
— The final instar larva has simple uni-
dentate mandibles.

Quadridentate mandibles are consid-
ered to be primitive because this is the
condition in various aculeate taxa consid-
ered to be relatively primitive on the basis
of other characters.

Although evaluation of the states of
the larval mandibular teeth is considerably
hampered by the complete absence of in-
formation on the larvae of the plumariid,
scolebythid, brachycistidid, sierolomorphid,

typhoctid, eotillid, chyphotid, apterogynid
and bradynobaenid groups, there never-
theless appear to be trends which may
reflect the evolution of the mandible. Thus
state 87.1 may have arisen on internode
15-16 (or possibly on internode 6-15 if it
is present in the sierolomorphids). State
87.2 is present in the apids and sapygids
but this may not be the primitive state for
the latter group since the larvae of Fedt-
schen\ia (the most generally primitive
member) are unknown. State 87.3 is pres-
ent in the rhopalosomatids, although it
does occur within the apid group (e.g..
Apis) and so is weakened, as well as not
indicating higher groups.

*88. Larval spiracles (Fig. 61). Primi-
tively, the final instar larva has ten pairs
of fully developed spiracles, all of similar
size and complexity. 88.1 — The final in-
star larva has only nine pairs of fully de-
veloped spiracles, the second thoracic
spiracle being much reduced in size and
complexity although still distinguishable.

Ten pairs of fully developed spiracles
are considered primitive because this is
the condition in most Hymenoptera, both
non-aculeates and aculeates (Riek, 1970).

Reduction of the second thoracic spi-
racle in the final instar larva (88.1) has
apparently occurred at least twice, on in-
ternode 6-7 and in the pompilid group.
Unfortunately, as for character 87, infor-
mation is lackiniT for nine taxa. The con-
dition observed in the larva of Sapyga
(Torchio, 1972) in which the spiracle is
apparently not reduced (included on the
tree as an apparent reversal to the primi-
tive condition) may have resulted from
the specialized cleptoparasitic habits of
this genus; a more generalized member
such as Fedtschenkja may have the spi-
racle reduced. Apart from this uncer-
tainty, the usefulness of this character is
also somewhat lessened by the fact that
the second spiracle is reduced in early
instar larvae of various other taxa like the

Phylogeny and Classification of the Aculeate Hymenoptera

537

bcihylids (c.i,^, Chrysis; Grancli, 1%1). It
is also apparently somewhat reduced in
even the final instar larva of some vcspids
(e.g., DoUchovespttla; Grandi, 1961).
Nevertheless, because of various differ-
ences in detail, this character does seem to
provide useful information on the rela-
tionship of the taxa above intcrnode 6-7,
as was suggested by Evans (1%5), al-
though his projection of the condition in
the anthoboscid group has proved in-
correct.

89. Number of prey. Primitively, each
larva is supplied with a single host or prey
individual and completes its development
on this individual (V, P or M, and T
appear once in the behavioral formula).
89.1 — Each larva is supplied with numer-
ous individuals of the host or prey and
develops at the expense of many such in-
dividuals (V, P or M, and T appear re-
peatedly in the behavioral formula).

Use of a single prey individual for each
larva is considered primitive because this
is the situation in many Terebrantia and
in various Aculeata considered relatively
primitive on the basis of other characters.

The use of many prey individuals is
apparently a trend shown in various taxa.
Thus state 89.1 has arisen in the formicid
and vespid groups and also in the apid
group but in a somewhat different sense
since the "prey" in the last instance is of
veiietable and not animal origin. State
89.1 has also arisen within the sphecid
group (in many members but not in
Ampulex, e.g.). Although an evaluation
is greatly hampered by the lack of bio-
logical information on the plumariid,
scolebythid, brachycistidid, sierolomorphid,
typhoctid, eotillid, chyphotid, apterogynid
and bradynobaenid groups, it appears that
state 89.1 is of no use in delineating groups
because of its multiple origins. It does,
however, give some indication of relative
advance on the tree.

90. Nest construction. Primitively, the
prey is left exposed or in a natural cavity
and no nest construction or excavation is
undertaken by the female (T, I and C do
not appear in the behavioral formula).
90.1 — The prey is relocated by the female
but no nest construction or excavation is
undertaken (T appears in the behavioral
formula but I and C do not). 90.1.1 — The
prey is relocated by the female which also
constructs a specialized nest or excavation
in which the prey is placed and which is
closed after provisioning and oviposition
(T, I and C appear in the behavioral
formula). 90.1.2 — The prey is relocated by
the female which also constructs a spe-
cialized nest or excavation in which the
prey is placed but which is not closed after
oviposition since provisioning is continu-
ous during the development of the young
(T and I but not C appear in the be-
havioral formula). 90.1.3 — The prey is
relocated by the female which does not
construct any specialized nest or excava-
tion but which does close off the cavity
in which the prey is placed (T and C but
not I appear in the behavioral formula).
90.2 — The prey is not relocated and no
specialized nest construction or excavation
is undertaken by the female which merely
closes off the cavity in which the prey was
discovered, after oviposition (C but not T
and I appears in the behavioral formula).

Lack of any type of prey transporta-
tion or nest construction is considered
primitive because this is the situation in
many Terebrantia and in various Aculeata
considered primitive on the basis of other
characters.

As with character 89, it is difficult to
determine the evolution of nest construc-
tion since data are lacking for nine of the
taxa considered. Nevertheless, state 90.1
has apparently arisen at least twice, on
internodes 4-5 (associating the sphecid and
apid groups) and 18-19 (linking the
formicid, vespid and possibly the scoliid

538

The University of Kansas Science Bulletin

groups). State 90.1.1 has arisen inde-
pendently in the apid and vespid groups;
90.1.2 is characteristic of the formicids;
and 90.1.3 has arisen (apparently de novo,
without prior origin of state 90.1) in the
pompilid group. State 90.2 is character-
istic of the myrniosid and mutillid groups
and probably of the sapygids also, thus
most likely having arisen on internode
7-(S. This state is also present in the
methochid group. Since behavior is by
its very nature more plastic and thus prob-
ably more susceptible to evolutionary
change than is morphology, thus aggra-
vating problems of parallelism and con-
vergence, this character appears to be of
minimal significance in delineating groups,
although it does provide some general
indications of possible groupings.

91. Oviposition sequence. Primitively, ovi-
position occurs on the prey or host after
it has been located [V (and usually P or
M, and T) appears before O in the be-
havioral formula]. 91.1 — Oviposition oc-
curs in the empty nest chamber before the
prey or host has been located (O appears
before V, P or M, and T in the behavioral
formula).

Oviposition after location of the host
is considered primitive because this is the
situation in the Terebrantia and in most
Aculeata.

Early oviposition (91.1) has apparently
arisen independently in the formicid and
vespid groups and is thus not of any sig-
nificance in establishing relationships, al-
though some connection may be implied
by the fact that both occurrences have ap-
peared on the branch subtended by inter-
node 1849.

92. Type of provisions. Primitively, the
larvae are reared on food of animal origin,
viz., other arthropods (subscripts "a" ap-
pear in the behavioral formula). 92.1 —
The larvae are reared on food of vegetable
origin, usually pollen and nectar (sub-

scripts "v" appear in the behavioral
formula).

Provisions of animal origin are con-
sidered primitive because these are the
type used by most Terebrantia and by
most Aculeata, including those considered
relatively primitive on the basis of other
characters.

The use of provisions of vegetable
origin is characteristic of only the apid
group, although a similar state has arisen
within the vespid group (most Masaridae
but not Euparagia; Williams, 1927) and
within the formicid group (the more ad-
vanced members; Wilson, 1971). This
character is thus of no use in the deriva-
tion of higher groupings.

Character States Primitive for the
Taxa Considered

The following accounts are composite
and intended to apply to primitive mem-
bers of the taxa concerned, i.e., the char-
acter states are those possessed by the
ancestral members of the taxon. Such an
ancestral form may be hypothetical since
no particular extant member may show
all the characters in the states given for a
taxon. However, all the states listed are
present in at least some modern species
for each taxon. The accounts are thus
usually not diagnostic in the usual sense
for the taxa involved. Unless otherwise
specified, each character state applies to
both sexes of the adults, except for some
characters of the mesosoma in taxa where
the female is apterous, in which case the
character states are those of the male. The
numbers refer to the preceding discussion
of the characters themselves.

The taxa are referred to below by in-
formal names (the ending "-id" does not
necessarily connote a group at the family
level) so that the final classification de-
rived from the cladogram may be more
readilv constructed from the taxa actuallv

Phylogeny and Classification oi- the Aculeate Hymenoptera

539

considered, wiUiout die accounts ut these
taxa appearing under names which may
dilTer from those adopted in the classifica-
tion, or which may bear connotations at
variance with those suggested by the final
result. Nevertheless, the order in which
the taxa are presented is as close to one
reflecting evolutionary advance as possible,
given the limitations of a linear arrange-
ment. The taxa vary greatly in scope and
categorical level, as may be judged from
the listing of species examined for each,
as well as the citation of which commonly
recognized groups are included. Since it
has proved impossible to check all pre-
viously identified specimens for accuracy
of identification, the names used are those
appearing on the determination labels as-
sociated with the specimens used, except
for a few instances where changes in
generic assignation have been made fol-
lowing recent revisions. Any inisidentifica-
tions that may exist are actually of little
or no significance for the derivation of the
character states for the taxon concerned.

1. Phimariid group.

Material examined: Mynnecopteriua fili-
cornis ( i ); Plumarius spp. ( i , Argen-
tina; 9 , Peru) ; Pliimuroides andalgalensis
( 6 ). ("Plumariidae")
(Data also from Evans, 1967; Nagy, 1969a;
Bradley, 1972. Nagy's "Heterogyninae" —
Heterogyna — is also considered a member
of this group.)

General — Sexual dimorphism considera-
ble although male not more slender than
female; male winged, female apterous
{Heterogyna also; Nagy, in litt.) (1-2).
No sterile caste (3). All pubescence sim-
pie (4).

Head — Cly[">eus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (upper
rim slightly produced in female) (6). Eye
oval with inner margin shallowly sinuate
(margin convex in female), not protuber-

ant; scattered short setae present (glabrous
in female) (7-9). Three ocelli present
(absent in female) (10). Genal secretory
organ absent (11). Antenna with 13 seg-
ments in both sexes; axis of radicle not
deviating much from that of scape (at an
oblic|ue angle in female) ; radicle not in-
serted into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented
(5-segiTiented in female); labial palpus 4-
segmented (3-segmented in female) (15-
17).

Mesosoma — Pronotum large, not reduced
mesally (reduced and short mesally in
male), freely articulating with mesotho-
rax; hind margin arcuate, almost straight
(strongly concave in male) ; anterior "col-
lar" very short (absent in female) ; postero-
lateral angle rounded, reaching tegula;
posteroventral margin approximately
straight; ventral angle rounded, not pro-
duced (18-23). Propleura mesally some-
what separated and shallowly diverging
posteriorly (24). Prosternum somewhat
sunken except anteriorly (25). Forecoxae
somewhat separated (almost contiguous in
female) {26). Mesonotum not anteriorly
produced (enlarged, anteriorly produced
in male) ; scutellum not inuch enlarged
(27-28). Prepectus not fused to mesepi-
sternum, discontinuous midventrally, the
halves contiguous (29). Mesepimeron run-
ning the height of mesopleuron (30).
Mesosternum essentially simple, without
posterior projections (31). Mid-coxae
slightly separated (almost contiguous in
female) (32). Meso- and metapleura ar-
ticulating (33). Metanotum not reduced
medially; metapostnotum forming a dis-
tinct transverse area, fused to propodeum,
slightly sunken but not invaginated (34-
35). Endophragmal pit very close to
anterior margin of metapleuron; no
metapleural gland (36-37). Metasternum
depressed anterolaterally to accommodate
mid-coxae (entirely depressed in male),

540

The University of Kansas Science Bulletin

well-dififerentiated from mesosternum, not
anteriorly produced (38-40). Hind coxae
slightly separated (41). Metathoracic-pro-
podeal pleural suture visible dorsal to
endophragmal pit, obliterated ventral to
pit (42). Propodeum not shortened; disc
and declivity not distinct but merging
(43-44). Forewing with venation reduced
apically; eight closed cells; veins 2s-m and
2m-cu absent (45-46). Pterostigma large
and sclerotized (47-48). Hind wing with
venation well-developed; three closed cells;
veins A and J absent; cu-e originating
proximal to point of separation of M and
Cu; free section of Cu present (49-53).
Basal hamuli dispersed along costa, not
clustered (54). Plical lobe indicated by a
moderate incision; jugal lobe absent (55-
56). Legs slender, generally unmodified
(femora and tibiae somewhat enlarged
in female); arolia well-developed; claws
toothed (simple in female) (57-59). Fore-
tibia with single calcar almost straight,
not highly modified (60). Mid-tibia with
scattered weak spines (moderate spines in
female); two simple spurs (61, 63-65).
Hind coxa simple (66). Hind tibia with
scattered weak spines; two simple spurs
(62, 64, 67-68). Modified mesosoma of
female with pronotum articulating with
mesonotum; propleura fused ventrally and
dorsally forming a rigid tube, slightly pos-
teriorly produced in the midline ventrally;
mesothorax with all sclerites fused, not
distinguishable; metathorax much re-
duced, fused to propodeum; meso-meta-
thoracic articulation very narrow and con-
stricted; metacoxal cavities considerably
separated from mesocoxal cavities so that
mid-coxae are closer to forecoxae than
hind coxae (69).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor petio-
late (anteriorly narrowed and almost peti-
olate in female) ; first tergum with lateral
margin overlying sternum, not fused to

it; first sternum not depressed, more or
less continuous in contour with the second
and overlying it (first sternum barely de-
pressed and separated from second in fe-
male) (72-76). Second segment not
apically constricted (77). Female with
seventh tergum not mesally reduced (al-
though enclosed by enlarged sixth ster-
num); articulation present within section
1 of gonocoxite IX; no valve on gonapo-
physis VIII; gonapophysis IX moderately
curved with apex directed obliquely down-
ward (78-81). Male with seventh sternum
reduced but not entirely hidden; hypo-
pygium simple, little modified although
slightly narrowed; cercus present; gonapo-
physes IX with notal fusion well-devel-
oped, especially basally (82-86).

Larva — Unknown (that ascribed to Plii-
mariits by Janvier, 1933:324, appears to be
that of a mutillid from the figure and
description).

Behavioral formula — Unknown.

2. Bethylid group.

Material examined: Anisepyris sitbvio-
laceiis ( ci , 9 ) ; Apenesia coliimbana ( cJ ) ;
Bethyhis amoeniis ( 9 ) ; Chrysis grandis
( (i , 9 ); cleptid spp. ( i , California; 9 ,
India) ; Deinodryintis henshawi ( 9 ) ;
dryinid sp. (9, Brazil); Eiichroeus pur-
puratiis ( (5 ) ; Goniozus joveolatHS ( 9 ) ;
Goniozus spp. ( 6 , New York; 9 , Flor-
ida); Hedychridiiun carinidatum {$)',
Omaliis aiiratits { $)\ Parasierola spp.
((5,9, California) ; Parnopes grandior
{i )\ Pristocera arm if era ( S ); Prosierola
bicarinata ( 9 ) ; Pseiidisobrachiiim coxalis
{i)\ P. fJavinervis ( 5 ) ; Rhabdepyris
megacephalits ( <5 , 9 ). ("Bethylidae,"
"Chrysididae," "Cleptidae," "Dryinidae")
(Data also from Haupt, 1938; Grandi,
1961; Maa & Yoshimoto, 1961; Evans,
1964a. The following taxa are also consid-
ered to be members of this group: "Em-
bolemidae," "Loboscelidiidae," "Sclero-
gibbidae.")

Phylogeny and Classification of the Aculeate Hvmenoptera

541

General — Sexual dimorphism minimal,
both sexes of similar form and lully
winged (1-2). No sterile caste (3). All
pubescence simple (4).

Head — Clvpeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin very shallowly
sinuate, not protuberant; scattered pores
without evident setae (7-9). Three ocelli
present (10). Genal secretory organ ab-
sent (11). Antenna with 13 segments in
both sexes; axis of radicle not deviating
much from that of scape; radicle not set
into basal invagination of scape (12-14).
Labio-maxillary complex not elongated;
maxillary palpus 6-segmented; labial pal-
[uis 3-segmented (15-17).

Mesoso.\l\ — Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin arcuate, almost straight;
anterior "collar" well-developed; posterolat-
eral angle rounded, reaching tegula; pos-
teroventral margin approximately straight;
ventral angle rounded, not produced (18-
23). Propleura mesally separated and di-
verging posteriorly (24). Prosternum not
sunken (25). Forecoxae somewhat sepa-
rated (26). Mesonotum not anteriorly
produced; scutellum not enlarged (27-28).
Prepectus not fused to mesepisternum, di-
vided midventrally, the two halves con-
tiguous (29). Mesepimeron running the
height of mesopleuron (30). Mesosternum
without posterior projections (31). Mid-
coxae somewhat separated (32) . Meso- and
metapleura articulating (33). Metanotum
not reduced medially; metapostnotum in-
distinct, indistinguishable from propo-
deum medially, apparently not invagi-
nated (34-35). Endophragmal pit very
close to anterior margin of metapleuron;
no metapleural gland (36-37). Metaster-
num depressed laterally to accommodate
mid-coxae, well-dififerentiated from meso-
sternum, not anteriorly produced (38-40).

Hind coxae almost contiguous (41). Meta-
thoracic-propodeal pleural suture barely
visible dorsal to endophragmal pit, ob-
literated ventral to pit (42). Propodeum
not shortened; disc and declivity slightly
distinct but merging (43-44). Forewing
with venation reduced apically; seven
closed cells; vein M absent distal to sepa-
ration from S; vein Cui absent distal to
separation from Cui;; crossveins Is-m,
2s-m and 2m-cu absent (45-46). Ptero-
stigma large and sclerotized (47-48). Flind
wing with venation reduced apically; one
closed cell (C) ; vein M4-Cu and all cross-
veins absent; veins A and J absent; traces
of cu-e originating proximal to separation
of M and Cu (49-52). Basal hamuli dis-
persed along costa, not clustered (54).
Plical lobe indicated by a moderate in-
cision; jugal lobe absent (55-56). Legs
slender, unmodified; arolia well-devel-
oped; claws toothed (57-59). Foretibia
with single calcar almost straight, not
highly modified (60). Mid-tibia with
some scattered setae spiniform; two simple
spurs (61, 63-65). Hind coxa simple (66).
Hind tibia with some scattered setae spin-
iform; two simple spurs (62, 64, 67-68).

Metasoxl\ — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-
olate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, more or less con-
tinuous in contour with the second (72-76).
Second segment not apically constricted
(77). Female with seventh tergum
partially exposed, not much reduced; ar-
ticulation present within section 1 of
gonocoxite IX; valve present on gonapo-
physis VIII; gonapophysis IX moderately
curved with apex directed obliquely down-
ward (78-81). Male with seventh sternum
not reduced nor hidden; hypopygium sim-
ple, unmodified; cercus present; gonapo-
physes IX with notal fusion well-devel-
oped (82-86).

542

The University of Kansas Science Bulletin

Final instar larva — Mandible quadriden-
tate (87). Ten pairs of fully developed
spiracles (88).

Behavioral formula — VaPaO (89-92).

3. Scolebythid group.

Material examined: Clystopsenella longi-
ventris ( 9 ) . ("Scolebythidae")
(Data also from Evans, 1963, 1967.)

General — Sexual dimorphism minimal,
both sexes of similar form and fully
winged (1-2). No sterile caste (3). All
pubescence simple (4).

Head — Clypeus very short, reduced; an-
tennae inserted very low on face (5). An-
tennal socket simple, with mesal margin
only slightly raised, no tubercle (6). Eye
oval with inner margin convex, not pro-
tuberant; scattered pores with short setae
(7-9). Three ocelli present (10). Genal
secretory organ absent (11). Antenna
with 13 segments in both sexes; axis of
radicle not deviating much from that of
scape; radicle not inserted in basal in-
vagination of scape (12-14). Labio-maxil-
lary complex not elongated; maxillary
palpus 6-segmented; labial palpus 4-seg-
mented (15-17).

Mesosoma — Pi'onotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin arcuate, almost straight;
anterior "collar" absent; posterolateral
angle rounded, reaching tegula; postero-
ventral margin approximately straight;
ventral angle rounded, not produced (18-
23). Propleura mesally considerably sepa-
rated and diverging posteriorly (24).
Prosternum large, not sunken (25). Fore-
coxae broadly separated basally but pos-
teriorly produced, thus becoming contig-
uous (26). Mesonotum not anteriorly
produced; scutellum not enlarged (27-28).
Prepectus not fused to mesepisternum, dis-
continuous midventrally, the halves con-
siderably separated (29). Mesepimeron
running the height of mesopleuron (30).

Mesosternum without posterior projec-
tions (31). Mid-coxae somewhat separated
(32). Meso- and metapleura articulating
i}^). Metanotum reduced and very short
medially; metapostnotum forming an in-
distinct transverse area, not clearly distin-
guishable from propodeum medially, not
invaginated (34-35). Endophragmal pit
very close to anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum depressed laterally to accom-
modate mid-coxae, well-diflFerentiated from
mesosternum, not anteriorly produced
(38-40). Hind coxae contiguous (41).
Metathoracic-propodeal pleural suture visi-
ble dorsal to endophragmal pit, obliterated
ventral to pit (42). Propodeum not short-
ened; disc and declivity not distinct but
merging (43-44). Forewing with venation
reduced apically; six closed cells; vein M
absent distal to separation from S; vein
Cui4-Cu2 reduced distal to junction with
Im-cu; vein E absent distal to junction
with cu-e; crossveins Is-m, 2s-m and 2m-
cu absent (45-46). Pterostigma large and
sclerotized (47-48). Hind wing with vena-
tion much reduced; no closed cells; veins
A and J absent; origin of cu-e not deter-
minable (49-52). Basal hamuli dispersed
along costa, not clustered (54). Plical lobe
indicated by a moderate incision; jugal
lobe absent (55-56). Legs stout with fe-
mora inflated; arolia small but distinct;
claws with blunt basal tooth (57-59).
Foretibia with single calcar almost straight,
not highly modified (60). Mid-tibia spine-
less; two simple spurs (61, 63-65). Hind
coxa simple (66). Hind tibia spineless;
two simple spurs (62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor petio-
late; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, more or less con-
tinuous in contour with the second and
overlying it (72-76). Second segment not

PlIYLOCIENY AND ClASSH-ICATION OF THli AcULLATK HVMF.NOPTERA

543

apicall}- constricted (77). Female with
seventh tergum barely exposed, not mes-
ally reduced; articulation [-jrescnt within
section 1 of gonocoxite IX; valve present
on gonap()[")hysis VIII; gonapophysis IX
moderately curved with apex directed ob-
liquely downward (78-(Sl). Male with
hypopygium simple, unmodified (S3) (ad-
ditional character states of terminalia
unknown).

Larva — Unknown.

Reiia\i()ral FORMULA — Unkuowu.

4. Sphecid group.

Material examined: ampulicid spp. (6,
Costa Rica; 9 , Sri Lanka); Aphilanthops
taiinila ( c5 , ? ) ; Astata occidentalis ( ci ,

9 ) ; Bembix hinei ( c5 , 9 ) ; Cerceris bi-
cornitta (6,9); Chlorion aerarium ( $ ,

9 ) ; Crabro largior ( c5 , 9 ) ; Dasyproctits
simillitnits (6,9); Epinysson bell us ( 6 ,

9 ) ; Gorytes provancheri ( c5 , 9 ) ; Hapa-
lomellinus albitomentosus (6,9); Hop-
lisoides gulielmi (6,9); Larropsis con-
jerta (6,9); Lyroda subita (6,9);
Mellinus sp. (6,9; Michigan); Motes
argentata (6,9); Nysson aequalis ( 6 ,

9 ) ; Oxybelus cornutus (6,9); Philan-
thus gibbosus (6,9); Podalonia commu-
nis (6,9); Prionyx joxi (6,9);
Pseneo montezuma (6,9); Pseudoplisus
phaleratus (6,9); Sphex ichneumoneus
(6,9); T achy sphex tarsatus (6,9);
Trypargilum politum (6,9); Zyzzyx
chilensis (6,9). ("Sphecidae" sensu lato)
(Data also from Williams, 1919; Leclercq,
1954; Evans & Lin, 1956; Evans, 1959a,
1966.)

General — Sexual dimorphism minimal,
male slightly smaller than female; both
sexes fully winged (1-2). No sterile caste
(3). All pubescence simple (4).

Head — Clypeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin shallowly sinuate,

not protuberant; scattered pores without
evident setae (7-9). Three ocelli present
(10). (icnal secretory organ absent (11).
Antenna with 13 segments in male, 12
segments in female; axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesoso.\l\ — Pronotum reduced mesally,
eliminating laterodorsal surface, fairly
freely articulating with mesothorax, espe-
cially dorsally; hind margin broadly U-
shaped; anterior "collar" well-developed;
posterolateral angle much reduced above
spiracular lobe, not reaching tegula; pos-
teroventral margin slightly concave; ven-
tral angle considerably produced mesally,
almost meeting its counterpart midven-
trally (18-23). Propleura mesally slightly
separated and shallowly diverging posteri-
orly (24). Presternum sunken except for
moderately large anterior rim (25). Fore-
coxae contiguous (26). Mesonotum some-
what anteriorly produced; scutellum not
much enlarged (27-28). Prepectus de-
pressed, fused to mesepisternum, contin-
uous midventrally (29). Mesepimeron re-
duced, very short but distinguishable the
entire height of mesopleuron (30). Meso-
sternum slightly posteriorly produced be-
tween coxal bases but not overlying them
(31). Mid-coxae slightly separated (32).
Meso- and metapleura not mutually mov-
able but not completely fused (33). Meta-
notum not reduced medially; metapostno-
tum mesally greatly expanded posteriorly,
forming most of apparent disc of propo-
deum, neither sunken nor invaginated
(34-35). Metepimeron expanded ventrally
anterior to endophragmal pit so that pit
is some distance from anterior margin of
metapleuron, additional ventral pit devel-
oped; no metapleural gland (36-37). Meta-
sternum depressed laterally to accommo-
date mid-coxae, well-differentiated from

544

The University of Kansas Science Bulletin

mesosternum except mesally, not anteriorly
produced (38-40). Hind coxae contiguous
(41). Metathoracic-propodeal pleural su-
ture entirely visible, more distinct dorsal
to endophragmal pit than ventrally (42).
Propodeum effectively not shortened dor-
sally except as a result o£ development of
metapostnotum; disc essentially absent but
merging with declivity (43-44). Forewing
with venation not much reduced apically;
ten closed cells (45-46). Pterostigma fairly
small and sclerotized (47-48). Hind wing
with venation not much reduced apically;
three closed cells; veins A and J present;
cu-e originating proximal to separation of
M and Cu; free section of Cu present
(49-53). Basal hamuli dispersed along
costa, not clustered (54). Plical lobe indi-
cated by a moderate incision; large jugal
lobe indicated by a shallow notch (55-56).
Legs slender, generally unmodified; arolia
well-developed; claws toothed (57-59).
Foretibia with single calcar almost straight,
not much modified (60). Mid-tibia with
scattered moderate spines; two spurs, both
slightly dorsally flattened with serrate
margins (61, 63-65). Hind coxa simple
(66). Hind tibia with scattered moderate
spines; two spurs, both slightly dorsally
flattened with serrate margins (62, 64,
67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor pet-
iolate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, not differentiated
from second but overlying it (72-76).
Second segment not apically constricted
(77). Female with seventh tergum hidden
and reduced but with continuous anterior
sclerotization; no articulation within sec-
tion 1 of gonocoxite IX; valve present on
gonapophysis VIII; gonapophysis IX al-
most straight with apex directed slightly
upward (78-81). Male with seventh ster-
num neither reduced nor hidden; hypo-

pygium simple, unmodified; cercus pres-
ent; gonapophyses IX with notal fusion
well-developed (82-86).

Final instar larva — Mandible quadriden-
tate (87). Ten pairs of fully developed
spiracles (88).

Behavioral formula — VaPaTaO (89-92).

5. Apid group.

Material examined: Andrena striatijrons
( ci , 9 ) ; Apis mellifera (5,2,^); Az-
tecanthidiiim xochipiUiiim ((5,9 );Brachy-
hesma incompleta ( c$ , 9 ) ; Callomelitta
sp. ( c5 , 9 ; Australia) ; Colletes niidus
((5,9); Euglossa cordata ( c? , 9 ) ; Fidelia
kobrowi ( <5 ) ; Fidelia sp. ( 9 , South
Africa) ; Halictiis ligatus ((5,9); Hylaeus
basalis ( c5 , 9 ) ; Leioproctus herbsti ( $ ,

9 ); Lithurgits apicalis ((5,9); Mega-
chile parallela ((5,9); Megalopta cen-
tralis ( c5 , 9 ) ; Melissodes microsticta ( 5 ,

9 ) ; Melitta tricincta ((5,9); Mesocheira
bicolor ((5,9); Nomada superba ((5,9);
Notnadopsis zebrata ((5,9); Nomia me-
landeri ( c5 , 9 ) ; Osiris spp. ( $ , Brazil;

9 , Panama) ; Protoxaea jerruginea ( $ ,

9 ) ; Ptilothrix fractifer ((5,9); Serico-
gaster jasciata ( c5 , 9 ) ; Sy strop ha curvi-
cornis ((5,9); Xylocopa micans ((5,9).
("Colletidae," "Halictidae," "Oxaeidae,"
"Andrenidae," "Melittidae," "Fideliidae,"
"Megachilidae," "Anthophoridae," "Api-
dae")

(Data also from Snodgrass, 1941; Mich-
ener, 1944, 1953, 1965, 1974; Daly, 1964;
Exley, 1968; Eickwort, 1969; Stephen,
Bohart & Torchio, 1969.)

General — Sexual dimorphism minimal,
male slightly smaller than female; both
sexes fully winged (1-2). No sterile caste
(3). Much erect pubescence plumose (4).

Head — Clypeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin shallowly sinuate,
not protuberant; scattered pores without

Phylogeny and Classification of the Aculeate Hymenoptera

545

evident setae (7-9). Three ocelli present
(10). Genal secretory organ absent (11).
Antenna with 13 segments in male, 12
segments in female; axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex slightly
elongated, especially in stipes and pre-
mentum; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum reduced mesally, eli-
minating laterodorsal surface, fairly freely
articulating with mesothorax, especially
dorsally; hind margin broadly U-shaped;
anterior "collar" well-developed; postero-
lateral angle much reduced above spirac-
ular lobe, reaching tegula; posteroventral
margin slightly concave; ventral angle
considerably produced mesally, almost
meeting its counterpart midventrally (18-
23). Propleura mesally slightly separated
and shallowly diverging posteriorly (24).
Prosternum sunken except for fairly long
anterior region (25). Forecoxae contigu-
ous (26). Mesonotum somewhat anteri-
orly produced; scutellum not greatly en-
larged (27-28). Prepectus depressed, fused
to mesepisternum, continuous midven-
trally (29). Mesepimeron much reduced,
vaguely distinguishable only at dorsal ex-
treme (30). Mesosternum very slightly
posteriorly produced between coxal bases
but not overlying them (31). Mid-coxae
slightly separated (32). Meso- and meta-
pleura not mutually movable but appar-
ently not fused (H). Metanotum not
reduced medially; metapostnotum mesally
greatly expanded posteriorly, forming most
of apparent disc of propodeum, neither
sunken nor invaginated (34-35). Mete-
pimeron ventrally expanded anterior to
endophragmal pit so that pit is some dis-
tance from anterior margin of metapleu-
ron, additional ventral pit developed; no
metapleural gland (36-37). Metasternum
depressed laterally to accommodate mid-
coxae, mesallv not well-differentiated from

mesosternum, not anteriorly produced (38-
40). Hind coxae contiguous (41). Meta-
thoracic-propodeal pleural suture visible
dorsal to endophragmal pit, not distin-
guishable ventrally (42). Propodeum ef-
fectively not shortened dorsally except as
a result of development of metapostnotum;
disc essentially absent but merging with
declivity (43-44). Fore wing with venation
somewhat reduced apically; ten closed
cells (45-46). Pterostigma fairly small and
sclerotized (47-48). Hind wing with vena-
tion only very slightly reduced apically;
two closed cells; vein C reduced distally;
vein A present, vein J absent; cu-e origi-
nating proximal to separation of M and
Cu; free section of Cu present (49-53).
Basal hamuli dispersed along margin, not
clustered (54). Plical lobe indicated by a
moderate incision; large jugal lobe indi-
cated by a moderate incision (55-56). Legs
slender, generally unmodified; arolia well-
developed; claws toothed (57-59). Fore-
tibia with single calcar almost straight,
little modified (60). Mid-tibia with scat-
tered weak spines; a single spur, slightly
dorsally flattened with serrate margins
(61, 63-65). Hind coxa simple (66). Hind
tibia with scattered weak spines; two
spurs, both slightly dorsally flattened with
serrate margins (62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor pet-
iolate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, not differentiated
from second but overlying it (72-76) . Sec-
ond segment not apically constricted (77).
Female with seventh tergum hidden, re-
duced, forming two separate sclerites; no
articulation within section 1 of gonocoxite
IX; valve present on gonapophysis VIII;
gonapophysis IX almost straight with
apex directed slightly upward (78-81).
Male with seventh sternum reduced and
concealed; hypopygium simple but con-

546

The University of Kansas Science Bulletin

cealed; cercus absent; gonapophyses IX
with notal fusion well-developed (82-86).

Final instar larva — Mandible bidentate
(87). Ten pairs of fully developed spi-
racles (88).

Behavioral formula— I (VvTv)"MOC (89-
92).

6. Anthoboscid group.
Material examined: Anthobosca ervthro-
nota ( <5 , 9 ) ; A. ?flamcornis (6,9);
Anthobosca spp. ( S , South Africa; 9 ,
Argentina); Lalapa liisa (S,9). ("An-
thoboscinae" of the "Tiphiidae")
(Data also derived from Turner, 1912;
Janvier, 1933; Pate, 1947a; Evans, 1965.)

General — Sexual dimorphism moderate,
male somewhat more slender than female;
both sexes fully winged (1-2). No sterile
caste (3). All pubescence simple (4).

Head — Clypeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin shallowly sinuate,
not protuberant; scattered pores without
evident setae (7-9). Three ocelli present
(10). Genal secretory organ absent (11).
Antenna with 13 segments in male, 12
segments in female; axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin shallowly arcuate; ante-
rior "collar" well-developed; posterolateral
angle rounded, reaching tegula; postero-
ventral margin approximately straight;
ventral angle rounded, not produced (18-
23). Propleura posteriorly contiguous mes-
ally (24). Prosternum sunken except for
relatively large anterior rim (25). Fore-
coxae very nearly contiguous (26). Meso-
notum not anteriorly produced; scutellum

not enlarged (27-28). Prepectus not fused
to mesepisternum, discontinuous midven-
trally, the halves considerably separated
(29). Mesepimeron running the height of
mesopleuron (30). Mesosternum with
platelike projections posteromesally over
coxal cavities (31). Mid-coxae slightly
separated (32). Meso- and metapleura
articulating (Si). Metanotum not reduced
medially; metapostnotum visible medially,
somewhat invaginated (34-35). Endo-
phragmal pit very close to anterior margin
of metapleuron ; no metapleural gland (36-
37). Metasternum depressed anterolater-
ally to accommodate mid-coxae, slightly
raised posteromesally, well-diflFerentiated
from mesosternum, not anteriorly pro-
duced (38-40). Hind coxae only slightly
separated (41). Metathoracic-propodeal
pleural suture entirely distinct (42). Pro-
podeum not shortened; disc and declivity
not distinct but merging (43-44). Fore-
wing with venation not much reduced
apically (slightly reduced in female) ; ten
closed cells (45-46). Pterostigma fairly
large and sclerotized (47-48). Hind wing
with venation not reduced apically; three
closed cells; veins A and J absent; cu-e
originating distal to separation of M and
Cu; free section of Cu present (49-53).
Basal hamuli dispersed along margin, not
clustered (54). Plical lobe indicated by a
moderate incision; moderate jugal lobe
present (55-56). Legs slender, generally
unmodified (stouter in female with fe-
mora and tibiae enlarged) ; arolia well-
developed; claws toothed (57-59). Fore-
tibia with single calcar almost straight,
not highly modified (60). Mid-tibia with
scattered moderate spines; two simple
spurs (spurs slightly dorsally flattened in
female) (61, 63-65). Hind coxa simple
(66). Hind tibia with scattered moderate
spines; two simple spurs (spurs slightly
dorsally flattened in female) (62, 64,
67-68).

Metasoma — No specialized "felt lines" or

Phylogeny and Classification of the Aculeate Hymenoptera

547

stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor pet-
iolate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, not highly differ-
entiated from second (72-76). Second seg-
ment not apically constricted (77). Fe-
male with seventh tergum hidden, reduced
but continuously sclerotized anteriorly; no
articulation within section 1 of gonocoxite
IX; valve present on gonapophysis VIII;
gonapophysis IX moderately curved with
apex directed obliquely downward (78-
81). Male with seventh sternum some-
what reduced although not hidden; hypo-
pygium slightly narrowed, lingulate, little
modified, concealed basally; cercus pres-
ent; gonapophyses IX with notal fusion
well-developed, especially basally (82-86).

Final ixstar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (88)
(pers. obs.).

Behavioral formula— VuPaO (89-92) .

7. Thynnid group.

Material examined: Diamma bicolor
( 9 ) ; Eiione sp. {6 , Australia) ; Elaph-
r opt era scoliaejormis ( <5 , 9 ) ; Eli doth yn-
nus basalis ( dl , 9 ) ; Hemithynnus aptenis
( (? , 9 ) ; thynnine spp. ( (^ , 9 ; Argentina,
Australia) ; Thynnoides fuwipennis ( S ,
9); Zaspilothynnus dilatatiis (c5,9).
("Thynninae" of the "Tiphiidae")
(Data also from Turner, 1910; Janvier,
1933; Clausen, 1940; Reid, 1941; Pate,
1947a; Given, 1954, 1958, 1959; Salter,
1957; RidsdiU Smith, 1970a & b.)

General — Sexual dimorphism considera-
ble although male only slightly more slen-
der than female; male winged, female ap-
terous (1-2). No sterile caste (3). All
pubescence simple (4).

Head — Clypeus of moderate size, slightly
dorsally produced; antennae inserted fairly
low on face (5). Antennal socket simple

although slightly raised dorsomedially,
without true tubercle (6). Eye oval with
inner margin very shallowly sinuate, not
protuberant; scattered pores without evi-
dent setae (7-9). Three ocelli present (10).
Genal secretory organ absent (11). An-
tenna with 13 segments in male, 12 seg-
ments in female; axis of radicle not devi-
ating much from that of scape; radicle not
set into basal invagination of scape (12-
14). Labio-maxillary complex not elon-
gated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin shallowly arcuate; ante-
rior "collar" well-developed; posterolateral
angle rounded, reaching tegula; postero-
ventral margin approximately straight;
ventral angle rounded, not produced (18-
23). Propleura posteriorly contiguous mes-
ally (24). Prosternum sunken except for
anterior rim (25). Forecoxae contiguous
(26). Mesonotum not anteriorly pro-
duced; scutellum not enlarged (27-28).
Prepectus not fused to mesepisternum,
discontinuous midventrally, the halves con-
siderably separated (29). Mesepimeron
running the height of mesopleuron (30).
Mesosternum with platelike projections
posteromesally over coxal cavities (31).
Mid-coxae sHghtly separated (32). Meso-
and metapleura articulating (33). Meta-
notum not reduced medially; meta-
postnotum visible medially, somewhat
invaginated (34-35). Endophragmal pit
very close to anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum depressed anterolaterally to
accommodate mid-coxae, slightly raised
posteromesally, well-difTerentiated from
mesosternum, not anteriorly produced (38-
40). Hind coxae contiguous (41). Meta-
thoracic propodeal pleural suture entirely
distinct (42). Propodeum not shortened;
disc and declivity not distinct but merg-
ing (43-44). Forewing with venation not

548

The University of Kansas Science Bulletin

much reduced apically; ten closed cells
(45-46). Pterostigma fairly large and
sclerotized (47-4(S). Hind wing with vena-
tion not much reduced apically; three
closed cells; veins A and } absent; cu-e
originating distal to separation of M and
Cu; free section of Cu present (49-53).
Basal hamuli dispersed along costa, not
clustered (54). Plical lobe barely indi-
cated; moderate jugal lobe present (55-56).
Legs slender, generally unmodified (some-
what stouter in female with femora and
tibiae somewhat enlarged); arolia well-
developed; claws toothed (57-59). Fore-
tibia with single calcar almost straight,
not highly modified (60). Mid-tibia with
scattered moderate spines; two simple
spurs (spurs slightly dorsally flattened in
female) (61, 63-65). Hind coxa with weak
dorsal carina (66). Hind tibia with scat-
tered moderate spines; two simple spurs
(spurs slightly dorsally flattened in fe-
male) (62, 64, 67-68). Modified mesosoma
of female with pronotum large, articu-
lating with mesothorax; mesonotum short;
scuto-scutellar suture present; separate pre-
pectal sclerite present; mesepimeron dif-
ferentiated from mesepisternum; meso-
and metapleura articulating; mesonotum
very short and transverse; endophragmal
pit very close to anterior margin of meta-
pleuron; metathoracic-propodeal notal and
pleural sutures visible (69).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-
olate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum very slightly posteriorly depressed,
not highly dilTerentiated from second (72-
76). Second segment not apically con-
stricted (77). Female with seventh tergum
hidden, reduced but continuously sclero-
tized anteriorly; no articulation within sec-
tion 1 of gonocoxite IX; valve present on
gonapophysis VIII; gonapophysis IX mod-
erately curved with apex directed obliquely

downward (78-81). Male with seventh
sternum somewhat reduced although not
hidden; hypopygium simple, unmodified;
cercus present; gonapophyses IX with
notal fusion well-developed, especially bas-
ally (82-86).

Final instar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (88)
(pers. obs.).

Behavioral formula — VaPaO (89-92).

8. Mvzinid group.

Material examined: Bnmnsomeria mutil-
loides ( 9 ) ; B. quadraticeps (5,9);
Meria injradentata ( 9 ) ; M. perornata
((5,9); M. se mini fa (5,9); Mesa cap-
itata (5,9); Myziniim diibiositm ( $ ,
9)\M. qiiinqiiecincttim (5,9); Pterom-
brus spp. (5, Mexico; 9, Argentina).
("Myzininae" of the "Tiphiidae")
(Data also from Williams, 1928; Krom-
bein, 1937, 1968; Clausen, 1940; Pate,
1947a; Evans, 1965.)

General — Sexual dimorphism marked,
male much more slender than female;
both sexes fully winged (1-2). No sterile
caste (3). All pubescence simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim some-
what raised by protuberance of frons,
forming a frontal ledge (6). Eye oval
with inner margin shallowly sinuate, not
protuberant; scattered pores without evi-
dent setae (7-9). Three ocelli present (10).
Genal secretory organ absent (H). An-
tenna with 13 segments in male, 12 seg-
ments in female; axis of radicle not deviat-
ing much from that of scape; radicle not
set into basal invagination of scape (12-
14). Labio-maxillary complex not elon-
gated; maxillary palpi 6-segmented; labial
palpi 4-segmented (15-17).

Mesosoma — Pronotum large, not reduced
mesally, freely articulating with mesotho-

Phylogeny and Classification of the Aculeate Hymenoptera

549

rax; hind mart^in shallowly arcuate; ante-
rior "collar" well-developed; posterolateral
angle rounded, reaching tegula; postero-
ventral margin approximately straight;
ventral angle rounded, not produced (18-
2.t). Propleura posteriorly contiguous mes-
ally (24). Prosternum sunken except for
anterior rim (25). Forecoxae very nearly
contiguous (26). Mesonotum not anteri-

i orly produced; scutellum not enlarged
(27-28). Prepectus not fused to mesepi-
sternum. discontinuous midventrally, the
halves considerably separated (29). Mes-

I epimeron running the height of meso-
pleuron (30). Mesosternum with platelike
projections posteromesally over coxal cavi-
ties (31). Mid-coxae slightly separated
(32). Meso- and metapleura articulating
(33). Metanotum not reduced medially;
metapostnotum not visible medially, in-
vaginated (34-35). Endophragmal pit very
close to anterior margin of metapleuron;
no metapleural gland (36-37). Metaster-
num depressed anterolaterally to ac-
commodate mid-coxae, slightly raised
posteromesally, well-differentiated from
mesosternum, not anteriorly produced (38-
40). Hind coxae slightly separated (41).
Metathoracic-propodeal pleural suture en-
tirely distinct (42). Propodeum not short-
ened; disc and declivity not distinct but
merging (43-44). Forewing with venation
not much reduced apically; ten closed cells
(45-46). Pterostigma fairly large and
sclerotized (47-48). Hind wing with vena-
tion not much reduced apically; three
closed cells; veins A and J absent; cu-e
originating distal to separation of M and
Cu; free section of Cu present (49-53).
Ikisal hamuli dispersed along margin, not
clustered (54). Plical lobe indicated by a
moderate incision; long jugal lobe marked
by a deep incision (55-56). Legs slender,
generally unmodified (stouter in female
with femora and tibiae enlarged); arolia
well-developed; claws toothed (57-59).
Foretibia with single calcar almost

straight, not highly modified (60). Mid-
tibia with scattered moderate spines; two
simple spurs (spurs slightly dorsally flat-
tened in female) (61, 63-65). Hind coxa
simple (66). Hind tibia with scattered
moderate spines; two simple spurs (spurs
slightly flattened in female) (62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-
olate (slightly constricted in male); first
tergum with lateral margin overlying ster-
num, not fused to it; first sternum exten-
sively depressed, most strongly so posteri-
orly, separated from second by a deep
constriction (72-76). Second segment not
apically constricted (77). Female with
seventh tergum hidden, reduced but con-
tinuously sclerotized anteriorly; no articu-
lation within section 1 of gonocoxite IX;
valve present on gonapophysis VIII ; gona-
pophysis IX moderately curved with apex
directed obliquely downward (78-81).
Male with seventh sternum neither re-
duced nor hidden; hypopygium a single
upcurved spine, hardly concealed basally;
cercus present; gonapophyses IX with no-
tal fusion extensive (82-86).

Final instar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (88).

Behavioral formula — VaPaO (89-92).

9. Methochid group.

Material examined: Methocha haemar-
rhoidalis ( 9 ) ; M. minima ( $ ) ; M.
mosotiiana (6,9); M. stygia ( (^ , 9 ).
("Methochinae" of the "Tiphiidae")
(Data also from Williams, 1919; Pate,
1947a; Burdick & Wasbauer, 1959; Ha-
mann, 1960; Evans, 1965; Wilson &
Farish, 1973.)

General — Sexual dimorphism considera-
ble, male much more slender than female;
male winged, female apterous (1-2). No

550

The University of Kansas Science Bulletin

sterile caste (3). All pubescence simple
(4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim only
very slightly raised, forming a very slight
frontal ledge (6). Eye oval with inner
margin oblic]uely convex, not protuberant;
many scattered pores with moderately long
setae (7-9). Three oceUi present (10).
Genal secretory organ absent (U). An-
tenna with 13 segments in male, 12 seg-
ments in female; axis of radicle not devi-
ating much from that of scape; radicle not
set into basal invagination of scape (12-14).
Labio-maxillary complex not elongated;
maxillary palpus 6-scgmented; labial pal-
pus 4-segmented (15-17).
Mesosoma— Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin shallowly arcuate; ante-
rior "collar" well-developed; posterolateral
angle rounded, reaching tegula; postero-
ventral margin slightly concave; ventral
angle rounded, not produced (18-23). Pro-
pleura posteriorly contiguous mesally (24) .
Prosternum sunken except for anterior
rim (25). Forecoxae contiguous (26).
Mesonotum not anteriorly produced; scu-
tellum not enlarged (27-28). Prepectus
not fused to mesepisternum, discontinuous
midventrally, the halves considerably sepa-
rated (29). Mesepimeron running the
height of mesopleuron (30). Mesosternum
essentially simple but with dentate projec-
tions anteromesal to coxal cavities (31).
Mid-coxae slightly separated (32). Meso-
and metapleura articulating (33). Meta-
notum not reduced medially; metapostno-
tum barely visible medially, invaginated
(34-35). Endophragmal pit very close to
anterior margin of metapleuron; no meta-
pleural gland (36-37). Metasternum de-
pressed anterolaterally to accommodate
mid-coxae, slightly raised posteromesally,
well-differentiated from mesosternum, not
anteriorly produced (38-40). Hind coxae

contiguous (41). Metathoracic-propodeal
pleural suture visible dorsal to endophrag-
mal pit, not distinguishable ventrally (42).
Propodeum not shortened; disc and de-
clivity not distinct but merging (43-44).
Forewing with venation not much re-
duced apically; nine closed cells; vein S
incomplete just proximal to junction with
r-s (45-46). Pterostigma fairly large and
sclerotized (47-48). Hind wing with vena-
tion only slightly reduced apically; two
closed cells; vein C reduced distally; veins
A and J absent; cu-e originating distal to
separation of M and Cu; free section of
Cu present (49-53). Basal hamuli dis-
persed along margin, not clustered (54).
Plical lobe indicated by a moderate in-
cision; moderate jugal lobe present (55-
56). Legs slender, generally unmodified;
arolia well-developed; claws toothed (57-
59). Foretibia with single calcar almost
straight, not highly modified (60). Mid-
tibia with scattered weak spines; two sim-
ple spurs (only one, modified as calcar by
dorsal pectination in female) (61, 63-65).
Hind coxa simple (with well-developed
dorsal carina in male) (66). Hind tibia
with scattered weak spines; two spurs, the
inner modified as calcar by dorsal pectina-
tion (only one spur, similarly modified, in
female) (62, 64, 67-68). Modified meso-
soma of female much narrowed; prono-
tum large, articulating with mesothorax;
separate prepectal sclerite present; meso-
thorax long dorsally; scuto-scutellar suture
present; mesepimeron not differentiated
from mesepisternum; meso- and meta-
pleura freely articulating; metathorax with
notum very short and transverse; meta-
pleuron reduced anterior to endophragmal
pit; metathoracic-propodeal notal and
pleural sutures visible dorsal to pit, not
discernible ventrally; metacoxal cavities
somewhat separated from mesocoxal cavi-
ties although mid-coxae still closer to hind
coxae than to forecoxae (69).

MetasoiMa— No specialized "felt lines" or

Phvlogi-ny and Classification of the Aculeate Hymenoptera

551

stridulaiory a[i[uratu,s (70-71). First seg-
ment neither apically constricted nor peti-
olate (slightly constricted in male); first
tergum with lateral margin overlying ster-
num, not fused to it; first sternum de-
pressed, especiallv posteriorly, separated
from second by a deep constriction (72-
76). Second segment not apically con-
stricted (77). Female with seventh tergum
hidden, reduced but continuously sclero-
tized anteriorly; no articulation within
section 1 of gonocoxite IX; valve present
on gonapophysis VlII; gonapophysis IX
considerably curved, especially basally,
with apex directed downward (78-81).
Male with seventh sternum not reduced
nor hidden; hypopygium a single up-
curved spine not concealed basally; cercus
present; gonapophyses IX with notal fu-
sion extensive (82-86).

Final instar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (88).

Behavioral formula — VaPaOC (89-92).

10. Tiphiid group.

y Material examined: Krombeinia nayarita
(6,2); Neotiphia rostrata (6,9); Para-
tiphia aequalis (6,9); P. neomexicana
(6,9); Tiphia letalis (6,9); T. popil-
liavora (6,9). ("Tiphiinae" of the
"Tiphiidae'')

(Data also from Allen & Jaynes, 1930;
Clausen & Gardner, 1932; Pate, 1947a;
Allen & Krombein, 1964; Evans, 1965;
Nagy, 1967; Allen, 1972.)

General — Sexual dimorphism slight, male
only a little more slender than female;
both sexes fully winged (1-2). No sterile
caste (3). All pubescence simple (4).

Head — Clypeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin shallowly sinuate,
not protuberant; scattered pores without
evident setae (7-9). Three ocelli present

(10). Genal secretory organ absent (11).
Antenna with 13 segments in male, 12
segments in female; axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin shallowly arcuate; ante-
rior "collar" well-developed; posterolateral
angle very slightly dorsally produced ante-
rior to tegula, thus appearing truncate but
reaching tegula; posteroventral margin
somewhat concave; ventral angle more or
less acute but not produced (18-23). Pro-
pleura posteriorly contiguous mesally (24).
Prosternum sunken except for anterior
rim (25). Forecoxae contiguous (26).
Mesonotum not anteriorly produced; scu-
tellum not enlarged (27-28). Prepectus not
fused to mesepisternum, discontinuous
midventrally, the halves considerably sep-
arated (29). Mesepimeron running the
height of mesopleuron (30). Mesosternum
with platelike projections posteromesally
over coxal cavities (31). Mid-coxae some-
what separated (32). Meso- and meta-
pleura not mutually movable although
not fused {O). Metanotum not reduced
medially; metapostnotum not visible me-
dially, invaginated (34-35). Metepister-
num somewhat expanded anterior to en-
dophragmal pit so that pit is some distance
from anterior margin of metapleuron but
pleural sulcus is very indistinct; no meta-
pleural gland (36-37). Metasternum de-
pressed anterolaterally to accommodate
mid-coxae, raised posteromesally, well-dif-
ferentiated from mesosternum, not anteri-
orly produced (38-40). Hind coxae slightly
separated (41). Metathoracic-propodeal
pleural suture not distinguishable (42).
Propodeum not shortened; disc and de-
clivity distinctly separated by a transverse
carina (43-44). Forewing with venation

552

The University of Kansas Science Bulletin

reduced apically; nine closed cells; vein S
incomplete just distal to separation from
M (45-46). Pterostigma fairly large and
sclerotized (47-48). Hind wing with vena-
tion reduced apically; three closed cells;
veins A and J absent; cu-e originating dis-
tal to separation of M and Cu; free section
of Cu present although reduced to a spur
(49-53). Basal hamuli dispersed along
margin, not clustered (54). Plical lobe
marked by a moderate incision; long jugal
lobe marked by an incision (55-56). Legs
fairly slender, little modified (stouter in
female with femora and tibiae enlarged);
arolia well-developed; claws toothed (57-
59). Foretibia with single calcar almost
straight, not highly modified (60). Mid-
tibia with scattered moderate spines; two
simple spurs (spurs slightly dorsally flat-
tened in female) (61, 63-65). Hind coxa
with weak dorsal carina basally (66).
Hind tibia with scattered moderate spines;
two simple spurs (spurs slightly dorsally
flattened in female) (62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment slightly constricted apically, not pet-
iolate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum posteriorly depressed, separated
from second by a deep constriction (72-
76). Second segment not apically con-
stricted (77). Female with seventh ter-
gum hidden, reduced but continuously
sclerotized anteriorly; no articulation
within section 1 of gonocoxite IX; valve
present on gonapophysis VIII; gonapo-
physis IX considerably curved, especially
basally, with apex directed downward (78-
81). Male with seventh sternum reduced
and hidden; hypopygium a single up-
curved spine, almost entirely concealed;
cercus absent; gonapophyses IX with notal
fusion extensive (82-86).

Final instar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (88).

Behavioral formula — VaPaO (89-92).

1 1 . Brack ycistidid group .

Material examined: Acanthetropis spp.
( i ; Arizona, New Mexico) ; Aglyptacros
siilcatiis ( ? ) ; Brachycistis alcanor ( <5 ) ;
B. aliitacea ( ? ) ; B. emarginata ( 5 ) ;
B. inaeqiialis { i )\ Bruesiella sp. ( 9 ,
Arizona) ; Coloastis spp. ( i ; Arizona,
New Mexico) ; Oiiemaya perpnnctata
{£). ("Brachycistidinae" of the "Tiphi-
idae")

(Data also from Mickel & Krombein,
1942; Pate, 1947a; Wasbauer, 1966, 1968;
Krombein, 1967.)

General — Sexual dimorphism considera-
ble although male not more slender than
female; male winged, female apterous
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size (very
short in female); antennae inserted fairly
low on face (5). Antennal socket simple,
without tubercle (6). Eye broadly oval,
almost round, with inner margin shal-
lowly sinuate, somewhat protuberant, scat-
tered pores with short setae (eye reduced,
oval with inner margin convex, flattened,
glabrous, without pores in female) (7-9).
Three ocelli present (absent in female)
(10). Genal secretory organ absent (11).
Antenna with 13 segments in male, 12
segments in female; axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex not elon-
gated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin shallowly arcuate
(deeply arcuate in male) ; anterior "collar"
well-developed; posterolateral angle very
slightly dorsally produced anterior to
tegiila thus appearing truncate but reach-
ing tegula; posteroventral margin some-

Phylogeny and Classiiication of the Aculeate Hymenoptera

553

what concave; vciiiral anijle rounded, al-
most acute hut not produced (18-23).
Propleura posteriorly contiguous mesally
(24). Prosternum sunken except for an-
terior rim (25). Forecoxae very nearly
contiguous (26). Mesonotum somewhat
anteriorly produced; scutellum not en-
larged (27-28). Prepectus not fused to
mesepisternum, discontinuous midven-
trally, the halves considerably separated
and much reduced in length (29). Mese-
pimeron running the height of meso-
pleuron (30). Mesosternum with platelike
projections posteromesally over coxal cavi-
ties (31). Mid-coxae somewhat separated
(32). Meso- and metapleura not mutually
movable although not fused (5i). Meta-
notum not reduced medially; metapost-
notum barely visible medially, invaginated
(34-35). Metepisternum somewhat ex-
panded anterior to endophragmal pit so
that pit is some distance from anterior
margin of metapleuron but pleural sulcus
is very indistinct; no metapleural gland
(36-37). Metasternum depressed anterolat-
erally to accommodate mid-coxae, slightly
raised posteromesally, well-differentiated
from mesosternum, not anteriorly pro-
duced (38-40). Hind coxae contiguous
(41). Metathoracic-propodeal pleural su-
ture visible dorsal to endophragmal pit,
barely discernible ventrally (42). Pro-
podeum not shortened; disc and declivity
fairly distinct (43-44). Forewing with
venation reduced apically; ten closed cells
(45-46). Pterostigma large and sclerotized
(47-48). Hind wing with venation re-
duced apically; three closed cells; veins A
and J absent; cu-e originating distal to
separation of M and Cu; free section of
Cu present although reduced to a spur
(49-53). Basal hamuli absent (54). Plical
lobe indicated by a shallow incision; mod-
erate jugal lobe present (55-56). Legs
slender, generally unmodified (stouter in
female with femora and tibiae enlarged);
arolia well-developed; claws simple (57-

59). Foretibia with single calcar almost
straight, not highly modified (58). Mid-
tibia with scattered moderate spines; a
single simple spur (spur slightly dorsally
Battened in female) (61, 63-65). Hind
coxa with weak dorsal carina basally (66).
Hind tibia with scattered moderate spines;
two simple spurs (spurs slightly dorsally
flattened in female) (62, 64, 67-68). Mod-
ified mesosoma of female with pronotum
large, articulating with mesothorax, its
hind margin almost straight; mesonotum
short; scuto-scutellar suture obliterated;
separate prepectal sclerite present although
much reduced; mesepimeron not differ-
entiated from mesepisternum externally;
meso- and metapleura articulating; meta-
notum extremely short and transverse,
fused to propodeum, essentially not dis-
cernible; metapleuron greatly shortened
anterior and dorsal to endophragmal pit;
metathoracic-propodeal notal and pleural
sutures discernible dorsal to pit, com-
pletely obliterated ventrally (69).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment slightly constricted apically (more
so in male), not petiolate; first tergum
with lateral margin overlying sternum,
not fused to it; first sternum entirely de-
pressed, separated from second by a deep
constriction (72-76). Second segment not
apically constricted (77). Female with
seventh tergum hidden, reduced but com-
pletely sclerotized anteriorly; no articula-
tion within section 1 of gonocoxite IX;
valve present on gonapophysis VIII ; gona-
pophysis IX considerably curved, espe-
cially basally, with apex directed down-
ward (78-81). Male with seventh sternum
reduced and hidden; hypopygium a single
upcurved spine, mostly concealed; cercus
absent; gonapophyses IX with notal fusion
extensive (82-86).

Larva — Unknown.

Behavioral formula — Unknown.

554

The University of Kansas Science Bulletin

12. Sapygid group.

Material examined: Fedtschenl{ia anthra-
cina ((5,9); Polochriim repandum ( <5 ) ;
Sapyga davicornis ( ? ) ; 5. martinu ( 5 ) ;
Sapyga sp. ( ? , Colorado). ("Sapygidae")
(Data also from Pate, 1947c; Bradley,
1955; Oeser, 1961; Guiglia, 1963, 1969;
Tobias, 1965; Bohart & Schuster, 1972;
Torchio, 1972.)

General — Sexual dimorphism minimal,
both sexes of similar form and fully
winged (1-2). No sterile caste (3). All
pubescence simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim some-
what raised, forming small tubercle (6).
Eye oval with inner margin very shal-
lowly sinuate, not protuberant; scattered
pores without evident setae (7-9). Three
ocelli present (10). Genal secretory organ
absent (11). Antenna with 13 segments
in male, 12 segments in female; axis of
radicle not deviating much from that of
scape; radicle not set into basal invagina-
tion of scape (12-14). Labio-maxillary
complex slightly elongated, with glossa
slightly produced and broadened; maxil-
lary palpus 6-segmented; labial palpus 4-
segmented (15-17).

Mesosoma — Pronotum large, not much re-
duced mesally, freely articulating with
mesothorax; hind margin shallowly arcu-
ate; anterior "collar" well-developed; pos-
terolateral angle rounded, reaching teg-
ula; posteroventral margin approximately
straight; ventral angle rounded, not pro-
duced (18-23). Propleura posteriorly con-
tiguous mesally (24). Prosternum sunken
except for fairly large anterior rim (25).
Forecoxae very nearly contiguous (26).
Mesonotum not anteriorly produced; scu-
tellum not enlarged (27-28). Prepectus not
much reduced although discontinuous
midventrally, the halves considerably sepa-
rated, fused to mesepisternum, the line of

fusion sulcate (29). Mesepimeron running
the height of mesopleuron (30). Meso-
sternum essentially simple but with weak
transverse carinae anteromesal to coxal
cavities (31). Mid-coxae slightly separated
(32). Meso- and metapleura articulating
(33). Metanotum not reduced medially;
metapostnotum barely discernible medi-
ally, invaginated (34-35). Endophragmal
pit close to anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum entirely depressed, well-dif-
ferentiated from mesosternum, not ante-
riorly produced (38-40). Hind coxae
contiguous (41). Metathoracic-propodeal
pleural suture entirely visible although
less distinct dorsal to endophragmal pit
(42). Propodeum not shortened; disc and
declivity not distinct but merging (43-44).
Forewing with venation not much re-
duced apically; ten closed cells (45-46).
Pterostigma fairly large and sclerotized
(47-48). Hind wing with venation not
much reduced apically; two closed cells;
vein C reduced distally; veins A and J
absent; origin of cu-e probably distal to
separation of M and Cu, not absolutely
determinable; free section of Cu absent
(49-53). Basal hamuli dispersed along
margin, not clustered (54). Plical lobe
barely indicated; short jugal lobe present
(55-56). Legs slender, generally unmod-
ified; arolia well-developed; claws toothed
(57-59). Foretibia with single calcar al-
most straight, not highly modified (60).
Mid-tibia with scattered moderate spines;
two simple spurs (spurs flattened and
marginally serrate in female) (61, 63-65).
Hind coxa with well-developed dorsal
carina {66). Hind tibia with scattered
moderate spines; two simple spurs (spurs
flattened and marginally serrate in fe-
male) (62, 64, 67-68)\

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-
olate; first tergum with lateral margin

Phylogeny and Classification of the Aculeate Hymenoptera

555

overlying sternum, not fused to it; first
sternum not depressed, not highly differ-
entiated from second (72-76). Second seg-
ment not apically constricted (77). Female
with seventh tergum hidden, reduced but
continuously sclerotized anteriorly; no ar-
ticulation within section 1 of gonocoxite
IX; no valve on gonapophysis VIII; gona-
pophysis IX moderately curved with apex
directed obliquely downward (78-Sl).
Male with seventh sternum somewhat re-
duced but not hidden; hypopygium sim-
ple, unmodified, not hidden; cercus pres-
ent; gonapophyses IX with notum
desclerotized, gonapophyses linked by
membrane basally only (82-86).

Final ixstar larva — Mandible bidentate
(87). Ten pairs of fully developed spi-
racles (88) (data for Sapyga only).

Behavioral formula — VaPaO (89-92)

13. Myrmosid group.

Material examined: Myrmosa brad ley i
( 9 ) ; M. f rater (£ ); M. melanocephala
((?,?); M. nigriceps {i)\ M. noctnrna
( (5 ) ; M. iinicolor ( c^ , 9 ) ; Myrmosida
parvula (5,9); M. rntilans ( 9 ) ; Pro-
tomiitilla fnicrosotna ( 9 ). ("Myrmosi-
nae" of the "Tiphiidae")
(Data also from RischofT, 1915; Bridwell,
1920; Krombein, 1940; Pate, 1947a & b;
Suarez, 1960; Wasbauer, 1973; Brothers,
1974b.)

General — Sexual dimorphism considera-
ble although male not more slender than
female; male winged, female apterous
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim much
produced, forming well-developed tuber-
cle (6). Eye oval with inner margin very
shallowly sinuate, not protuberant; many
scattered pores with moderately long setae
(7-9). Three ocelli present (10). Genal

secretory organ absent (11). Antenna with
13 segments in male, 12 segments in fe-
male; axis of radicle perpendicular to that
of scape; radicle set into basal invagina-
tion of scape (12-14). Labio-maxillary
complex not elongated; maxillary palpus
6-segmented; labial palpus 4-segmented
(15-17).

Mesosoma — Pronotum large, not much re-
duced mesally, freely articulating with
mesothorax; hind margin shallowly arcu-
ate; anterior "collar" well-developed; pos-
terolateral angle rounded, reaching teg-
ula; postero ventral margin approximately
straight; ventral angle rounded, not pro-
duced (18-23). Propleura posteriorly con-
tiguous mesally (24). Prosternum sunken
except for anterior rim (25). Forecoxae
contiguous (26). Mesonotum barely an-
teriorly produced; scutellum not much
enlarged (27-28). Prepectus not much re-
duced although discontinuous midven-
trally, the halves considerably separated,
fused to mesepisternum, the line of fusion
only marked by two pits ventrally (29).
Mesepimeron running the height of meso-
pleuron (30). Mesosternum essentially
simple but with weak transverse carinae
anteromesal to coxal cavities (31). Mid-
coxae very nearly contiguous (32).
Meso- and metapleura articulating {ii).
Metanotum not reduced medially; meta-
postnotum barely discernible medially, in-
vaginated (34-35). Metepisternum and
metepimeron slightly expanded anterior
to endophragmal pit so that pit is shifted
slightly posteriorly; no metapleural gland
(36-37). Metasternum entirely depressed,
well-differentiated from mesosternum, not
anteriorly produced (38-40). Hind coxae
contiguous (41). Metathoracic-propodeal
pleural suture almost entirely obliterated,
position barely discernible (42). Propo-
deum not shortened; disc and declivity
not distinct but merging (43-44). Fore-
wing with venation not much reduced
apically; ten closed cells (45-46). Ptero-

556

The University of Kansas Science Bulletin

stigma fairly large and sclerotized (47-48).
Hind wing with venation not much re-
duced apically; two closed cells; vein C
reduced distally; veins A and J absent;
origin of cu-e probably distal to separation
of M and Cu, not absolutely determinable;
free section of Cu absent (49-53). Basal
hamuli clustered near point of separation
of veins C and SC+R+S (54). Plical lobe
barely indicated; short jugal lobe present
(55-56). Legs slender, unmodified; arolia
well-developed (reduced in female); claws
toothed (simple in female) (57-59). Fore-
tibia with single calcar almost straight,
not highly modified (60). Mid-tibia with
scattered moderate spines; two simple
spurs (spurs flattened and marginally ser-
rate in female) (61, 63-65). Hind coxa
with lamellate dorsal carina (66). Hind
tibia with scattered moderate spines; two
simple spurs (spurs flattened and mar-
ginally serrate in female) (62, 64, 67-6S).
Modified mesosoma of female with pro-
notum large, articulating with mesotho-
rax; prepectal sclerite fused to mesepister-
num; meso- and metathorax and propo-
deum fused completely dorsally; pleura
with only meso-metapleural suture faintly
visible, flattened (69).

Metasoma — No specialized "felt lines"
(70). Stridulitrum present as a single
small shield-shaped area medially at base
of third tergum (71). First segment nei-
ther apically constricted nor petiolate
(slightly constricted in male); first tergum
with lateral margin overlying sternum,
not fused to it; first sternum posteriorly
depressed, separated from second by a
deep constriction (72-76). Second segment
not apically constricted (77). Female with
seventh tergum hidden, reduced but con-
tinuously sclerotized anteriorly; no articu-
lation within section 1 of gonocoxite IX;
valve present on gonapophysis VIII; gona-
pophysis IX considerably curved, espe-
cially basally, with apex directed down-
ward (78-81). Male with seventh sternum

not reduced nor hidden; hypopygium sim-
ple but concealed; cercus present; gonapo-
physes IX with notum desclerotized, gona-
pophyses linked by membrane basally only

(82-86).

Final instar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (88)
(pers. obs.).

Behavioral formula — VaPaOC (89-92).

14. Mutillid group.

Material examined: Specimens of ap-
proximately 150 genera and subgenera.
(See section on Mutillidae below, Table
III, for complete list.) ("Pseudophotop-
sidinae - Sphaeropthalminae - Mutillinae -
Rhopalomutillinae" of the "Mutillidae")
(Data also from Bischoff, 1920-21; Mickel,
1928; Schuster, 1947, 1949; Brothers, 1972;
Debolt, 1973.)

General — Sexual dimorphism considera-
ble, although male not more slender than
female; male winged, female apterous
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim much
produced, forming well-developed tuber-
cle (6). Eye oval with inner margin very
shallowly sinuate, not protuberant; scat-
tered sensory pores without evident setae
(7-9). Three ocelli present (10). Genal
secretory organ absent (11). Antenna
with 13 segments in male, 12 segments in
female; axis of radicle perpendicular to
that of scape; radicle set into basal invagi-
nation of scape (12-14). Labio-maxillary
complex not elongated; maxillary palpus
6-segmented; labial palpus 4-segmented
(15-17).

Mesosoma — Pronotum fairly large, some-
what reduced mesally; articulation with
mesothorax slightly restricted but with no

PlIYLOGENY AND CLASSIFICATION' OF THE AcULEATE HyMENOPTERA

557

fusion; hind margin concave; anterior
"collar" well-developed; posterolateral an-
gle very slightly dorsally produced ante-
rior to tegula, thus appearing truncate but
reaching tegula; posteroventral margin
approximately straight; ventral angle
rounded, not produced (18-23). Propleura
posteriorly contiguous mesally (24). Pro-
sternum sunken except for anterior rim
(25). Forecoxae contiguous (26). Meso-
notum only slightly anteriorly produced;
scutellum not enlarged (27-2S). Prepectus
not much reduced although discontinuous
midventrally, the halves considerably sepa-
rated, fused to mesepisternum, the line of
fusion marked by two pits ventrally (29).
Mesepimeron running the height of meso-
pleuron although poorly differentiated
ventrally (30). Mesosternum essentially
simple but with weak transverse cari-
nae anteromesal to coxal cavities (31).
Mid-coxae very nearly contiguous (32).
Meso- and metapleura articulating (33).
Metanotum slightly reduced medially;
metapostnotum barely discernible medi-
ally, invaginated (34-35). Metepisternum
and metepimeron slightly expanded ante-
rior to endophragmal pit so that pit is
shifted slightly posteriorly; no metapleural
gland (36-37). Sternum entirely depressed
but slightly ventrally produced antero-
mesal to coxal cavities, well-difFerentiated
from mesosternum, not anteriorly pro-
duced (38-40). Hind coxae contiguous
(41). Metathoracic-propodeal pleural su-
ture almost entirely obliterated, position
barely discernible (42). Propodeum not
shortened; disc and declivity not distinct
but merging (43-44). Forewing with ve-
nation reduced apically; ten closed cells
(45-46). Pterostigma fairly large and
sclerotized (47-48). Hind \\ing with vena-
tion reduced apically; two closed cells;
vein C reduced distally; veins A and J
absent; cu-e apparently originating distal
to separatifjn of M and Cu; free section
of Cu apparently absent (49-53). Basal

hamuli clustered near point of separation
of veins C and SC+R+S (54). Plical lobe
barely indicated; extremely short jugal
lobe present (55-56). Legs slender, un-
modified; arolia well-developed; claws
toothed (57-59). Foretibia with single cal-
car almost straight, not highly modified
(60). Mid-tibia with scattered moderate
spines; two simple spurs (spurs flattened
and serrate in female) (61, 63-65). Hind
coxa with weak dorsal carina basally (66).
Hind tibia with scattered moderate spines;
two simple spurs (spurs flattened and ser-
rate in female) (62, 64, 67-68). Modified
mesosoma of female with pronotum fused
to mesothorax but with suture visible; pre-
pectus fused to mesepisternum; meso- and
metathorax fused completely dorsally, with
pleurointersegmental suture faintly visible
laterally; metathoracic-propodeal notal and
pleural sutures visible only above endo-
phragmal pit; pleura flattened (69).

Metasoma — Tendency toward develop-
ment of felted secretory areas ("felt lines")
on second segment, on both tergum and
sternum laterally (70). Stridulitrum pres-
ent as a small shield-shaped area medially
at base of third tergum (71). First seg-
ment neither apically constricted nor peti-
olate (slightly constricted in male); first
tergum with lateral margin overlying ster-
num, not fused to it; first sternum pos-
teriorly depressed, separated from second
by a deep constriction (72-76) . Second seg-
ment not apically constricted (77). Female
with seventh tergum hidden, reduced but
continuously sclerotized anteriorly; no ar-
ticulation within section 1 of gonocoxite
IX; valve present on gonapophysis VIII;
gonapophysis IX considerably curved,
especially basally, with apex directed
downward (78-81). Male with seventh
sternum somewhat reduced but not hid-
den; hypopygium simple, unmodified, not
hidden; cercus present; gonapophyses IX
with notum desclerotized, gonapophyses
linked by membrane basally only (82-86).

558

The University of Kansas Science Bulleth

Final instar larva — Mandible quadriden-
tate (87). Nine pairs of fully developed
spiracles, i.e., second pair reduced (S8).

Behavioral formula— VaP..,OC (89-92) .

15. Sierolomorphid group.

Material examined: Sierolomorphu

} apache {$)\ S. canadensis ( (? , 9 ) ; S.

nigrescens ( <? , 9 ) ; S. similis ( 6 ) .

("Sierolomorphidae")

(Data also from Evans, 1961; Nagy, 1971.)

General — Sexual dimcjrphism minimal,
both sexes of similar form and fullv
winged (1-2). No sterile caste (3). All
pubescence simple (4).

Head— Clypeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple but with dorsomesal rim
very slightly raised, no true tubercle (6).
Eye oval with inner margin vcrv shallowly
sinuate, not protuberant; scattered pores
with short setae (7-9). Three ocelli pres-
ent (10). Genal secretory organ absent
(11). Antenna with 13 segments in male,
12 segments in female: axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum large, not reduced
mesally, freely articulating with mesotho-
rax; hind margin shallowly arcuate (some-
what more deeply so in male); anterior
"collar" well-developed; posterolateral an-
gle rounded, reaching tegula; posteroven-
tral margin slightly concave; ventral angle
rounded, not produced (18-23). Propleura
posteriorly contiguous mesally (24). Pro-
sternum sunken except for anterior rim
(25). Forecoxae contiguous (26). Meso-
notum not anteriorly produced (very
slightly produced in male); scutellum not
enlarged (27-28). Prepectus not fused
to mesepisternum, discontinuous midven-
trally, the halves considerably separated

(29). Mesepimeron running the height of
mesopleuron (30). Mesosternum essen-
tially simple but with weak ridges antero-
mesal to coxal cavities (31). Mid-coxae
slightly separated (32). Metanotum not
reduced medially; metapostnotum visible
medially, slightly depressed but not in-
vaginated (33-34). Metepimeron slightly
expanded anteroventral to endophragmal
pit so that pit is a short distance from
anterior margin of metapleuron; no meta-
pleural gland (35-36). Metasternum en-
tirely depressed except for slight ridges
anterior to coxae, well-differentiated from
mesosternum, not anteriorly produced (38-
40). Hind coxae contiguous (41). Meta-
thoracic-propodeal pleural suture barely
discernible dorsal to endophragmal jiit,
not visible ventrally (42). Propodeum not
shortened; disc and declivity somewhat
distinct although merging (43-44). Fore-
wing with venation reduced apically;
seven closed cells; vein S absent between
its separation from M and junction with
r-s; crossveins 2s-m and 2m-cu absent (45-
46). Pterostigma large and sclerotized
(47-48). Hind wing with venation re-
duced apically; two closed cells; vein C
reduced distally; veins A and J absent;
cu-e originating proximal to separation of
M and Cu; free section of Cu present
(49-53). Basal hamuli dispersed along mar-
gin although tending toward concentra-
tion at separation of vein SC+R+S from
margin (54). Plical lobe barely indicated;
jugal lobe absent (55-56). Legs slender,
generally unmodified; arolia well-devel-
oped; claws toothed (57-59). Foretibia
with single calcar almost straight, not
highly modified (60). Mid-tibia with scat-
tered weak spines; two simple spurs (61,
63-65). Hind coxa with weak dorsal carina
(66). Hind tibia with scattered weak
spines; two simple spurs (62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-

Phylogeny and Classification of the Aculeate Hymenoptera

559

elate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, not highly differ-
entiated from second (72-76). Second seg-
ment not apically constricted (77). Female
with seventh tergum hidden, reduced but
continuously sclerotized anteriorly; no ar-
ticulation within section 1 of gonocoxite
IX; valve present on gonapophysis VIII;
gonapophysis IX gently curved, almost
straight, with apex only weakly directed
obliquely downward (78-81). Male with
seventh sternum not reduced nor hidden;
hypopygium reduced, very narrow, hid-
den except for extreme apex; cercus pres-
ent; gonapophyses IX with notal fusion
well-developed, especially basally (82-86).

Larva — Unknown.

Behavioral formula — Unknown.

16. Pompilid group.

Material examined: Anopliiis amethysti-
niis ( (i , 9 ) ; C crop ales maculata ((5,9);
Chirodatyjniis pyrrhomelas (9); Crypto-
cheilits severini ((5,9); Episyron qiiin-
quenotatiis ((5,9); Eiiclavelia jasciata
( 9 ) ; Evagetes parvus ((5,9); Paracy-
phononyx junereus ((5,9); Pepsis thisbe
((5,9); Poeciloponipilus interruptus ( $ ,
9 ) ; Pompilus seeks t us ((5,9); P sort h as-
pi s connexa ((5,9); Tachypompilus fer-
rugineus ((5,9). ("Pompilidae")
(Data also from Evans, 1953, 1959b, 1972;
Daly, 1955.)

General — Sexual dimorphism minimal,
male slightly smaller than female but of
similar form; both sexes fully winged
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket simple, without tubercle (6).
Eye oval with inner margin shallowly
sinuate, not protuberant; scattered pores
without evident setae (7-9). Three ocelli
present (10). Genal secretory organ ab-

sent (11). Antenna with 13 segments in
male, 12 segments in female; axis of rad-
icle not deviating much from that of
scape; radicle not set into basal invagina-
tion of scape (12-14). Labio-maxillary
complex not elongated; maxillary palpus
6-segmented; labial palpus 4-segmented
(15-17).

Mesosoma — Pronotum large, not mesally
reduced, freely articulating with mesotho-
rax; hind margin arcuate; anterior "col-
lar" well-developed; posterolateral angle
rounded, reaching tegula; posteroventral
margin slightly concave; ventral angle
rounded, not produced (18-23). Propleura
posteriorly contiguous mesally (24). Pro-
sternum sunken except for anterior rim
(25). Forecoxae contiguous (26). Meso-
notum not anteriorly produced; scutellum
not enlarged (27-28). Prepectus not much
reduced although discontinuous midven-
trally, the halves considerably separated,
fused to mesepisternum, the line of fusion
retained as a sulcus (29). Mesepimeron
running the height of mesopleuron (30).
Mesosternum essentially simple but with
weak ridges anteromesal to coxal cavities
(31). Mid-coxae contiguous (32). Meso-
and metapleura not mutually movable al-
though not fused {H). Metanotum not
reduced medially; metapostnotum well-
developed, visible medially, slightly de-
pressed but not invaginated (34-35). Mete-
pimeron slightly expanded anteroventral
to endophragmal pit so that pit is a short
distance from anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum entirely depressed except for
slight ridges anterior to coxae, well differ-
entiated from mesosternum, not anteriorly
produced (38-40). Hind coxae contiguous
(41). Metathoracic-propodeal pleural su-
ture entirely visible (42). Propodeum not
shortened; disc and declivity somewhat
distinct although merging (43-44). Fore-
wing with venation not reduced apically;
ten closed cells (45-46). Pterostigma fairly

560

The University of Kansas Science Bulletin

small with sclerotization reduced (47-48).
Hind wing with venation not reduced
apically; two closed cells; vein C reduced
distally; veins A and J absent; cu-e origi-
nating proximal to separation of M and
Cu; free section of Cu present (49-53).
Basal hamuli strongly concentrated into
a basal cluster at separation of vein
SC+R+S frf)m margin (54). Plical lobe
barely indicated; moderate jugal lobe pres-
ent (55-56). Legs slender, generally un-
modified although somewhat elongated;
arolia well-developed; claws toothed (57-
59). Foretibia with single calcar almost
straight, not highly modified (60). Mid-
tibia with scattered moderate spines; two
spurs slightly dorsally flattened (61, 63-65).
Hind coxa simple (66). Hind tibia with
scattered moderate spines; two spurs
slightly dorsally flattened, the inner one
modified as a calcar by development of
basal brush of setae dorsally (62, 64, 67-6(S).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-
olate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed, not highly differ-
entiated from second (72-76). Second seg-
ment not apically constricted (77). Female
with seventh tergum hidden, reduced but
continuously sclerotized anteriorly; no ar-
ticulation within section 1 of gonocoxite
IX; no valve on gonapophysis VIII;
gonapophysis IX gently curved, almost
straight, with apex only weakly directed
obliquely downward (78-81). Male with
seventh sternum reduced but not hidden;
hypopygium essentially simple, not re-
duced nor hidden; cercus present; gonapo-
physes IX with notal fusion well-devel-
oped (82-86).

Final instar larva — Mandible tridentate
(87). Nine pairs of fully developed spi-
racles, i.e., second pair reduced (88).

Behavioral formula— VaPaTaOC (89-92).

17. Rhopalosomatid group.

Material examined: Olixon ban\sii (5,
9 ) ; 0. testaceum ( 2 ) ; ? Olixon sp. {$ ,
Cameroon); Rhopalosoma spp. (9; Ma-
dagascar, Panama). ("Rhopalosomatidae")
(Data also from Perkins, 1908; Reid, 1939;
Brues, 1943; Gurney, 1953.)

General — Sexual dimorphism minimal,
male slightly smaller than female but of
similar form; both sexes fully winged
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket simple, without tubercle (6).
Eye oval with inner margin shallowly
sinuate, not protuberant; scattered pores
with short setae (7-9). Three ocelli pres-
ent (10). Genal secretory organ absent
(11). Antenna with 13 segments in male,
12 segments in female; axis of radicle not
deviating much from that of scape; radi-
cle not set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum somewhat reduced
mesally, fairly freely articulating with
mesothorax; hind margin concave; ante-
rior "collar" fairly well-developed; pos-
terolateral angle rounded, very slightly
produced dorsally and thus truncate al-
thotigh reaching tegula; posteroventral
margin slightly concave; ventral angle
rounded, not produced (18-23). Propleura
posteriorly contiguous mesally (24). Pro-
sternum sunken except for anterior rim
(25). Forecoxae contiguous (26). Meso-
iKjtum somewhat anteriorly produced;
scutellum not enlarged (27-28). Prepectus
not fused to mesepisternum, much re-
duced, short and tall, hidden under pro-
notum, discontinuous midventrally, the
halves considerably separated (29). Mese-
pimeron running the height of meso-
pleuron (30). Mesosternum with platelike

Phylogeny and Classification of the Aculeate Hymenoptera

561

projections posteromesally over coxal cavi-
ties (31). Mid-coxae contiguous (32).
Meso- and metapleura articulating (33).
Metanotum not reduced medially, meta-
postnotum visible medially, depressed but
not invaginated (34-35). Metepimeron
somewhat expanded anteroventral to en-
dophragmal pit so that pit is some distance
from anterior margin of metapleuron; no
metapleural gland (36-37). Metasternum
entirely depressed, well-differentiated from
mesosternum, not anteriorly produced (38-
40). Hind coxae contiguous (41). Meta-
thoracic-propodeal pleural suture entirely
visible (42). Propodeum not shortened;
disc and declivity not distinct but merg-
ing (43-44). Forewing with venation
slightly reduced apically; eight closed cells,
including cell C which is much reduced
by fusion of veins C and SC4-R4-S except
distally; crossveins 2s-m and 2m-cu absent
(45-46). Pterostigma small with sclerotiza-
tion reduced (47-48). Hind wing with
venation slightly reduced apically; two
closed cells; vein C reduced distally; veins
A and J absent; cu-e originating proximal
to separation of M and Cu; free section of
Cu present (49-53). Basal hamuli strongly
concentrated into a basal cluster at separa-
tion of vein SC+R4-S from margin (54).
Plical lobe well-marked by a deep incision ;
moderate jugal lobe present (55-56). Legs
slender, little modified (tarsi flattened and
broadened, forefemur swollen and arolia
enlarged in female); arolia well-devel-
oped; claws toothed (57-59). Foretibia
with single calcar almost straight, not
highly modified (60). Mid-tibia with scat-
tered weak spines; two simple spurs (61,
63-65). Hind coxa simple (66). Hind
tibia with scattered weak spines; two sim-
ple spurs, the inner one modified as a
calcar by development of basal brush of
setae dorsally (62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment apically somewhat constricted, not

petiolate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum not depressed although fairly
well-diiTerentiated from second (72-76).
Second segment not apically constricted
(77). Female with seventh tergum hid-
den, reduced but continuously sclerotized
anteriorly; no articulation within section
1 of gonocoxite IX; valve present on gona-
pophysis VIII; gonapophysis IX gently
curved, with apex directed obliquely up-
ward (78-81). Male with seventh sternum
reduced but not hidden; hypopygium es-
sentially simple, not reduced nor hidden;
cercus present; gonapophyses IX with
notal fusion well-developed (82-86).

Final instar larva — Mandible essentially
unidentate (inner margin serrate) (87).
Ten pairs of fully developed spiracles (88).

Behavioral formula— Va (Pa = ) O (89-92) .

18. Formicid group.

Material examined: Aphuenogaster jiilva
((^,9,5); Atta cephalotes ( c5 , 9 , ^ ) ;
Camponotiis herciileanits ( c? , ? , ^ ) ; Cre-
matogaster coarctata ( c^ , 9 , 5 ) ; Dinopo-
nera sp. ( 5 , Argentina) ; Eciton vagans
( (5 , ^ ) ; Ectatomma tuberculatum (5,9,
^ ) ; Labidus coecus ( 5 , ^ ) ; Liotneto-
piutn apiculatum (5,9,5); Myrmecia
spp. ( 5, Queensland); Neoponera villosa
(9,5); Nomamyrmex crassicomis ( $ ,

5 ) ; Odontomachus haematoda (5,9);
Paltothyreus sp. (5,9,5; Mozambique) ;
Paraponera clauata (5,9,5); Pogono-
myrmex barbatus (5,9,5). ("Formici-
dae")

(Data also from Haskins & Ensman,
1938; Brown & Nutting, 1950; Creighton,
1950; Haskins & Haskins, 1950; Wheeler

6 Wheeler, 1952, 1971; Brown, 1954, 1969;
Brown & Wilson, 1959; Hermann &
Blum, 1966; Wilson, Carpenter & Brown,
1967; Gotwald, 1969; Wilson, 1971; Markl,
1973.)

General — Sexual dimorphism moderate.

562

The University of Kansas Science Bulletin

male somewhat smaller but not much
more slender than female; both sexes fully
winged (1-2). Wingless sterile female
worker caste present (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted low on face (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin shallowly sinuate,
not protuberant; scattered pores with
short setae (7-9). Three ocelli present
(10). Genal secretory organ absent (11).
Antenna with 13 segments in male, 12
segments in female; axis of radicle not
deviating much from that of scape; radi-
cle set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma— Pronotum large, somewhat re-
duced mesally, fairly freely articulating
with mesothorax; hind margin deeply
arcuate; anterior "collar" well-developed;
posterolateral angle very slightly dorsally
produced posteriorly, thus appearing trun-
cate although reaching tegulae; postero-
ventral margin slightly concave; ventral
angle acute, somewhat mesally produced
(l(S-23). Propleura posteriorly contiguous
mesally (24). Prosternum sunken except
for anterior rim (25). Forecoxae contigu-
ous (26). Mesonotum slightly anteriorly
produced; scutellum not enlarged (27-28).
Prepectus fused to mesepisternum, not
distinguishable, apparently very short and
narrow, hidden under pronotum (29).
Mesepimeron reduced, distinguishable only
at dorsal extreme (30). Mesosternum sim-
ple, without posterior projections (31).
Mid-coxae slightly separated (32). Meso-
and metapleura fused, not mutually mov-
able (possibly not fused in Sphecomyrma)
(33). Metanotum not reduced medially;
metapostnotum not visible medially, in-
vaginated (34-35). Metepisternum and
metepimeron expanded anterior to endo-
phragmal pit so that pit is some distance

from anterior margin of metapleuron;
metapleural gland present, opening above
hind coxa (36-37). Metasternum entirely
depressed, well-differentiated from meso-
sternum, not anteriorly produced (38-40).
Hind coxae contiguous (41). Metatho-
racic-propodeal pleural suture entirely dis-
tinguishable (42). Propodeum not short-
ened; disc and declivity not distinct but
merging (43-44). Forewing with venation
not much reduced apically; eight closed
cells; veins 2s-m and 2m-cu absent (45-
46). Pterostigma moderate and sclerotized
(47-48). Hind wing with venation not
much reduced apically; two closed cells;
vein C reduced distally; veins A and J
absent; cu-e originating proximal to sepa-
ration of M and Cu; free section of Cu
present (49-53). Basal hamuli absent (54).
Plical lobe indicated by a shallow notch;
moderate jugal lobe present (55-56). Legs
slender, generally unmodified; arolia well-
developed; claws toothed (57-59). Fore-
tibia with single calcar almost straight, not
much modified (60). Mid-tibia with few
moderate spines apically only; two spurs,
the inner dorsally carinate-pectinate form-
ing a calcar, the outer simple (61, 63-65).
Hind coxa simple {bd). Hind tibia with
few moderate spines apically only; two
spurs, the inner dorsally carinate-pectinate
forming a calcar, the outer simple (62,
64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment somewhat apically constricted, not
petiolate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum slightly depressed, differentiated
from second by a slight constriction (72-
76). Second segment not apically con-
stricted (77). Female with seventh tergum
hidden, reduced but continuously sclero-
tized anteriorly; no articulation within
section 1 of gonocoxite IX; valve present
on gonapophysis VIII; gonapophysis IX
almost straight, with apex weakly directed

I'

Phylogeny and Classification of the Aculeate Hymenoptera

563

obliquely upward (78-81). Male with
seventh sternum not reduced nor hidden;
hypopygium simple, unmodified, not hid-
den; cercus present; gonapophyses IX
with notal fusion well-developed (82-86).

Final ixstar larva — Mandible tridentate
(87). Ten jxiirs of fully developed spi-
racles (88).

Behavioral formula— I [ lO" ( VaPaT„) ^ ] '
(89-92).

19. Scoliid group.

Material examined: Austroscolia nitida

( 5 ) ; Campsojyteris mar gin el I a ( S ,9 );

Scolia nobilitata ((5,5); Trielis octomac-

ttlatd (5,2); Trisciloa zonata ( <5 ) .

("Scoliidae")

(Data also from Betrem, 1928; Bradley,

1950; Grandi, 1961; luga, 1968.)

General — Sexual dimorphism slight, male
onlv slitjhtlv more slender than female;
both sexes fully winged (1-2). No sterile
caste (3). All pubescence simple (4).

Head — Clypeus somewhat enlarged; an-
tennae inserted moderately high on face
(somewhat higher in male) (5). Antennal
socket simple, without tubercle (6). Eye
oval with inner margin deeply incised, not
protuberant; pores and setae absent (7-9).
Three ocelli present (10). Genal secretory
organ absent (11). Antenna with 13 seg-
ments in male, 12 segments in female;
axis of radicle not deviating much from
that of scape; radicle not set into basal
invagination of scape (12-14). Labio-max-
illary complex somewhat elongated, with
both glossa and paraglossa produced; max-
illary palpus 6-segmented; labial palpus
4-segmented (15-17).

Mesosoma — Pronotum considerably re-
duced and short mesally, not freely mov-
able against mesothorax but tightly ap-
pressed to it; hind margin strongly
concave; anterior "collar" well-developed;
posterolateral angle slightly produced

above tegula, thus appearing notched but
reaching tegula; posteroventral margin
strongly concave; ventral angle acute,
somewhat produced (18-23). Propleura
posteriorly contiguous mesally (24). Pro-
sternum entirely sunken, without differ-
entiated anterior rim (25). Forecoxae
contiguous (26). Mesonotum enlarged
and anteriorly produced; scutellum not
enlarged (27-28). Prepectus fused to
mesepisternum, much reduced but tall,
discontinuous midventrally, the halves
considerably separated, hidden under pro-
notum (29). Mesepimeron running the
height of mesopleuron although pleural
sulcus much reduced and not differen-
tiated externally (30). Mesosternum flat,
laterally broadened, with small plate-like
projections over coxal cavities (31). Mid-
coxae very broadly separated (32). Meso-
and metapleura not mutually movable
although not fused {y:)). Metanotum not
reduced medially; metapostnotum not
visible medially, invaginated (34-35). En-
dophragmal pit a short distance from
anterior margin of metapleuron and
metepisternum posteriorly expanded; no
metapleural gland (36-37). Metasternum
large, laterally broadened, with small plate-
like projections over coxal cavities, little
differentiated from mesosternum but not
anteriorly produced (38-40). Hind coxae
broadly separated (41). Metathoracic-pro-
podeal pleural suture entirely distinct (42).
Propodeum not much shortened; disc and
declivity fairly distinct (43-44). Forewing
with venation reduced apically; ten closed
cells (45-46). Pterostigma reduced and
partially desclerotized (47-48). Hind wing
with venation reduced apically; three
closed cells; veins A and J absent; cu-e
originating proximal to separation of M
and Cu; free section of Cu present (49-
53). Basal hamuli absent (54). Plical lobe
not indicated; long jugal lobe indicated
by a deep incision (55-56). Legs fairly
slender, unmodified (stouter in female

564

The University of Kansas Science Bulletin

with femora and tibiae somewhat en-
larged); aroUa well-developed; claws sim-
ple (57-59). Foretibia with single calcar
slightly curved, posteriorly hollowed and
spatulate (60). Mid-tibia with scattered
stout spines; one spur, slightly dorsally
flattened (61, 63-65). Hind coxa simple
(66). Hind tibia with scattered stout
spines; two slightly flattened spurs (62,
64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment slightly apically constricted, not peti-
olate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum entirely depressed, highly differ-
entiated from the second by a marked
constriction (72-76). Second segment not
constricted apically (77). Female with
seventh tergum hidden, reduced but con-
tinuously sclerotized anteriorly; no articu-
lation within section 1 of gonocoxite IX;
valve present on gonapophysis VIII; gona-
pophysis IX moderately curved, with apex
directed obliquely downward (78-81).
Male with seventh sternum neither re-
duced nor hidden; hypopygium apically
produced as three subequal spines, almost
entirely concealed (except only for tips of
spines); cercus absent; gonapophyses IX
with notum desclerotized, gonapophyses
linked by membrane along most of their
length (82-86).

Final instar larva — Mandible tridentate
(87). Ten pairs of fully developed spi-
racles (88).

Behavioral formula — VaPa(?Ta)0 (89-
92).

20. Vespid group.

Material examined: Ancistroccrtis anti-
lope ((5,9); Belonogaster spp. ( $ ,
Uganda; 9, South Africa); Brachygastra
lecheguana (5,9); Ceramius bicolor ( $ ,
9 ) ; C. lichtensteinii ((3,9); Euinenes
jraterniis ((5,9); Eitodynents annidatus

((5,9); Eitparagia sciitellaris ((5,9); E.
Umberlakj ((5,9); Monobia quadridens
((5,9); Polistes me trie us ((5,9); ?oly-
bia sericea ( 9 ) ; Provespa barthelemyi
((5,9); Pseudomasaris vespoides (5,9);
Pteroeheilus quinquejasciatum (5,9);
Stenodynerus papagorutn (5,9); Sym-
morphus eanadensis (5,9); Synagris cor-
nuta ( 5 ) ; Synoeca surinama ( 9 ) ; Vespa
crabro (5,9); Vespula norvegicoides
(5,9); Zethiis tnatzicatzin (5,9).
("Masaridae," "Eumenidae," "Vespidae")
(Data also from Duncan, 1939; Reid,
1942; Grandi, 1961; Richards, 1962; Evans,
1966; Torchio, 1970; Charnley, 1973.)

General — Sexual dimorphism minimal,
male not much more slender than female;
both sexes fully winged (1-2). No sterile
caste (3). All pubescence simple (4).

Head — Clypeus somewhat enlarged; an-
tennae inserted moderately high on face
(5). Antennal socket simple, without tu-
bercle (6). Eye oval with inner margin
fairly deeply incised, not protuberant; scat-
tered pores without evident setae (7-9).
Three ocelli present (10). Genal secretory
organ absent (11). Antenna with 13 seg-
ments in male, 12 segments in female; axis
of radicle not deviating much from that
of scape; radicle not set into basal invagi-
nation of scape (12-14). Labio-maxillary
complex slightly elongated, with both
glossa and paraglossa somewhat produced;
maxillary palpus 6-segmented; labial pal-
pus 4-segmented (15-17).

Mesosoma — Pronotum extremely reduced
and short mesally, not freely movable
against mesothorax but tightly appressed
to it; hind margin strongly concave; an-
terior "collar" slightly reduced; postero-
lateral angle slightly produced above teg-
ula and exceeding its anterior margin, the
angle thus appearing acute; posteroventral
margin strongly concave; ventral angle
acute, slightly ventrally produced (18-23).
Propleura posteriorly contiguous mesally

I

Phylogeny and Classification of the Aculeate Hymenoptera

565

(24). Prosternum entirely sunken, with-
out differentiated anterior rim (25).
Forecoxae contiguous (26). Mesonotum
enlarged and anteriorly produced; scutel-
lum not enlarged (27-28). Prepectus not
fused to mesepisternum, much reduced,
very short, discontinuous midventrally,
the halves very considerably separated,
hidden under pronotum (29). Mes-
epimeron running the height of meso-
pleuron although much reduced (30).
Mesosternum essentially simple but with
anteromesal rims of coxal cavities acutely
produced posteriorly (31). Mid-coxae con-
tiguous (32). Meso- and metapleura not
mutually movable although apparently
not fused (ii). Metanotum not reduced
medially; metapostnotum not visible me-
dially, invaginated (34-35). Metepister-
num and metepimeron expanded anterior
to endophragmal pit so that pit is some
distance from anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum entirely depressed except for
anteromesal ridge, to accommodate mid-
coxae, well-diflferentiated from mesoster-
num, not anteriorly produced (38-40).
Hind coxae contiguous (41). Metatho-
racic-propodeal pleural suture entirely dis-
tinct (42). Propodeum shortened dorso-
medially so that disc virtually absent,
merging with declivity (43-44). Forewing
with venation not much reduced apically;
ten closed cells (45-46). Pterostigma fairly
small and slightly desclerotized (47-48).
Hind wins; with venation not much re-
duced apically; three closed cells; veins A
and J absent; cu-e originating proximal
to separation of M and Cu; free section
of Cu present (49-53). Basal hamuli ab-
sent (54). Plical lobe barely indicated;
moderate jugal lobe present (55-56). Legs
slender, unmodified; arolia well-devel-
oped; claws toothed (57-59). Foretibia
with single calcar almost straight, not
much modified (60). Mid-tibia with scat-
tered weak spines; two simple spurs (61,

63-65). Hind coxa simple (66). Hind
tibia with scattered weak spines; two
spurs, the inner carinate and fringed dor-
sally, forming a calcar, the outer simple
(62, 64, 67-68).

Metasoma — No specialized "felt lines" or
stridulatory apparatus (70-71). First seg-
ment neither apically constricted nor peti-
olate; first tergum with lateral margin
overlying sternum, not fused to it; first
sternum depressed, especially posteriorly,
highly differentiated from second by a
marked constriction (72-76). Second seg-
ment not apically constricted (77). Female
with seventh tergum hidden, reduced but
continuously sclerotized anteriorly; no
articulation within section 1 of gonocoxite
IX; no valve on gonapophysis VIII; gona-
pophysis IX almost straight, with apex
only weakly directed obliquely downward
(78-81). Male with seventh sternum re-
duced although not entirely hidden; hypo-
pygium simple, unmodified, not hidden;
cercus absent; gonapophyses IX with notal
fusion well-developed (82-86).

Final ixstar larva — Mandible tridentate
(87). Ten pairs of fully developed spi-
racles (88).

Behavioral formula— IO(VaPaTa)"C (89-
92).

21. Eotillid group.

Material examined: Eotilla 7nic\eU ( $ ,
9 ) ; £". SHperba {$). ("Eotillinae" of the
"Mutillidae")

(Data also from Schuster, 1949; Mickel,
1968.)

General — Sexual dimorphism considera-
ble, male not much more slender than fe-
male; male winged, female apterous (1-2).
No sterile caste (3). Much pubescence of
sub-plumose setae and plumose scales (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket simple, without tubercle (6).
Eye oval with inner margin very shallowly

566

The University of Kansas Science Bulletin

sinuate, not protuberant; scattered pores
with short setae (7-9). Three ocehi pres-
ent (absent in female) (10). Genal secre-
tory organ absent (11). Antenna with 13
segments in male, 12 segments in female;
axis of radicle not deviating much from
that of scape; radicle not set into basal
invagination of scape (12-14). Labio-max-
illary complex not elongated; maxillary
palpus 6-segmented; labial palpus 4-seg-
mented (15-17).

Mesosoma — Pronotum large, very little re-
duced medially, freely articulating with
mesothorax; hind margin arcuate; ante-
rior "collar" well-developed; posterolateral
angle rounded, reaching tegula; postero-
ventral margin approximately straight;
ventral angle rounded, not produced (18-
23). Propleura posteriorly contiguous mes-
ally (24). Prosternum sunken except for
anterior rim (25). Forecoxae nearly con-
tiguous (26). Mesonotum not much an-
teriorly produced; scutellum not enlarged
(27-28). Prepectus not fused to mesepi-
sternum (probably — not determinable
without dissection), discontinuous mid-
ventrally, the halves considerably sepa-
rated, very short and tall (29). Mese-
pimeron reduced, distinguishable only at
dorsal extreme (30). Mesosternum simple,
without posterior projections (31). Mid-
coxae somewhat separated (32). Meso-
and metapleura not mutually movable
although not fused (ii). Metanotum not
reduced medially; metapostnotum not vis-
ible medially, invaginated (34-35). Mete-
pimeron very slightly expanded antero-
ventral to endophragmal pit so that pit
is a short distance from anterior margin
of metapleuron; no metapleural gland
(36-37). Metasternum depressed only lat-
erally to accommodate mid-coxae, mesally
little differentiated from mesosternum and
slightly anteriorly produced (38-40). Hind
coxae contiguous (41). Metathoracic-pro-
podeal pleural suture visible dorsal to
endophragmal pit, not distinguishable ven-

trally (42). Propodeum not shortened;
disc and declivity somewhat distinct al-
though merging (43-44). Forewing with
venation reduced apically; ten closed cells
(45-46). Pterostigma fairly large and
sclerotized (47-48). Hind wing with vena-
tion reduced apically; two closed cells;
vein C reduced distally; veins A and J
absent; cu-e originating proximal to sepa-
ration of M and Cu; free section of Cu
present (49-53). Basal hamuli strongly
clustered at separation of vein SC+R+S
from margin (54). Plical lobe well-marked
by a deep incision; moderate jugal lobe
present (55-56). Legs slender, generally
unmodified; arolia reduced; claws toothed
(57-59). Foretibia with single calcar al-
most straight, not highly modified (60).
Mid-tibia with scattered weak spines; two
simple spurs (61, 63-65). Hind coxa with
weak dorsal carina basally (66). Hind
tibia with scattered weak spines; two
simple spurs (62, 64, 67-68). Modified
mesosoma of female with pronotum large,
articulating with mesothorax; mesonotum
extremely short and transverse; scuto-
scutellar suture absent; prepectus probably
fused to mesepisternum but hidden under
pronotum; mesepimeron not differentiated
from mesepisternum; meso- and meta-
pleura fused but suture distinct; metatho-
rax not distinguishable but apparently
transverse (69).

Metasoma — "Felt line" developed on sec-
ond segment, on tergum only, as line of
dense pubescence (70). Stridulitrum pres-
ent as a shield-shaped area on each side
at base of fourth tergum (71). First seg-
ment somewhat apically constricted, with
short petiole; first tergum with lateral
margin only very slightly overlying ster-
num, apparently not movable against ster-
num although not fused to it except
perhaps at basal extreme; first sternum
entirely depressed, slightly more so pos-
teriorly, separated from second by a deep
constriction (72-76). Second segment not

Phylogeny and Classification of the Aculeate Hymenoptera

567

apically constricted (77). Female with
seventh tergum hidden; apex o£ gonapo-
physis IX not directed downward (addi-
tional details of female terminalia un-
known) (78, 81). Male with seventh
sternum somewhat reduced although not
hidden; hypopygium simple, unmodified;
cercus present; gonapophyses IX with
notal fusion well-developed (82-87).

Larva — Unknown.

Behavioral formila — Unknown.

12. Typhoctid group.

Material examined: Typhoctes peciiliaris
( 6\ 9 ) ; T. striolatus ( ? ) ; T. williamsi
(i)'-, Typhoctoides aphelonyx (2).
("Typhoctinae" of the "Mutillidae")
(Data also from Reid, 1941; Schuster,
1949; Krombein & Schuster, 1957; Broth-
ers, 1970.)

General — Sexual dimorphism considera-
ble, male smaller but not much more
slender than female; male winged, female
apterous (1-2). No sterile caste (3). Much
pubescence of sub-plumose setae (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket simple, without tubercle (6).
Eye oval with inner margin very shallowly
sinuate (slightly incised in male), not pro-
tuberant; scattered pores with short setae
(7-9). Three ocelli present (absent in fe-
male) (10). Genal secretory organ absent
(11). Antenna with 13 segments in male,
12 segments in female; axis of radicle not
deviating much from that of scape; radicle
not set into basal invagination of scape
(12-14). Labio-maxillary complex not
elongated; maxillary palpus 6-segmented;
labial palpus 4-segmented (15-17).

Mesosoma — Pronotum large, very little re-
duced medially, freely articulating with
mesothorax; hind margin arcuate; ante-
rior "collar" well-developed; posterolateral
angle rounded, reaching tegula; postero-
ventral margin approximately straight;

ventral angle rounded, not produced (18-
23). Propleura posteriorly contiguous mes-
ally (24). Prosternum sunken except for
anterior rim (25). Forecoxae nearly con-
tiguous (26). Mesonotum not much an-
teriorly produced; scutellum not enlarged
(27-28). Prepectus not fused to mesepi-
sternum, discontinuous midventrally, the
halves considerably separated, each half
very short and tall (29). Mesepimeron
reduced, distinguishable only at dorsal ex-
treme (30). Mesosternum simple, without
posterior projections (31). Mid-coxae
somewhat separated (32). Meso- and met-
apleura not mutually movable although
not fused {i?)). Metanotum not reduced
medially; metapostnotum not visible me-
dially, invaginated (34-35). Metepimeron
very slightly expanded anteroventral to
endophragmal pit so that pit is a short
distance from anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum depressed only laterally to
accommodate mid-coxae, mesally little dif-
ferentiated from mesosternum and very
slightly anteriorly produced (38-40). Hind
coxae contiguous (41). Metathoracic-pro-
podeal pleural suture visible dorsal to
endophragmal pit, not distinguishable ven-
trally (42). Propodeum not shortened;
disc and declivity somewhat distinct al-
though merging (43-44). Forewing with
venation reduced apically; ten closed cells
(45-46). Pterostigma small and sclerotized
(47-48). Hind wing with venation re-
duced apically; three closed cells; veins A
and J absent; cu-e originating proximal to
separation of M and Cu; free section of
Cu present (49-53). Basal hamuli strongly
clustered at separation of veins C and
SC+R+S (54). Plical lobe not indicated;
jugal lobe absent (55-56). Legs slender,
generally unmodified; arolia well-devel-
oped; claws toothed (57-59). Foretibia
with single calcar almost straight, not
highly modified (60). Mid-tibia with scat-
tered moderate spines; two simple spurs

568

The University of Kansas Science Bulletin

(61, 63-65). Hind coxa with weak dorsal
carina basally (66). Hind tibia with scat-
tered moderate spines; two simple spurs
(62, 64, 67-68). Modified mesosoma of
female with pronotum large, articulating
with mesothorax; mesonotum extremely
short and transverse; scuto-scutellar suture
absent; prepectus fused to mesepisternum;
mesepimeron not differentiated from
mesepisternum; meso- and metapleura
fused but suture distinct; metanotum very
short and transverse, not distinguishable,
fused to propodeum; endophragmal pit
barely distinguishable, apparently within
meso-metapleural suture; metathoracic-
propodeal pleural suture barely visible dor-
sal to endophragmal pit, not distinguish-
able ventrally (69).

Metasonl\ — "Felt line" developed on sec-
ond segment, on tergum only, as line of
dense pubescence (70). Stridulitrum pres-
ent as a shield-shaped area on each side
at base of fourth tergum (71). First seg-
ment somewhat apically constricted, with
short petiole; first tergum with lateral
margin only very slightly overlying ster-
num posteriorly, not movable against ster-
num although not fused to it except per-
haps at extreme base; first sternum entirely
depressed, slightly more so posteriorly,
separated from second by a deep constric-
tion (72-76). Second segment not apically
constricted (77). Female with seventh ter-
gum hidden, reduced but continuously
sclerotized anteriorly; no articulation
within section 1 of gonocoxite IX; no
valve on gonapophysis VIII; gonapophysis
IX almost straight with apex weakly di-
rected obliquely upward (78-81). Male
with seventh sternum somewhat reduced
although not hidden; hypopygium sim-
ple, unmodified; cercus present; gonapo-
physes IX with notal fusion well-de-
veloped although slightly dcsclerotized
(82-86).

Larva — Unknown.

Behavioral formula — Unknown.

23 . Chyph otid gro up.

Material examined: Chyphotes attenii-
atiis ( <:5 ) ; C. aiiripihis ( 9 ) ; C. beljragei
((5,9); C. elevatus ( c? ) ; C. evansi ( 9 ) ;
C. melaniceps {i)\ C. niibecitliis ( <5 ) ;
C. petiolatiis ( 9 ) ; C. si mills ( 5 ) ; Chy-
photes spp. (6,9; Mexico). ("Chypho-
tini" of the "Mutillidae")
(Data also from Buzicky, 1941; Reid,
1941; Mickel, 1967; Debok, 1973.)

General — Sexual dimorphism considera-
ble, although male not much more slender
than female; male winged, female wing-
less (1-2). No sterile caste (3). All pu-
bescence simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim pro-
duced, forming a moderate tubercle (6).
Eye almost round with inner margin
sinuate (convex in female), protuberant;
scattered pores with short setae (appar-
ently without pores and glabrous in fe-
male) (7-9). Three ocelli present (absent
in female) (10). Genal secretory organ
absent (11). Antenna with 13 segments
in male, 12 segments in female; axis of
radicle deviating from that of scape at an
oblique angle; radicle not set into basal
invagination of scape (12-14). Labio-max-
illary complex not elongated; maxillary
palpus 6-segmented; labial palpus 4-seg-
mented (15-17).

Mesosoma — Pronotum large, not reduced
medially (somewhat reduced in male),
freely articulating with mesothorax; hind
margin almost straight (strongly arcuate
in male); anterior "collar" well-developed;
posterolateral angle rounded, reaching teg-
ula; posteroventral margin very slightly
concave; ventral angle more or less acute,
slightly produced (18-23). Propleura pos-
teriorly contiguous mesally (24). Proster-
num sunken except for anterior rim (25).
Forecoxae nearly contiguous (26). Meso-
notum not anteriorly produced (slightly

Phylogeny and Classification of the Aculeate Hymenoptera

569

produced in male); scutellum not enlarged
(27-2S). Prepectus fused to mesepister-
num, discontinuous midventrally, the
halves considerably separated, each half
very short and tall (29). Mesepimeron re-
duced, distinguishable only at dorsal ex-
treme (30). Mesosternum simple, without
posterior projections (31). Mid-coxae
somewhat separated (32). Meso- and met-
.iplcura not mutually movable, apparently
fused (ii). Metanotum not reduced me-
dially; metapostnotum barely visible me-
dially, invaginated (34-35). Metepisternum
very slightly expanded anteroventral to
cndophragmal pit so that pit is a short
distance from anterior margin of meta-
pleuron; no metapleural gland (36-37).
Metasternum depressed only laterally to
accommodate mid-coxae, mesally little dif-
ferentiated from mesosternum and very
slightly anteriorly produced (38-40). Hind
coxae contiguous (41). Metathoracic-pro-
podeal pleural suture visible dorsal to
cndophragmal pit, not distinguishable ven-
trally (42). Propodeum not shortened;
disc and declivity somewhat distinct al-
though merging (43-44). Forewing with
venation reduced apically; ten closed cells
(45-56). Pterostigma fairly large and
sclerotized (47-48). Hind wing with ve-
nation reduced apically; two closed cells;
vein C reduced distally; veins A and J
absent; cu-e originating proximal to sepa-
ration of M and Cu; free section of Cu
present (49-53). Basal hamuli strongly
concentrated into a very small basal cluster
at separation of vein SC4-R4-S from mar-
gin (54). Plical lobe not indicated; mod-
erate jugal lobe present (55-56). Legs
slender, generally unmodified (slightly
stouter in female); aroHa well-developed
(much reduced in female) ; claws toothed
(57-59). Foretibia with single calcar al-
most straight, not highly modified (60).
Mid-tibia with few moderate spines; two
simple spurs (spurs dorsally flattened in
female) (61, 63-65). Hind coxa appar-

ently simple (with moderate dorsal carina
in male) (66). Hind tibia with very few
moderate spines apically only; two simple
spurs (spurs dorsally flattened in female)
(62, 64, 67-68). Modified mesosoma of
female with pronotum large, articulating
with mesothorax; meso- and metathorax
and propodeum entirely fused dorsally and
laterally; no sutures visible except that
meso-metanotal suture barely evident;
mesonotum large, posteriorly expanded;
prepectus fused to mesepisternum; pro-
podeum reduced dorsally; disc lost (69).

METAso^L\ — "Felt line" developed on sec-
ond segment, on tergum only, as line of
dense pubescence (70). Stridulitrum pres-
ent as a shield-shaped area on each side
at base of fourth tergum (71). First seg-
ment somewhat apically constricted, with
marked petiole; first tergum with lateral
margin only very slightly overlying ster-
num posteriorly, not movable against ster-
num and fused to it basally, much nar-
rowed along petiole (tergum apparently
absent from petiole in female) ; first ster-
num depressed, slightly more so posteri-
orly, separated from second by a deep
constriction (72-76). Second segment not
apically constricted (77). Female with
seventh tergum hidden, reduced but con-
tinuously sclerotized anteriorly; no articu-
lation within section 1 of gonocoxite IX;
no valve on gonapophysis VIII; gonapo-
physis IX almost straight, with apex
weakly directed obliquely upward (78-81).
Male with seventh sternum somewhat re-
duced, not entirely hidden; hypopygium
basally concealed by sixth and seventh
sterna, apically produced as three spines,
the middle one produced and upcurved;
cercus present; gonapophyses IX with no-
tal fusion extensive (82-86).
Larva — Unknown.
Behavioral formula — Unknown.

24. Apterogynid group.

Materl\l examined: Apterogyna algirica

570

The University of Kansas Science Bulletin

( c$ ) ; A. aliwalica (9); A. bembesia
(9); A. bimacitlata (9)\ A. bitlawayona
(S); A. geyri {$)\ A. globularis {S);
A. guillannodi ( 9 ); A. \arroa ($); A.
\ochi (S ); A. olivieri (6 ,9); A. patrizii
{i)\ A. savignyi ( <5 , 9 ) ; A. schiiltzei
( t5 , 9 ) ; A. villiersi {$). ("Apterogy-
nini" of the "Mutillidae")
(Data also from Reid, 1941; Giner Mari,
1944; Invrea, 1957.)

General — Sexual dimorphism considera-
ble, although male not more slender than
female; male winged, female apterous
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (somewhat
higher in male) (5). Antennal socket
with dorsomesal rim produced, forming
well-developed tubercle (6). Eye oval,
somewhat shortened and almost round,
with inner margin extremely shallowly
sinuate (convex in female), protuberant;
pores and setae apparently absent (7-9).
Three ocelli present (absent in female)
(10). Genal secretory organ well-devel-
oped (11). Antenna with 13 segments in
male, 12 segments in female; axis of rad-
icle deviating from that of scape at an
oblique angle; radicle not set into basal
invagination of scape (12-14). Labio-max-
illary complex not elongated; maxillary
palpus 6-segmented; labial palpus 4-seg-
mented (15-17).

Mesosoma — Pronotum large, not reduced
medially (much reduced in male), freely
articulating with mesothorax; hind mar-
gin shallowly arcuate (deeply concave in
male); anterior "collar" well-developed;
posterolateral angle rounded, reaching teg-
ula; posteroventral margin very slightly
concave; ventral angle acute, slightly mes-
ally produced (18-23). Propleura posteri-
orly contiguous mesally (24). Prosternum
sunken except for anterior rim (25). Fore-
coxae contiguous (26). Mesonotum not
anteriorly produced (considerably pro-

duced in male) ; scutellum somewhat en-
larged and swollen posterodorsally (27-
28). Prepectus not fused to mesepisternum,
discontinuous midventrally, the halves
considerably separated, each half ex-
tremely short and tall (29). Mesepimeron
reduced, distinguishable only at dorsal ex-
treme (30). Mesosternum simple, without
posterior projections (31). Mid-coxae
somewhat separated (considerably sepa-
rated in female) {^2). Meso- and meta-
pleura not mutually movable although
apparently not fused (i^). Metanotum
somewhat reduced medially; metapostno-
tum not visible medially, invaginated (34-
35). Metepisternum expanded anterior to
endophragmal pit so that pit is some dis-
tance from anterior margin of metapleu-
ron; no metapleural gland (36-37). Met-
asternum depressed only laterally to
accommodate mid-coxae, mesally little-
dififerentiated from mesosternum and con-
siderably anteriorly produced (38-40).
Hind coxae contiguous (41). Metatho-
racic-propodeal pleural suture visible dor-
sal to endophragmal pit, not distinguish-
able ventrally (42). Propodeum slightly
shortened dorsally; disc and declivity
somewhat distinct although merging (43-
44). Forewing with venation much re-
duced apically; five closed cells; cells R,
IS, 2S, S+M, IM absent (45-46). Ptero-
stigma very small and sclerotized (47-48).
Hind wing with venation considerably
reduced apically; three closed cells; veins
A and J absent; cu-e originating proximal
to separation of M and Cu; free section of
Cu present (49-53). Basal hamuli absent
(54). Plical lobe not indicated; moderate
jugal lobe present (55-56). Legs slender,
generally unmodified (slightly stouter in
female); arolia much reduced; claws
toothed (57-59). Foretibia with single cal-
car considerably curved and flattened with
an outer spine apically (60). Mid-tibia
with very few moderate spines apically
only; two spurs flattened and with

Phvlogeny and Classification of the Aculeate Hymenoptera

571

strongly toothed margins (61, 63-65).
Hind coxa with moderate dorsal carina
(66). Hind tibia with very few moderate
spines apically only; two spurs flattened
and with strongly toothed margins (62,
64, 67-68). Modified mesosoma of female
with pronotum large, articulating with
mesothorax; meso- and metathorax and
propodeum entirely fused dorsally and lat-
erally; meso-metanotal suture barely evi-
dent; mesonotum large, somewhat pos-
teriorly expanded; prepectus fused to
mesepisternum; meso- and metapleura
fused but suture visible; line of fusion be-
tween nota and pleura of meso- and met-
athorax weakly marked; metathoracic-
propodeal suture entirely obliterated;
propodeum reduced dorsally; disc lost
(69).

Metasoma — "Felt line" developed on sec-
ond segment, on tergum only, as line of
dense pubescence (as invaginated cuticu-
lar line in female) (70). Stridulitrum
absent (71). First segment considerably
constricted apically, nodose, with marked
petiole; first tergum with lateral margin
only very slightly overlying sternum pos-
teriorly, not movable against sternum and
fused to it basally, much narrowed along
petiole; first sternum depressed, slightly
more so posteriorly, separated from the
second by a deep constriction (72-76).
Second segment considerably constricted
apically (77). Female with seventh ter-
gum hidden, reduced but continuously
sclerotized anteriorly; no articulation
within section 1 of gonocoxite IX; valve
present on gonapophysis VIII; gonapophy-
sis IX almost straight with apex weakly
directed obliquely downward (7(S-81).
Male with seventh sternum reduced but
not entirely hidden; hypopygium not con-
cealed basally, apically produced as three
spines, the middle one produced and up-
curved; cercus absent; gonapophyses IX
with notal fusion extensive (82-86).

La rva — Un k now n .

Beha\i()ral formula — Unknown.

25. Bradyuobaenid group.

Material examined: Bradynobaeniis aus-
tral is ( c5 ) ; B. gayi {i ,9)\ B. wagen-
l{nechti ( <i ) ; Bradynobaenus spp. ( 9 ? ;
Chile, Argentina). ("Bradynobaeninae"
of the "Tiphiidae")

(Data also from Spinola, 1851; Zavattari,
1910; Reid, 1941; Pate, 1947a; Perez
D'Angello, 1968.)

General — Sexual dimorphism considera-
ble although male not more slender than
female; male winged, female apterous
(1-2). No sterile caste (3). All pubescence
simple (4).

Head — Clypeus of moderate size; anten-
nae inserted fairly low on face (5). An-
tennal socket with dorsomesal rim very
slightly produced but no true tubercle;
interantennal frons slightly produced and
forming a weak frontal ledge (6). Eye
broadly oval with inner margin extremely
shallowly sinuate, not protuberant; pores
and setae apparently absent (7-9). Three
ocelli present (10). Genal secretory organ
present (not clearly distinguishable in
male) (11). Antenna with 13 segments in
male, 12 segments in female; axis of rad-
icle deviating from that of scape at an
oblique angle; radicle not set into basal
invagination of scape (12-14). Labio-max-
illary complex reduced; maxillary palpus
2-segmented; labial palpus 3-segmented
(1547).

Mesosoma — Pronotum large, not reduced
medially (extremely reduced in male),
freely articulating with mesothorax; hind
margin shallowly arcuate (very deeply
concave in male); anterior "collar" mod-
erate; posterolateral angle rounded, reach-
ing tegula, slightly posteriorly produced be-
low tegula; posteroventral margin slightly
concave; ventral angle acute, slightly mes-
ally produced (18-23). Propleura posteri-

572

The University of Kansas Science Bulletin

orly contiguous mesally (24). Prosternum
sunken except for small anterior rim (25).
Forecoxae nearly contiguous (26). Meso-
notum not anteriorly produced (greatly
produced in male) ; scutellum much en-
larged and swollen posterodorsally, over-
hanging metanotum (27-28). Prepectus
not fused to mesepisternum, discontinuous
midventrally, the halves considerably sep-
arated, each half extremely short and tall
(29). Mesepimeron reduced, distinguish-
able only at dorsal extreme (30). Meso-
sternum simple, without posterior projec-
tions (31). Mid-coxae somewhat separated
(considerably so in female) (32). Meso-
and metapleura not mutually movable
although apparently not fused (33).
Metanotum somewhat reduced medially;
metapostnotum not visible medially, in-
vaginated (34-35). Metepisternum ex-
panded anterior to endophragmal pit so
that pit is some distance from anterior
margin of metapleuron but sutures ante-
rior to pit are indistinct; no metapleural
gland (36-37). Metasternum depressed
only laterally to accommodate mid-coxae,
mesally little differentiated from mesoster-
num and considerably anteriorly produced
(moderately produced in female) (38-40).
Hind coxae contiguous (41). Metatho-
racic-propodeal pleural suture visible dor-
sal to endophragmal pit, very indistinctly
distinguishable ventrally (42). Propodeum
considerably shortened dorsally; disc vir-
tually non-existent, distinct from declivity
(43-44). Forewing with venation ex-
tremely reduced apically; three closed
cells; cells R, SC+R, IS, 2S, S+M, IM,
ICu absent (45-46). Pterostigma very
small and sclerotized (47-48). Hind wing
with venation extremely reduced apically;
two closed cells; vein M+Cu absent; veins
A and J absent; cu-e originating proximal
to separation of M and Cu; free section
of Cu present (49-53). Basal hamuli absent
(54). Plical lobe not indicated; moderate
jugal lobe present (55-56). Legs fairly

slender, little modified (considerably
stouter in female); arolia much reduced;
claws simple (57-59). Foretibia with single
calcar considerably curved and flattened
with an outer spine apically (spine absent
in female) (60). Mid-tibia with few
strong spines apically only; no spurs (61,
63-65). Hind coxa simple (66). Hind
tibia with few strong spines apically only;
two spurs flattened with a few teeth on
margins (nematiform and smooth in fe-
male) (62, 64, 67-68). Modified mesosoma
of female with pronotum large, articulat-
ing with mesothorax; meso- and metatho-
rax and propodeum entirely fused dorsally
and laterally; meso-metanotal suture barely
evident; mesonotum large, somewhat pos-
teriorly expanded; prepectus fused to
mesepisternum; meso-metapleural suture
faintly visible; line of fusion between nota
and pleura of meso- and metathorax dis-
tinct; metathoracic-propodeal suture en-
tirely obliterated; propodeum reduced dor-
sally; disc lost (69).

Metasoma — "Felt line" developed on sec-
ond segment, on tergum only, as invagi-
nated cuticular line in both sexes (70).
Stridulitrum absent (71). First segment
slightly constricted apically, with short
petiole; first tergum with lateral margin
only very slightly overlying sternum pos-
teriorly, not movable against sternum and
fused to it basally, much narrowed along
petiole; first sternum depressed, more so
posteriorly, separated from second by a
deep constriction (72-76). Second segment
not apically constricted (77). Female with
seventh tergum hidden, reduced but con-
tinuously sclerotized anteriorly; no articu-
lation within section 1 of gonocoxite IX;
no valve on gonapophysis VIII; gonapo-
physis IX strongly curved, especially bas-
ally, with apex downwardly directed (78-
81). Male with seventh sternum reduced
but not entirely hidden; hypopygium not
concealed basally, apically produced as
three subequal spines; cercus absent;

Phyloceny and Classification of the Aculeate Hymenoptera

573

gonapophyses IX with iiotal fusion ex-
tensive (82-86).

LAr<\A — Unknown.

Behavioral formula — Unknown.

Discussion

Cladistic relationships

Surprisingly, there have apparently
been few intensive attempts to derive a
phylogeny of the aculeate Hymenoptera.
Although various schemes have been pro-
posed (e.g., Ashmead, 1896b; Borner, 1919;
Evans, 1958, 1966b; Ross, 1965; Evans &
West Eberhard, 1970), these generally do
not explicitly state the data on which they
are based, or else they are based on a very
limited data set so that detailed compari-
sons with the present scheme are of little
value. A recent scheme of the evolution
of the Hymenoptera is that of Evans &
West Eberhard (1970), the upper section
of which is shown in Fig. 62. This dia-
gram expresses the commonly accepted
ideas on aculeate evolution and should be
used as the frame of reference for discus-
sion of the present scheme (Fig. 2).

Basically, a sequence of more or less
straight-line evolutionary advancement
seems to have been emphasized, so that
the more "primitive" groups (e.g., bethy-
loids and scolioids) are shown as arising
early and the more "advanced" groups
(e.g., apoids) are derived at the apex of
the tree (Fig. 62). It now seems that
some of the most "advanced" groups origi-
nated near the base of the tree and subse-
quently diverged greatly from the original
stock. (This in essence reemphasizes the
inappropriateness of the terms "primitive"
and "advanced" as applied to entire taxa.)
It is of some interest that Spradbery (1973)
has recently presented a diagram which
is somewhat more similar to that derived
here in that he shows the line leading to
the bees arising above the bethyloids but

Apoidea

Sphecoidea

Pompiloidea

Bethyloidea

Vespoidea

Formicoidea

Scolioidea

Fk,. 62. Phyl(jgcny of the traditional superfamilies

of Aculeata, according to Evans & West Eberhard

(1970).

below the scolioids. On the other hand,
he places the chrysidids in the Terebrantia,
which is at variance with most other
schemes.

In the present cladogram the scoleby-
thids are extremely close to the bethylid
group and may even have been derived
from a stock very similar to modern
"Bethylidae," indicating that this group
is clearlv more strontjly associated with
the bethylid group than any other.
(Evans, 1963, by contrast, included the
scolebythids in the "Scolioidea.") Al-
though the plumariids possess many char-
acters in states which have otherwise oc-
curred only in lines on or above internode
1-4, they are linked to the bethylids and
scolebythids by the common development
of an articulation within section 1 of
gonocoxite IX in the female and by the
loss of the jugal lobe in the hind wing.

The remaining aculeates are strongly
associated, most strikingly by common pos-
session of sexual dimorphism in the num-

574

The University of Kansas Science Bulletin

ber of antennal segments (13 segments in
male and 12 segments in female), and re-
duction and desclerotization of the seventh
metasomal tergum in the female. Thus
the basal division of the Aculeata into two
holophyletic lines as proposed by Oeser
(1961) is not contradicted here.

The common ancestor of the sphecid
and apid lines apparently diverged very
early from the other aculeates, thus estab-
lishing a separate and highly divergent
line. This is most strikingly characterized
by the forms of the pronotum, metapost-
notum and metapleuron, which serve to
associate the apid and sphecid groups very
strongly.

The remaining line of aculeates is not
associated by any very strong character
states, although the posterior approxima-
tion of the propleura, shortening of the
prepectus, anterior depression of the meta-
sternum, loss of veins A and J (hind
wing) and deepening of the axillary in-
cision may all have occurred on a line
(internode 4-6) associating this group.
These character states have also evolved
elsewhere, however, although the first and
second are probably the strongest and ap-
pear on the cladogram only once each.

Within this grouping there appear to
be two lines, one of which is much more
firmly established than the other. The
group containing the tiphiids, mutillids
and their relatives (subtended by inter-
node 6-7) is associated by a number of
reasonably good character states, such as
the distal position of crossvein cu-e (hind
wing), the partial invagination of the
metapostnotum and the reduction of the
second larval spiracle (which has appar-
ently been reversed in, or within, the
sapygids). Although these are individually
not particularly impressive, the entire com-
plex of character states involved is evi-
dently significant. Within this group the
association of the sapygids, myrmosids and
mutillids is based on some quite strong

character states, especially the loss of the
free section of vein Cu (hind wing) (al-
though with apparent reversal within the
mutillid group), reduction of the jugal
lobe of the hind wing, and desclerotization
and basal retraction of the gonapophyseal
notum in the male. Although the myrmo-
sids have recently commonly been in-
cluded in the "Tiphiidae" (e.g., Krom-
bein, 1940; Pate, 1947a), their association
with the mutillids appears much stronger,
being based particularly on the form of
the scape, fusion of the prepectus to the
mesepisternum, the form of the modified
female mesosoma, and the mesal stridu-
litrum. The general similarity between
the myrmosids (males) and Fedtschenkja
(sapygid group) is also quite marked,
especially in wing venation, reinforcing
the grouping of these taxa. It is of some
interest that Fedtschenkja was first de-
scribed as a member of the Mutillidae
(Saussure, 1880) and a short time later
was placed in a separate tribe of the mu-
tillids since "ils forment le trait d'union
entre les Mutillines et les Sapygides"
(Saussure, 1892). If indeed Fedtschenkja
is correctly considered to be a relatively
"primitive" member of the "Sapygidae"
(a placement that was not found to be
contradicted during the present study),
then the present investigation supports
this relationship which was first suggested
more than 80 years ago. (GuigUa, 1955,
1969, has outlined the taxonomic history
of Fedtschenkja^

The group containing the tiphiids con-
sists of six taxa associated (on internode
7-10) by rather few characters of which
the major one is probably the development
of lamellate projections over the mid-coxae
(although these have been reduced in the
methochids). The group is also associated
by a somewhat similar body form, espe-
cially of the females. The five taxa beyond
the anthoboscids share a tendency towards
reduction or loss of the wings in the fe-

Phylogeny and Classii-ication of the Aculeate Hymenoptera

575

male, resulting; in similar mesosomal mod-
ifications, although the tiphiid group does
not exhibit complete aptery in any mem-
ber. The four taxa beyond the thynnids
are strongly associated by the modification
of the male hypopygium as a simple up-
curved aculeus. The myzinid and metho-
chid groups share males of a peculiar
elongate form, in particular, and the tiphi-
ids and brachycistids both have the sev-
enth sternum and hypopygium of the male
almost completely concealed, amongst
other shared character states.

The remaining taxa (subtended by in-
ternode 6-15) are very weakly associated
basally and may not represent a holophy-
letic group. Internodes 6-15, 15-16 and
16-18 are all rather weak (each bearing
relatively few character states, most of
which occur numerous times elsewhere on
the cladogram) and may not be supported
when additional data are considered. De-
spite these limitations, the sierolomorphids
do appear to be similar to the pompilid
group and also the typhoctids, especially
in the structure of the female genitalia.
The pompilids and rhopalosomatids are
quite strongly associated, particularly by
the similar development of a calcar on
the hind tibia, and to some extent by the
venation of the hind wing (especially the
loss of the apical section of vein E).

The ten taxa above the sierolomorphids
are associated (on internode 15-16) by the
clustering of the basal hamuli, the reduc-
tion in the size of the pterostigma and per-
haps the reduction in the number of teeth
on the larval mandible (tridentate or less),
all rather weak characters, especially the
last since the larva is unknown for the
entire branch subtended by internode 18-
21. Association of the eight apical groups
(above internode 16-18) is mainly on the
basis of the complete invagination of the
metapostnotum, which has also occurred
elsewhere.

The formicid-vespid-scoliid line is as-

sociated by the common production of the
ventral angle of the pronotum and ante-
rior expansion of the metepisternum and
metepimeron (which is least obvious in
the scoliids). These are probably not very
strong character states. However, the
formicid and vespid groups share other
states, such as the extreme reduction of
the prepectus, the use of numerous prey
individuals and early oviposition, which
may additionally indicate some associa-
tion based on parallel development of these
character states. By contrast, the somewhat
unexpected association of the scoliids and
vespids seems to be based on a number
of good character states. These include
especially the form of the pronotum and
its close association with the mesothorax,
as well as the reniform eyes, dorsally pro-
duced clypeus, sunken prosternum and
membranous pterostigma. (They of course
also share all derived states appearing on
internodes below internode 19-20.) Both
of these taxa also possess highly muscular
and similarly modified venom reservoirs
with the muscle strands running trans-
versely (Robertson, 1968), which supports
this relationship, although Robertson came
to somewhat different conclusions in her
study which was based on a rather limited
data set.

The remaining five taxa are very
strongly associated (on internode 18-21)
by a number of characters, such as the
form of the metasternum, the basic form
of the modified mesosoma in females (all
of which are apterous), the presence of
dorsal "felt lines" and the paired stridu-
litra (which are, however, lost in die
apterogynids and bradynobaenids). Within
this group the eotillids and typhoctids are
strongly linked by the form of the meso-
soma in the female. The other three taxa
share an angulate scape, a similarly modi-
fied mesosoma in the female, a similarly
modified first metasomal segment, and a
trispinose hypopygium in the male. The

576

The University of Kansas Science Bulletin

association of the apterogynids and brady-
nobaenids has been mentioned previously
by Reid (1941) and in passing by Bischoff
(1927); it is strongly supported, particu-
larly by the specialized genal organ, the
enlarged scutellum, the identical modifica-
tion of wins venation and the form of the
foretibial calcar.

When life history data are considered,
the cladogram does not contradict the
idea that the ancestral aculeates were asso-
ciated with hemimetabolous insects, such
as Orthoptera or Hemiptera, which may
have been more plentiful than the more
recently evolved holometabolous groups.
Thus, for example, various members of
the bethylid group attack Hemimetabola
[dryinids and embolemids on Homoptera
(Ponomarenko, 1972; Evans, 1964a), me-
sitiines on Orthoptera (Gryllidae) (Niel-
sen, 1932)— although Nagy, 1969c, found
Mesitiiis horvathi parasitic on a clythrine
beetle — and sclerogibbids on Embioptera
(Callan, 1939; Richards, 1939)], as do the
most generalized sphecoids [ampulicids
on Blattodea and sphecines mainly on
Orthoptera (Clausen, 1940)], these taxa
representing the most basal groups on the
cladogram. Within each of these taxa
there have been departures to the use of
Holometabola [e.g., Coleoptera larvae by
many "BethyUdae" (Evans, 1964a), Dip-
tera and Lepidoptera by higher "Nyssoni-
nae" (Evans, 1966a)]. The rhopalo-
somatids attack Orthoptera (Gryllidae)
(Gurney, 1953) and the pompiUds utiHze
spiders (Evans, 1953), another ancient
group. Above internode 4-5, the apids
have undergone a drastic change to the
use of pollen and nectar (Malyshev, 1968).
On internode 7-8 a change to parasitiza-
tion of other Hymenoptera nesting in the
soil apparently occurred. This is the basic
habit of the sapygid (Bohart & Schuster,
1972), myrmosid and mutillid (Mickel,
1928) groups, although many sapygids
have become cleptoparasitic on bees

(Malyshev, 1968; Torchio, 1972) and a
few mutillids parasitize hosts other than
Hymenoptera (Brothers, 1972). All the
taxa above internode 7-10 utilize cryptic
coleopterous larvae (Pate, 1947a), as do
the scoliids (Clausen, 1940) (this probably
resulting in the convergence of body form
between various tiphioids and the scoliids).
(The hosts are actually unknown for the
brachycistidid group, but see Wasbauer,
1966.) Primitive formicids have adopted
a predatory habit connected with their so-
ciality (Wilson, 1971). The vespids use
many prey individuals of various types
with which they provision their complex
nests, and a large group of these (the
masarids) has adopted pollen and nectar
(although some of the most primitive
masarids — Eitparagia — use the larvae of
weevils; Williams, 1927). Unfortunately,
nothinii is known of the life histories o£
plumariids, scolebythids (although para-
sitism of woodboring beetle larvae is
probable; Evans, 1963), sierolomorphids,
typhoctids, eotillids, chyphotids, apterogy-
nids or bradynobaenids, so that they can-
not be related to this scheme.

Although such attempts can be made
to correlate prey or host group with a
cladogram derived almost entirely from
morphological characters, much more
work is needed on the life histories of
most groups of Aculeata before any more
definite indications of the patterns of evo-
lution of host relationships can be given.
Data on entirely new aspects of behavior
will also undoubtedly prove useful. For
example. Parish (1972) examined groom-
ing behavior in 51 species of aculeates (as
well as other Hymenoptera) and applied
the results to a somewhat unusual phylog-
eny which he had constructed from pre-
viously published information. Although
very incomplete (due to poor representa-
tion in some groups), the patterns derived
seem to confirm the distinctness of the
Aculeata. Again, additional investigations

■'

Phyloceny and Classification of the Aculeate Hymenoptera

577

are needed, but these initial results seem
promising. Despite the prematurity of the
attempts by workers such as Iwata (1942)
and Malyshev (1968) to derive a unified
scheme of the evolution of behavioral pat-
terns in the Hymenoptera, such studies
are of great heuristic value in emphasizing
those areas most needing additional study.

Geographic distribution patterns often
supply additional data which may support
or refute a particular cladogram, especially
when considered in the liiiht of recent ad-
vances in the understanding of plate tec-
tonics (see e.g., Axelrod & Raven, 1972),
and using the principles derived by Hen-
nig (1966b). Unfortunately, such data are
of minimal use in the present study since
most groups are essentially cosmopolitan.
Furthermore, members of at least the
bethylid ( Procleptes ), sphecid { Archisphex
and Lisponema), formicid (Sphecomyrma)
and possibly plumariid (Cretaviis) groups
are known from the Cretaceous (Sharov,
1957; Evans, 1967, 1969b; Wilson, Car-
penter & Brown, 1967), suggesting that
most of the gross evolution within the
aculeates took place before the breakup of
Pangaea had proceeded very far. The ex-
istence of fossil ants older than any pom-
pilids (Evans, 1969b) places further doubt
on the accuracy of the internodes associat-
ing these groups; although the cladogram
does not preclude an earlier origin for the
ants than for the pompilids, such a con-
dition would seem unlikelv. The incom-
pleteness of the fossil record may be the
most important factor in this regard,
however.

When the present cladogram (Fig. 2)
is compared with a commonly accepted
scheme such as that given by Evans &
West Eberhard (1970) (Fig. 62), a few
striking differences may be noted. The
most important of these involves the posi-
tion and arrangement of the "Scolioidea."
This group, which contains many of the
taxa generally considered primitive in the

Aculeata, is shown as a holophyletic unit
in Evans & West Eberhard's scheme. The
taxa included in this group appear on four
distinct branches in the present cladogram,
however, as emphasized in Fig. 63, and the
scolebythids and plumariids (considered
by some authors to belong to the "Sco-
lioidea" and by Evans & West Eberhard
to be intermediate between the "Bethy-
Icjidea" and "Scolioidea") form two addi-
tional branches. Thus the "Scolioidea" is
a highly paraphyletic (if not polyphyletic)
group. If the ancestral aculeate (below
node 1) is considered to be a scolioid, then
most of the other commonly accepted
superfamilies have independently evolved
from various points within the "Scoli-
oidea," a situation which can be expressed
as in Fig. 64. This closely resembles the
scheme provided by Evans (1958) (Fig.
65), except as regards the "Pompiloidea."
Actually, in view of the fundamental de-

scolioids Vespoidea

Formicoidea

scolioids

Pompiloidea

Bethyloidea

Apoidea
Sphecoidea

Fig. 63. Phylogeny of the traditional superfamilies

of Aculeata, from the present study (see Fig. 2),

emphasizing the paraphyletic (or polyphyletic) nature

of the Scolioidea (scolioids).

578

The University of Kansas Science Bulletin

Pompiioidea Apoidca
Vespoidea

Formicoidea

64

thyloidea

SCOLIOIDEA

Pompiioidea Apoidea

Vespoidea
Formicoidea

65

SCOLIOIDEA

Figs. 64-65. Relationships of the traditional super-
families of Aculeata. 64, from the present study,
considering the ancestral aculeate to be a member
of the Scolioidea; 65, according to Evans (1958).

rived characters present on internode 1-4,
it is not appropriate to consider the an-
cestor below node 1 to be a member of
the "Scolioidea," so that this scheme is
also inadequate. It is also singularly lack-
ing in information content.

The "Pompiioidea" has commonly been
considered as derived from the line lead-
ing to the sphecoids and apoids, appar-
ently because the posterolateral angle of
the pronotum is rather rounded in the
pompilids, a condition which has been
regarded as a precursor to that in the
sphecoids and apoids. Actually the primi-
tive condition in the pompilids is not un-
like that in many of the more generalized
scolioids, such as Anthobosca or Sierolo-
morpha, although the lobe over the spi-
racle is somewhat emphasized by the pres-
ence of a dorsoventral groove in the
pompilids. Furthermore, the pompilids do
not possess any of the other unique modi-

fications, such as the form of the meta-
postnotum and metapleuron, which are
characteristic of the apoids and sphecoids.
It is thus clear that there is no good char-
acter to link the pompilids to the sphecoid-
apoid line. Instead, the pompilids show
more affinities with the formicoid-vespoid
line.

Derivation of a classification

Given the present cladogram (Fig. 2),
the question arises as how best to trans-
late it into a classificatory scheme and
whether the traditional superfamilies
("Hethyloidea," "Scolioidea," "Pompiioi-
dea," "Vespoidea," "Formicoidea," "Sphe-
coidea" and "Apoidea") should logically
continue to be recognized as such. The
basic branching pattern (the result of
cladogenesis) is only one aspect to be con-
sidered; degree of relative morphological
or phenetic distinctness of taxa (the result
of anagenesis) is another. The latter is
difficult to measure objectively in such a
way that it can be incorporated directly
into a cladogram.

A simple method of estimating phe-
netic divergence might merely involve
summing the number of derived character
states on each particular internode, thus
deriving an index of phenetic divergence
for that internode. (For any taxon in
which the state of a character is unknown,
the primitive state is assumed unless other-
wise indicated by the distribution of de-
rived states in related taxa.) This is not
sufficient, however, since degree of clas-
sificatory (or taxonomic) distinctness is
somewhat dependent on the sizes of the
groups involved, as is expressed in Mayr's
(1969:92) recommendation that, for taxa
above the species level, "the size of the
[phenetic] gap [separating a taxon from
another of the same categorical level] be
in inverse ratio to the size of the taxon."
This is purely a formalistic requirement
designed to facilitate the data retrieval or

Phylogeny and Classification of the Aculeate Hymenoptera

579

reference functions of a classification by
minimizing the proliferation of named
taxa each containing only one or a few
rather distinctive species. It has further
been suggested by Ashlock (pers. comm.)
that it may be logical to judge the relative
importance of the various inter nodes in
terms of the "success" of the groups de-
rived from those internodes, possibly by
considering the number of species in each
group. Although "success" undoubtedly
involves many other factors, most of these,
such as biomass, number of individuals or
energy flow, are incapable of adequate
t|uantification in practice. (See Van Valen,
1973, for a discussion of criteria for com-
parison of taxa at the same categorical
level in different phyla.)

The importance ("distinctness") of a
particular internode may thus be estimated
by devising some index that combines the
number of evolutionary innovations (most
conveniently represented as hatch marks)
on that internode and the number of spe-
cies that has been produced by the various
lines derived from it. A simple product
of these numbers appears to serve this
purpose, as has been suggested by Ashlock
(pers. comm.). Such a distinctness mea-
sure based on numbers or frequencies (df )
may be expressed as

dlin — Kiii^i:

(1)

where km equals the number of derived
character states (hatch marks) on inter-
node (or final branch) m (assuming each
such state to appear once on the branch)
and Sim equals the total number of species
subtended by internode m.

There is, however, an additional factor
to consider: the fact that the distinctness
of a group depends on the uniqueness of
the derived character states delimiting the
group. Thus a group characterized by the
possession of many states uniquely derived
(i.e., possessed by that group only) may
logically be regarded as more distinct than

one which possesses an ecjual number of
states which have been derived more than
once on the tree. The weight to be at-
tached to any derived character state in
assessing its significance as a delimiter of
taxa for any internode, may logically be
considered to be proportional to the ratio
of its frequency of occurrence on that in-
ternode (assumed as unity) to its total fre-
quency on the entire tree. This value is
the reciprocal of the total number of times
that it appears on the tree. Based on these
considerations, another measure of dis-
tinctness for each internode (and final
branch) utilizes the reciprocal of the har-
monic mean (i.e., the mean of the re-
ciprocals) of the number of times that
each derived character state falling on that
internode appears on the entire tree. (Al-
though some character states appear only
once on the tree, they may have been de-
rived within other groups elsewhere on
the tree. Such states are obviously not as
efficient in delimiting groups as are those
which appear only once on the tree and
have not been derived elsewhere. It is
impossible to differentiate between these
two types in practice when dealing with
very large groups, however, so that only
occurrences shown on the tree can be in-
cluded.) Maximal distinctness of a branch
(all character states uniquely derived) is
given by a value of 1, with decreased dis-
tinctness being less than 1 but greater than
0 (unless the internode bears no derived
states). This distinctness measure based
on character state efficiencies (de) may be
expressed as

2 Nj-^

dem =

j=i

kr

(2)

where Nj equals the total number of times
that derived character state j appears on
the entire cladogram and km (as before)
equals the number of derived character

580

The University of Kansas Science Bulleti

N

States (hatch marks) on internode (or
final branch) m (assuming each such state
to appear once on the branch). This in-
dex is identical to the mean of the "unit
character consistencies" (Farris, 1969) for
the "characters" on the internode if each
character state is considered as a separate
"character" (i.e., each has a range of one
unit).

Neither of these measures alone is suf-
ficient to provide an adequate idea of dis-
tinctness; since all the factors mentioned
must be taken into account, a combina-
tion should be more useful. This is easily
derived as the product of dj and de. Thus,
a more complete measure of distinctness
(dc) for any internode or final branch m
may be given as

dcm = df.ndem = Sm [ S Nj-M (3)

Once such a measure of distinctness
has been obtained for each branch, it is a
simple matter to construct a matrix pre-
senting the taxonomic distinctness (dt) of
each taxon from every other taxon. This
is done by summing the distinctness values
for each branch linking the two taxa (X
and Y) under consideration. This may be
expressed as

dtx,Y

p

^=- 2 dCm

m = l

r / ^ \ 1

Sm 2 Nj-M

ra = l

- \ j 1 / '".

(4)

where p is the number of internodes and
final branches between taxon X and taxon
Y including the most recent common an-
cestor of X and Y, and the other symbols
have the same meaning as before. These
final figures are essentially weighted mea-
sures of total "patristic diflference" (Far-
ris, 1967) or "number of steps" in the
sense of Camin & Sokal (1965). They
may be used in judging appropriate levels
of distinctness for allocating the various

taxa to different categorical levels (such
as genus, family, etc.) in the final clas-
sification.

There is a complication introduced into
the present study as a result of differential
evolution of the two sexes. In some in-
stances a particular derived state is present
in one sex only although it could poten-
tially occur in both. Such a derived state
is obviously not as effective in distinguish-
ing between taxa as is one which does not
exhibit such dimorphic development, and
thus should not be given equal weight in
the calculation of a distinctness measure.
(This of course does not deny that such
a character state may be equally valid in
a cladistic sense as a reflection of signifi-
cant genetic modification as is one which
occurs in both sexes.) For states appear-
ing on the tree at least once in both
sexes, it thus seems logical to score each
occurrence of the state in one sex only as
0.5, and when affecting both sexes as 1.0.
Character states occurring on the tree in
only one sex or logically applicable to one
sex only (such as those of the female
genitalia) should also be coded as 1.0 since
there is no a priori reason to consider these
less significant than those appearing in
both sexes.

These adjustments (or any other type
of weighting to take similar factors into
account) invalidate the assumptions made
earlier that each state appearing on an
internode appears there once. The pre-
viously derived equations (1-4) must thus
be made more general. When this is done
it becomes obvious that k,n in equation 2
is not the same factor as that symbolized
by km in equation 1. The generalization
follows:

Assume no limitation on the number of
occurrences of each derived character state
j on any internode m. Now,

DF.

(5)

Phylogeny and Classification of the Aculeate Hymenoptera

581

where ij is the number of occurrences i
of each derived state j, and k is the num-
ber of derived states, both on internode
(or final branch) m; and Sm equals the
total number of species subtended b\' in-
ternode m, as before.

Since unitar\' occurrence of each de-
rived state on an\' one internode is now
not assumed, the measure of distinctness
based on efficiencies can no longer be based
directly on the harmonic mean of the
number of times each state appears on
the entire tree. Instead, the "efficiency"
of each state ] is directly calculated as the
ratio of the frequency of its occurrence on
the internode to its frequency on the en-
tire cladogram, and the mean of these
ratios for all k states is then utilized. The
limits of 0 and 1 still hold, as long as no
state is considered to occur less than once
in the entire tree. Thus,

DE,:

m = l

k.

(6)

\shere '>m is the number of occurrences of
derived character state ] on internode (or
hnal branch) m, km is the number of
derived character states (hatch marks) on
internode m (as before), and n is the
number of internodes (and final branches)
in the cladogram.

There is no longer any cancellation of
terms in the formula when the product of
DF and DE is found, so that DC cannfjt
be calculated directlv. Now,

DC„, = DF,„DE„,

(7)

The taxonomic distinctness of anv two
taxa, X and Y, is thus measured as

DT>

XY

2 DC,
111=1

(S)

where p is the number of internodes and

final branches between taxon X and taxon
Y, including the most recent common an-
cestor of X and Y.

These generalized ecjuations (5-S) are
those used in the present study. For in-
vestigations where the assumption of one
occurrence on any internode for each de-
rived state is valid, the earlier equations
(1-4) are easier to apply, especially since
DC can then be calculated directly.

A word is necessary to clarify the
method of determining the number of oc-
currences of a character state. There are
two aspects which may cause confusion.
First, a particular derived state is some-
times apparently reversed to the primitive
or a less derived condition. Since the
original primitive state cannot be placed
on any internode, the only appearances of
the "primitive" state will be the result of
such reversals. As such, it seems most ap-
propriate to consider these as "pseudo-
primitive" states and treat them like any
other derived state, merely totalling the
number of times that each appears on the
tree. On the other hand, a reversal to an
already derived state (e.g., from state C.2.1
to C.2) represents another appearance of
a state which is already placed elsewhere
on the tree. Although this actually repre-
sents a further derivation (in the above
example it might be numbered C.2. 1.1)
in a cladistic sense, it nevertheless is
equivalent to the less-derived state (C.2)
in a phenetic sense, and so should be
counted among the occurrences of that
less-derived state. The second problem
concerns instances in which a highly de-
rived state apparently arises de novo, there
being no evidence of an intermediate state
which is found elsewhere on the tree (e.g.,
state C.3.1.1 is derived directly from C.3
althouiih elsewhere on the tree it is de-
rived from (^.3.1). The usual practice in
constructi ntr the cladogram is to enter the
missing intermediate state on the inter-
node with its logical derivative. If there

582

The University of Kansas Science Bulletin

is absolute certainty that such an inter-
mediate is required for that subsequent
derivation to be possible, then this place-
ment of states is justified. Such certainty
is usually not possible, however, since a
similar highly derived state may often
logically have arisen in dilTerent ways
from difTerent intermediates. Some of
these intermediates may not be present
in the material examined, so that they do
not appear in the coding of character
states. Thus, inferring the occurrence of
a particular derived state on an internode
merely because of the presence of a state
which may logically be derived from it,
is probably unwarranted in most instances.
Such inferred derivations should thus not
usually be included when counting the
occurrences of a particular state.

It should be emphasized at this point
that the measures of taxonomic distinct-
ness (DT) should not be considered to
provide absolute criteria or definite nu-
merical values at which groups are auto-
matically considered distinct at some pre-
determined taxonomic level. (As Sneath
& Sokal, 1973, emphasize, groups delimited
in an analogous fashion on phenograms
by means of phenon lines "approach nat-
ural taxa more or less closely" but are
more appropriately termed "phenons" and
are not allocated levels in the formal tax-
onomic hierarchy.) Instead, these mea-
sures may be used as indices from which
some indication of the size of the phenetic
gap between two groups, as well as the
relative variation within a cluster of
groups, may be gauged. As is indicated
by Sneath & Sokal (1973), the most im-
portant criteria on which to base decisions
on taxonomic rank are probably the in-
ternal diversity of a taxon and the sizes
of the gaps between a taxon and those
adjacent to it. Despite the significance
generally attached to the "gap criterion"
by practicing taxonomists of the "evolu-
tionary" school (Michener, 1970), it is re-

jected by Sneath & Sokal (mainly because
a gap is almost incapable of being meas-
ured using the usual numerical phenetic
techniques). The distinctness measures
derived above seem to be interpretable in
terms of both criteria, however.

Because of the branching structure of
a cladogram, involving hierarchical levels
of dichotomy, use of the absolute number
of species for calculation of DF is not
appropriate. Instead, a multiplicative
(logarithmic or exponential) conversion

Fig. 66. Cladogram o£ 25 taxa of Aculeata, as in

Fig. 2, emphasizing the number of levels of

dichotomy.

Phylogeny and Classification of the Aculeate Hymenoptera

583

scale should be used. It may be proposed
that the logarithm to the base 2 is appro-
priate since this represents the number of
levels of dichotomy in a perfectly regular
ucc which gives rise to the number of
end points (species) involved. Cladograms
are practically never perfectly regular,
however. If one were able to count the
number of levels involved in an actual
cladogram, the number would almost in-
variably be greater than for a perfectly
regular branching pattern involving the
same number of species. For example, the
[iresent cladogram of the Aculeata (with
25 hnal branches) may be considered to
involve 9 levels of dichotomy (see Fig. 66
tor derivation) although a perfectly reg-
ular dendrogram with the same number
of final taxa would have about half that
number of levels (4 < logL'25 < 5). A
scale based on consideration of the num-
ber of levels of dichotomy is thus not
logically rec^uired. Furthermore, a loga-
rithmic scale has the effect of drastically
reducing the weight of a very large group
in comparison with the small ones. This
effect is not as extreme if an exponential
scale is used. In the present study some
groups were extremely large, and the cube
root of the estimated number of species
in each group was thus used for the values
of Sill. (The number of species in each
group is a best estimate based on the in-
formation contained in a wide variety of
papers cited elsewhere in this account.)

Values of 2i, k, S, DF, DE and DC
for the Aculeata appear in Table I, and
those of DT are shown in Table II. The
taxa are arranged from the base of the
tree to the apex, taking each branch in
turn.

The taxonomic distinctness measures
(DT) for the Aculeata range from 5 to
104(S (mean 454). In general, the long
diagonal in Table II contains the values
between the closest pairs on the tree except
where different branches are involved.

These values are all below the mean for
the study, except for two. These are the
values between the plumariids and sphe-
cids (566) and between the apids and
anthoboscids (713). Furthermore, the
values between all the basal taxa on the
branches represented by these taxa (con-
sidering in addition sphecids to sierolo-
morphids, 489) are greater than the mean.
This suggests a fundamental division into
three groups, these subtended by inter-
nodes 1-2, 4-5, and 4-6 respectively.

Within the first of these groups (con-
taining plumariids, bethylids and scoleby-
thids) the values of DT are all quite low
(mean is 77) although that between the
scolebythids and bethylids is much lower
than the others. This indicates relatively
slight internal variation within this group.
Furthermore, none of these values is much
more than one c]uarter of the lowest value
between an included taxon and a taxon in
one of the other large groups (389, plu-
mariids to sierolomorphids), again indi-
cating a large gap between the three
groups already identified.

The second group (sphecids and apids)
has somewhat greater internal variation
(246), but the value is only about half the
mean value for the entire study. Further-
more, this figure is also only about half of
the lowest value between a member of
this group and one of the others (489,
sphecids to sierolomorphids), reemphasiz-
ing the presence of a large gap.

The third group contains 20 taxa and
is more complex to analyze. The variation
within this group is quite considerable,
and the mean value of DT within the
group (339) is insignificantly less than the
minimal value between a member of this
group and one of the other two (389,
sierolomorphids to plumariids). The most
distinct taxon included here is the formi-
cids, which is more distinct from every
other taxon (minimal DT is 373, to
typhoctids, and mean is 507) than the

584

The UNivtRsiTY of Kansas Science Bulletin

mean value for the group. It is also
about as distinct within the group as the
group is from one of the other groups.
This suggests that the formicids may form
a fourth group of the same rank as the

three already recognized. The scoliids and
vespids are almost as distinct from other
members of this group as are the formicids
(minimal value is 323, vespids to typhoc-
tids) but less so from each other (183).

Table I. Occurrences (2i) and numbers (k) of derived states, number of species subtended (S) and
various distinctness measures (DF, DE, DC) fcjr the internodes and final branches of the cladogram

of Aculeata (Fig. 2).

Internode

k

DF

DE

DC

1-2
2-Plumar.

2-3
3-Bethyl.
3-Scoleb.
1-4
4-5
5-Sphecd.
5-Apids
4-6
6-7
7-8
8-Sapygd.

S-9

9-Myrmos.

9-Mutild.

7-10

10-Anthob.

10-11

11-Thvnnd.

11-12

12-13

13-Myzind.

13-Methoc.

12-14
14-Tiphid.
14-Brachy.

6-15

15-Sierol.

15-16

16-17

17-Pomiiil.

17-Rhopal.

16-18

18-19

19-Formic.

19-20
20-Vespid.
20-Scolid.
18-21
21-22
22-Typhoc.
22-Eotild.

21-23
23-Chypho.

23-24
24-Aptero.
24-Bradvn.

6.0

17.0
4.0
1.0

12.0
6.0

15.0
3.0

16.0
5.0
7.5

10.5
4.0

12.5
6.0
9.0
2.0
1.5
1.0
3.0
3.5
3.0
1.0

17.0

13.5
3.0

11.0
3.0

12.0
3.0
7.0
8.0

14.0
4.0
7.0

18.0
9.0

10.0

23.0

19.0
6.0
3.0
4.0

14.5
8.5

16.0
5.5

21.5

6

22
4
1

12
6

15
3

16
5
8

11
4

13
7
9
2
2
1
3
4
3
1

18

14
3

14
3

12
3
7
8

14
4
7

18
9

10

23

19
6
3
4

17
9

18
6

7?

10030

20

10010

10000

10

58500

32000

12000

20000

26500

7700

5200

150

5050

50

5000

2500

100

2400

1000

1400

600

500

100

800

700

100

18800

20

18780

3050

3000

50

15730

15500

12000

3500

3000

500

230

20

10

10

210

100

110

100

10

129.42

46.07

86.20

21.54

25.80

232.92

476.25

68.67

434.24

149.05

148.12

181.86

21.24

214.50

22.08

153.90

27.14

6.96

13.39

30.00

39.16

25.29

7.94

78.88

125.28

26.64

51.04

79.77

32.52

79.68

101.50

115.36

51.52

100.32

174.51

412.02

136.62

144.20

182.62

116.47

16.26

6.45

8.60

86.13

39.44

76.64

25.52

46.22

0.41

0.42

0.78

0.40

0.78

0.69

0.63

0.48

0.49

0.80

0.46

0.56

0.66

0.54

0.55

0.52

0.38

0.31

1.00

0.21

0.63

0.67

0.33

0.48

0.25

0.68

0.36

0.33

0.33

0.67

0.39

0.39

0.37

0.31

0.40

0.57

0.77

0.55

0.57

0.50

0.52

0.26

0.38

0.54

0.42

0.75

0.61

0.77

53.06

19.35

67.24

8.62

20.12

160.71

300.04

32.96

212.78

119.24

68.14

101.84

14.02

115.83

12.14

80.03

10.31

2.16

13.39

6.30

24.67

16.94

2.62

37.86

31.32

18.12

18.37

26.32

10.73

53.38

39.58

44.99

19.06

31.10

69.80

234.85

105.20

79.31

104.09

58.24

8.46

1.68

3.27

46.51

16.56

57.48

15.57

35.59

■'Estimated number of species (includmg those yet to be described). Calculations utilize \XS.

Phylogeny and Classihication of tiil Aculeate Hymenoptera 585

%.

'•^</^

Q

^S X 2S ^%

'<4

a
«^
rt
u

"5

<

u

h
Q

c

E

0

c

0
X

n
h

<

in

^ ON -.

^^^^ o o

^0.-> U-N ~ -^

^ t^ (^ U~\

<5^ . m r<-) o o
t^ r<-) in

^o-

t^

o.

o

r^

r^

^s-

^3-

n-

o

in

r^

rn

ro

TT

M-

•^

V

ro

i-O

-r

NO

CN

(N

^<f

OC

CN

r^i

r^

rn

in

' — 1

m

r<1

<N^

-T-

■^

V

CN

^*-

rn

in

NO

CN

CN

^^

•^

I^

r^

CN

OO

O

T

n-

rr.

ro

n-

■<r

in

>i-.

'^C

■<i-

rl-

CN

CO

O

_

OO

oo

flf

CN

■<r

NO

in

NO

F— .

NO

oo

ro

m

ro

r—t

C^l

CN

CN

\

Oc/

rl-

^'

o

in

n-

NO

t^

m

rn

NO

CN

I^

CN

OO

OO

<^

ON

1— (

n-

CO

rn

^^

1— (

CN

CN

ro

V-

CN

ro

O

o

^r

-1-

vn

r^

rn

m

'<$^

•<r

CN

o

in

r^

NO

NO

•— •

t^

ON

' — '

' — '

■^r

rO

ro

•— '

C^l

CN

CN

■%..

ro

o

'J-

NO

in

r^

cc

n-

t-

^^

<3

r<-l

C?

OC'

'O

m,

•t-

-t-

CN

in

t^

,

^r

in

m

i^

r^

t^

in

in

in

NO

NO

X.

C^i

in

m

NO

r<-i

^o

oo

NO

oo

CN

NC

NO

'^

ON

ro

NO

rn

^~^

NO

oo

t^

r^

l>l

OO

o

f^i

n-

T

t^

NO

NO

'T

n-

in

in

NO

^o.

V'-

(N

O'

.—1

oo

CN

CN

CN

"3-

ro

in

NO

rn

rn

'^S'

TT

-^

CN

•^

CN

CN

Tl-

t^

NO

NO

r^

a\

r-^

CN

CN

ro

ro

in

m

in

ro

ro

•^

TT

•^

<^-..

^t^'^

f

OO

NO

T'O

O

Tf-

CN

f— <

NO

in

t^

OO

in

in

<JV

^-i

(N

CN

o

ro

O

oo

r'-,

in

'3-

■^

ON

in

r^

\

CN

ro

rn

CN

ro

ro

in

m

in

ro

ro

m

■^

■<r

'^

>o

^r

oc

NO

m

O

_fc.

__^

NO

m

r~

oo

■d-

■^

\

ro

C^l

CN

o

ro

cb

oo

f^,

in

•^

■<r

ON

in

r^

CN

ro

ro

CN

(^

oo

in

m

m

ro

m

OO

•^

■^

%

•^

rj-

CN

ro

OC

NO

o

r^

NO

o

CN

m

o

o

"•>y/..

C2

C>

r<-i

o

o

ro

^—

oo

r^

3

in

in

o

NO

oo

'r

(N

CO

■>J-

CN

oo

c^

in

in

in

m

m

•<r

■^

■^

%

f/.

o

CN

CN

»--

oo

NO

'■'-)

o

'J-

C-l

'— '

NO

in

r--

oo

in

in

•<r

vo

vo

OO

ON

NO

I^

o

r^

in

o

(N

— ^

.— «

NO

CN

■^

CN

rf\

.— .

ro

CN

in

m

in

ro

m

rn

T

M-

%,

X

a

^s

o

o

cc

in

ni

NO

"^t-

<-o

oo

J^

CN

o

:^

t^

in

oo

oo

oo

■T-

NO

(^4

ro

NO

rn

^^

NO

oo

t^

r^

m

oo

o

CN

(Vl

^-1

CN

OS

m

rr

•>r

CNl

CN

m

rn

■^

o

in

o

OO

<^)

oo

in

CN

NO

NO

o

O

rn

CN

ON

\o

c^

5i

CN

CN

^r

_

.— i

T-

^-

CN

*7~

t^

NO

NO

.— .

NO

<X3

CN

ro

— '

CN

CN

1-

"7-

'T-

CN

CN

rn

m

fO

cc

^o

oo

oo

O

O

cc

CN

NO

O

oo

r-^

CN

_

rn

•^

^

•—4

VC

o

c\

CN

(__i

ro

CN

NO

ON

t^

-7-

CN

c-^i

— ^

~^

NO

CN

M-

r^

oo

t-^

t^

OO

CN

CN

NO

t^

l^

o

c^

o

OO

OO

oo

OS

ON

^^ vc .-, o

■Vo ~ "^ "~

J ri ir, iTv

"

oc

-r

a:

CC

NO

O

OO

ON

NO

o

oo

I--

CN

,—,

rn

^r

.— t

<— 1

cc

r^i

_

. — .

ro

in

— ^

oo

•— •

CN

NO

— «

T-

m

rn

oo

■^

NO

in

>o

^c;

o

NO

1^

CO

•^

NO

in

OO

oo

00

NO

NO

NO

t^

t^

%>^ vc .c -. =
/> vc -7- rn m

-''/ ir\ t-^ -7- ^

a:

1-

3C

oo

VO

oc

CN

I^

oo

t^

(^1

CN

rn

-7-

oc

(^J

^^

ro

in

OO

— ^

CN

NO

.— 1

-T-

r<-)

rn

OO

"7-

NO

rr

in

in

in

m

NO

t^

ro

m

T-

I^

t^

r^

in

m

in

^

NO

</^ ir\ 'o n c\ r~H

^^/o Cs (^1 C3 OO O

V \,D ic TT in

in

o

in

in

"N

1 ~

in

NO

ro

t^

1-

CN

OO

o

r^

I^

rr

oo

r^

r^

CN

o

t^

-7-

f^

^7-

C^J

t^

CN

OO

CN

n-

CN

»-H

in

in

in

in

in

t^

t^

■7-

in

in

OO

I^

r^

in

m

NO

NO

r~

vDr> CN r-, — -7- — oc

O Csl CZ> I-<1 CZ3 —

vc

r^I

vc

'O

•T-

oo

NO

r^

in

ON

NO

in

,

o

,

r^i

CN

c^

in

oc

oc

o

— •

OO

ir

oo

in

rn

oo

— ^

o

o

in

O

(N

■-< o cc m in

m

in

in

in

NO

r^

r~.

Tt-

in

m

OO

t^

oo

NO

NO

NO

t^

t^

586

The University of Kansas Science Bulletin

These two may thus also form a separate

group.

The remaining taxa form a paraphy-
letic group and show a wide range of
variation, although no one taxon is mini-
mally as distinct from any other as are
the formicids, vespids and scoliids. Thus,
the basal taxa on each of the three major
branches involved have the following
values: anthoboscids to sapygids — 128;
anthoboscids to sierolomorphids — 118;
sapygids to sierolomorphids — 221. This
indicates that these branches are not
highly distinct, these values all being less
than the mean value for all included taxa
(280). When members at the extremes of
the included branches are considered, how-
ever, it is seen that some of the taxa are
highly distinct from each other (e.g., mu-
tillids to bradynobaenids, 674). Further-
more, each pair of taxa along these
branches does not show great differentia-
tion of its members, so that at no point is
it possible to identify a further distinct
group. Since a paraphyletic taxon should
logically only be recognized if it is ap-
proximately as homogeneous as other rec-
ognized holophyletic taxa at the same
categorical level, the extreme variability
of this group suggests that it should not
be recognized as such. Further subdivi-
sion is, however, impossible because of the
absence of internal gaps. If the formicids,
vespids and scoliids are again added in,
it is found that the total variation within
the group is only increased by about one-
sixth (maximum is now 7S1, mutillids to
formicids). This produces a holophyletic
group only slightly more variable than the
paraphyletic one, so that the recognition
of this large holophyletic group is prob-
ably the best course of action. The Acu-
leata is thus considered to comprise three
subgroups of equal categorical rank,
which may be designated as superfamilies,
viz., Bethyloidea, Sphecoidea and Vespoi-
dea. This is somewhat reminiscent of the

suggestion by Bradley (1958) who pro-
posed two superfamilies, Vespoidea (in-
cluding those taxa here placed in the
Bethyloidea) and Sphecoidea.

Within the Bethyloidea, in the absence
of any detailed studies of the included taxa
other than those of previous workers, it
seems best to maintain the commonly ac-
cepted family groups (cf., Maa & Yoshi-
moto, 1961), Bethylidae, Scolebythidae,
Cleptidae, Chrysididae, Loboscelidiidae,
Dryinidae, Embolemidae, Sclerogibbidae
and Plumariidae. (Krombein, 1957 & in
litt., has suggested that the Cleptidae and
Loboscelidiidae may more appropriately be
considered to fall within the Chrysididae.)
Various of these taxa contain few species
with apparently relict distributions, prob-
ably indicative of a remote origin. For
example, recent Scolebythidae occur in
Brazil and Madagascar (Evans, 1963), and
Plumariidae (including Heterogyninae)
occur in arid areas in South America,
South Africa and the Mediterranean area
(Rhodes) (Bradley, 1972; Nagy, 1969a).

Within the Sphecoidea the two main
taxa have been treated differently in the
past. The wasp component has recently
been considered to comprise a single fam-
ily, the Sphecidae (Leclercq, 1954; Evans,
1964b) while the bees comprise nine fam-
ilies (Michener, 1965; Rozen, 1965; Rob-
erts, 1973). Since both groups contain
comparable numbers of species and appear
to include an approximately equivalent
range of phenetic variation, it seems rea- ""
sonable to subdivide both groups to a
similar degree. If this is done, the Sphe-
coidea may be considered to contain either
two families (one of wasps and one of
bees) or two groups each consisting of a
number of families. Since an important
function of a classification involves infor-
mation retrieval, the size of the group
involved is of some concern, for efficiency
is lost if a group is very large. Based on
this consideration, it is suggested that the

Phylogeny and Classification of the Aculeate Hymenoptera

587

eight subfamilies of sphecids recognized
by Evans (1964b) be considered valid taxa
at the family level, and that the presently
accepted bee families be retained at that
level. The Sphecoidea then consists of the
Ampulicidae, Sphecidae, Larridae, Melli-
nidae, Pemphredonidae, Astatidae, Phi-
lanthidae, Nyssonidac, Colletidae, Halicti-
dae, Oxaeidae, Andrenidae, Melittidae,
Fideliidae, Megachilidae, Anthophoridae
and Apidae. The first eight may be in-
cluded in an informal grouping, the Sphe-
ciformes, and the last nine in the Api-
formes. (It is of some interest that, despite
his suiTnestion in 1944 that the bees should
be placed "as a division of the Sphecoidea,"
Michener has retained the superfamily
"Apoidea" in his subsequent papers, illus-
trating that some collective term for "bees"
is necessary.)

Despite the general absence of marked
gaps within the Vespoidea, recognition of
subtaxa, mainly on the basis of areas of
low distinctness, is possible. Thus the six
taxa above internode 7-10 show generally
low mutual distinctness and a slightly
higher minimal value to surrounding taxa.
! They may thus logically be recognized as
I a single taxon at the family level with the
component taxa designated as subfamilies.
The Tiphiidae thus comprises the Antho-
boscinae, Thynninae, Myzininae, Metho-
chinae, Tiphiinae and Brachycistidinae.
Although many of the values are some-
what higher for the five taxa subtended
by internode 18-21, the gaps between any
of these and any member of the larger
group are considerably greater. Here
again it is probably appropriate to recog-
nize these taxa as a group at the family
level. The lowest value in the matrix is
that between two members of this group,
a figure so small (5) that the question
arises as to whether recognition of these
taxa as separate is justified. In the absence
of convincing evidence to the contrary,
recognition of these members at a very

low level is probably appropriate. The
Bradynobaenidae thus comprises the Ty-
phoctinae (with tribes Eotillini and Ty-
phoctini), Chyphotinae, Apterogyninae
and Bradynobaeninae.

The only other relatively low values
(less than 100) are those between the
pompilids and rhopalosomatids (64) and
the myrmosids and mutillids (92). In
both instances consideration should be
given to inclusion of both taxa in the
same family. In the case of the rhopaloso-
matids and pompilids this does not seem
appropriate since the habits of the two
groups are completely different (rhopalo-
somatids develop as ectoparasites of active
crickets and pompilids store paralyzed
spiders as larval provisions). The myrmo-
sids and mutillids, however, have ex-
tremely similar habits, both being para-
sitoids of the prepupae or pupae of other
Hymenoptera (see Knerer, 1973, e.g.).
This suggests that they may be considered
to fall in a single family, the Mutillidae
(see section on Mutillidae below for de-
tails).

Each remaining taxon is best con-
sidered as valid at the family level at the
least. The vespid group may additionally
be considered to comprise three families
(Masaridae, Eumenidae, Vespidae) as
recommended by Richards (1962). Since
the formicid group is highly distinctive
within the Vespoidea, it also seems appro-
priate to recognize the Formiciformes in-
formally to contain the family Formicidae,
the remainder of the Vespoidea being the
Vespiformes. (Incidentally, it is of some
interest that a form more similar to
Sierolomorpha than to Anthobosca seems
to be indicated as the ancestral type that
gave rise to ants; Anthoboscinae have in
the past been considered to be the closest
to the ancestral form, Wilson, Carpenter
& Brown; 1967.)

The superfamily Vespoidea is therefore
considered to contain the following 12

588

The University of Kansas Science Bulletin

Vespidae Eumenidae
Masaridae

Apidae

Anthophoridae

Megachilidae
Fideliidae

Formicidae

Mutillidae
Sapygidae

Nyssonidae
Philanthidae

Astatidae

Pemphredonidae

Mellinidae
Larridae

Scolebythidae
Bethylidae

Cleptidae
Chrysididae

Loboscelidiidae

BETHYLOIDEA

Fig. 67. Phylogeny of the families of Hymenoptera Aculeata (relationships within the Vespoidea from the

present study; within the Sphecoidea from Michener, 1974, for the Apiformes, and Evans, 1964b, for the

Spheciformes; within the Bethyloidea from the present study and personal impressions).

PnVLOGliNY AND ClASSI 1-ICATION OV THE AcULEATE HyMENOPTERA

589

families: Tiphiidac, Sapyi;itlac, Mutilli-
dae, Sicrolomorphidae, Pompilidae, Rho-
palosomatidae, Scoliidae, Masaridae, Eu-
menidac, Vespidae and Bradynobaenidae
in the Vespiformes, and Formicidae in the
Formici formes.

The phylogeny of the Aculeata to the
level of family thus appears as in Fig. 67.

INVESTIGATION OF THE

MYRMOSID-MUTILLID COMPLEX

(MUTILLIDAE)

The methods used in the study of the
myrmosid-mutillid group were similar to
those utilized in the study of the entire
Aculeata. Initially, all available specimens
(approximately 10,000. in about 1250 spe-
cies and including representatives of 89
per cent of the valid described genera and
subgenera as well as many undescribed
genera) were surveyed for about 50 char-
acters, mainly ones which had been used
previously to differentiate members at the
generic and suprageneric levels. (The ma-
terial examined is summarized in Table
III. This is actually a highly conservative
estimate since the extensive collections of
many museums on four continents were
also surveyed, but only the most signifi-
cant specimens were borrowed and thus
included in the figures given here.) This
survey demonstrated that many characters
were highlv variable, and these were dis-
carded. New characters were discovered
and incorporated. On the basis of the ini-
tial survey, and considering published
views such as those of Bischoff (1920-21)
and Schuster (1947, 1949), members of 135
genera and subgenera of mutillids were
tentatively grouped into 16 suprageneric
complexes and these were surveyed with
respect to 46 characters. At this point
various genera (or subgenera) showed
identical distributions of the character
states and many such superfluous mem-
bers were discarded, one genus generally
being retained for each such set of con-

cordant states. Wagner trees (by elec-
tronic computer) and cladograms (by
hand) were constructed, and characters
having states with many independent
origins or reversals were discarded. Addi-
tional characters were discovered and
added, and further groups showing con-
cordance of character states were reduced
to representative genera. Representatives
of three genera of myrmosids were added
and a few additional characters were intro-
duced as a result.

The machine- and hand-derived clado-
grams were then compared and refined as
a few more characters were added, elimi-
nated or coding was modified. Finally,
identical cladograms based on 43 charac-
ters (involving 61 derived states) were
derived both by the machine and by hand.
When the characters were allocated to
two data sets, one containing the 20 char-
acters of females and the other those of
males, and cladograms were derived using
these data sets independently, the result-
ing schemes were fully compatible al-
though not identical because of the lack
of characters for females on two of the
internodes. A final check was made that
the character states differentiating each
taxon on the tree were present in all the
genera and subgenera represented by that
taxon.

States of the Characters Considered

The 43 "best" characters used in the
analysis of the myrmosid-mutillid complex
are listed below, with comments on their
evolutionary patterns. This is followed
by a listing of the characters which were
eliminated during the analysis, with the
reasons for their elimination. In both cases
characters of females are listed before those
of males, the applicable sex being indi-
cated by the prefixes F (female) and M
(male). Even in cases where characters
are apparently identical in the two sexes,
they have been considered separately be-

590

The University of Kansas Science Bulletin

Table III. Genera and subgenera of Mutillidae, including the approximate number of species examined.
(* indicates that the type species was seen. Information on synonymies and new sex associations will be

presented elsewhere.)

MYRMOSINAE Mvrmosa (3 spp.,*; S ,9 \ ii'cl. Ischioceras)

M>rmosina (1 sp.,*:S )

Myrmosula (2 spp.*: S ,9 )

Protomutilla (1 sp.\9)
Not examined: Krombeinella {iiul. Paramyrmosa Stidrez), Leiomyrmosa, Morysmula, Paramyrmosa

SclllSStdT.

PSEUDOPHOTOPSIDINAE

Pscudophotopsis (10 .<pp.;S-9; incl. Alloneurion, Ephutomma,
Sphinctomutilla Andre, 1899)

TICOPLINAE

Not examined: Ticoiila.

Areotilla (5 spp.*\$ ,9 )
NanomutiUa (3 spp.,*\ $ ,9)
Smicromyrmilla (17 spp.; $ ,9)

RHOPALOMUTILLINAE

Rhopalomutilla (13 spp.*\$.9)

SPHAEROPTHALMINAE
DASYLABRINI

Not examined: Craspedopyga.

Apteromutilla (4 spp.*:i ,9 ; incl. Aptcrotilla)

Brachymutilla (7 spp.*: S .9 )

Chrcstomutilla (5 spp.*: S ,9)

Dasylabris (28 spp.*; $ ,9 : incl. Allomutilla)

Dasylabroides (21 spp.*; $ ,9)

Seyrigilla (1 sp.*;9)

Stenomutilla (16 spp.*; $ ,9 : incl. Xenomutilla)

Tricholabiodes (H spp.*; S ,9)

SPHAEROPTHALMINI
PSEUDOMETHOCINA

Ancipitotilla (1 sp.,*; $ )

Anomophotopsis (1 sp.*:$)

Atillum (9 spp.*; $,9)

Calomutilla (3 spp.*;9)

Darditilla (10 spp.:S ,9)

Dimorphomutilla (6 spp.: S ,9 )

Euspinolia (1 1 spp.,*; S ,9 : incl. Rcedia .-Is/imcad)

Gurisita (2 spp.*; 9 )

Hoplocrates ( 15 spp.*; S ,9 ; incl. Hoplomutilla Andre)

Hoplognathoca (3 spp.,*; 9 )

Hoplomutilla Ashmcad (18 spp.,*: i ,9 ; incl. Tilluma)

Horcomutilla (6 spp.,*: 9 )

Invreiella (3 spp.,*:9)

Jamaitilla (I sp.,*:i ,9)

Lynchiatilla (4 spp.; S ,9 )

Mammomutilla (I sp.,*;$)

Mickelia (1 sp.:9 ) " '

Myrmilloides (1 /p.,*;(5,9)

Pappognatha (5 spp.; S ,9 )

Patquiatilla (Isp.,*: 9 )

Pertyella (6 spp.,*; 9 )

Pseudomethoca (23 spp.,*; $ ,9 : incl. Nomiaephagus)

Seabratilla (1 sp.,*;9 )

Sphinctopsis (33 spp.,*;$ ,9 ; incl. Sphinctomutilla Andre, 1909)

Tallium (13 spp.,*: $,9)

Vianatilla (I sp..*;9)

Not examined: Allotilla.

Phylogeny and Classification of the Aculeate Hymenoptera 591

Table III. Genera and subgenera of Mutillidae, including the approximate number of species examined.
(* indicates that the tyjic species was seen. Information on synonymies and new sex associations will be

presented elsewhere.) (Continued.)

SPHAEROPTH.\LMIN.-\ Acanthophotopsis (2 spp.,*; $ )

Acrophotopsis (1 sp.*;$)

Ascetotilla (8 spp.*;i,9)

Bothriomutilla (1 sp.*\ S ,9)

Ccphali>mutilla (7 spp.; $ )

Ceratophotopsis (2 spp.,*\$)

Cystomutilia (2 spp.*; $,9)

Dasymutilla (110 spp.*; <^ , 9 : ind. Bruesia, Pycnomutilla)

Dilophotopsis (6 spp.*; c5 , 9 )

Ephutomnrpha sensii stricto (6 spp.,*; $ )

Eurymutilia ( 1 1 spp.*; S ,9 )

Huacotilla (I sp.;9 )

Leucospilomutilla (1 sp.,*; $ ,9 )

Limaytilla (4 spp.*;S ,9)

Lomachaela (6 spp.*; c? , 9 )

Lophomutilla (3 spp.,*; 9 )

Lophostigma (6 spp.; 9 )

Micromutilla (7 spp.,*; 6,9)

Morsyma (3 spp.*; £ ,9 )

Nanotopsis (1 sp.,*; $ )

Neomutilla (3 spp.,*; $ ,9 )

Odontophotopsis (16 spp.,*; $ ; inch Tetraphotopsis)

Paramutilla (2 spp.}*};$,9 ?)

Periphotopsis (1 sp.,*; $ )

Photomorphina Schuster, 1952 (8 spp.; $ .9 ; ind. Photomorphina
Schuster, 1958)

Photomorphus (3 spp.,*; $ ,9)

Photopsioides (5 spp.*; i ,9 )

Physetapsis (1 sp.,*; $ )

Photopsis (13 spp.,*; $ ,9 '■, ind. Agama Blak.e, Neophotopsis,
Pyrrhomutilla)

Protophotopsis (2 spp.,*; $ ,9 )

Ptilomutilla (4 spp.,*; 9 )

Reedomutilla (4 spp.,*; S ,9 ; ind. Reedia Andre)

Scaptodactyla (6 spp.*; S ,9)

SmicromutiJla (1 sp.,*; $ ,9 )

Sphaeropthalma (2 spp.,*; S ,9 : ind. Sphacrophthalma)

Suarcztilia (6 spp.; $ ,9)

Tobantilla (4 spp.; S -9 )

Traumatomutilla (31 spp.,*; $ ,9)

Xenophotopsis (1 sp.,*; S )

Xystromutilla (6 spp.; S ,9)
Not examined: Chasciuitilia, Protophotopsieila, Xenomorphus.

In addition to other scattered tindesa'ibed genera, representatives of at least 20 undescribed genera
from the Australasian region, mainly Sphaeropthalmina but a few Pseudomethocina, were e.xamined.

MYRMILLINAE Blakeius (2 spp.,*; $ ,9 \ md. Bisigilla)

Ceratotilla (9 spp..*;$ ,9 )
Clinotilla (1 sp.,*;$,9)
Labidomilla (12 spp.,*; $,9)
Liomutilla (1 sp.*; 9 )
Liotiila (1 sp.,*;9)
Myrmilla (10 spp.,*; S ,9 ; ind. Edrionotus, Eurygnathilla,

Pseudomutilla, Rudia)
Myrmotiila (1 sp.,*; 9 )
Odontotilla (9 spp.,*;S ,9)
Platymyrmilla (1 sp.,*; $ ,9)
Pygomilla (6 spp.,*; 9 )

592

The University of Kansas Science Bulletin

Table III. Genera and subgenera of Mutillidae, includinsj the approximate number of species examined.
(* indicates that the tyjie species was seen. Information on synonymies and new sex associations will be

presented elsewhere.) (Concluded.)

Sigilla (1 .<p.*;S,9)
Spilomutilla (3 spp.: S ,9)
Squamulodlla (26 spp.; S ,9)
Viereckia (0 spp.*: $,9)
Not examitjt'd: Omotilla, Saganotilla.

xMUTILLINAE
MUTILLINI
MUTILLINA

Not exaniuitd: Physetopoda.

Hadrotilla (1 sp..*:9 )

Mudlla (23 spp.*; £ ,9 \ incl. Barymutilla, Pycnotilla, Ronisia)
Odontomutilla (40 spp.: i ,9 : iucl. Pcringueya, Radoszkowskius)
TropidotiUa (4 spp.*:S .9 )

SMICROMYRMINA

Not e.xamined: Hildcbrandtia,
jails here.

Antennotilla (4 spp.*: $ )
Artiotilla (1 sp.*:9)
Aureotilla (2 spp.*:$ ,9)
BisuicotiUa (1 sp.*:S)
Carinotilla (3 spp.: $ ,9)
Chrysotilla (4 spp.; 9 )
Corytilla (3 spp.; $,9)

Ctenotilla (15 spp.*; $ ,9 ; i"cl. Cephalotilla, Pseudocephalotilla)
Dolichomutilla (7 spp.*; $ ,9 )
Eremomyrme (6 spp.; $ )
Glossotilla (16 spp.*; $,9)
Gynandrotilla (2 spp.; S )
Lophotilla (4 spp.;S ,9 )
Mimecomutilla (3 spp.,*: $ ,9)
Pristomutilla (10 spp.; S -,9 )
Promecilla (18 spp.,*; $,9)
Psammotherma (2 spp.,*; $ )
Smicromyrme (100 spp.,*; $ ,9)
Spinulotiila (6 spp.,*;S .9)
Sulcotilla (3 spp.,*:S ,9 )

TimuUa (175 spp.,*; i ,9 ; i"il. Lobotilla, Trogaspidia)
Trispilotilla (5 spp.: S .9 )
Rhombotilla, Sylvotilla, Zeugomutilla. Chactotilla prohahly also

EPHUTINI

Not examined: Arcasina.

Ephuamclia (1 sp.;S)

Ephuchaya (1 sp.,*;£ )

Ephuseabra (1 sp.,*:S )

Ephusuarezia (2 spp.; S ,9 ■")

Ephuta (50 spp.,*; d ,9 : imi. Ephutopsis)

Xenochile (I sp.;^)

cause the distribution of the various states
often differs in the females from that in
the males. (See below for treatment in
calculations, however.) Some of the less
useful characters that had been used in
the study of the Aculeata and which dif-
ferentiate the myrmosids and mutillids

(e.g., subtle differences in the form of the
posterolateral angle of the pronotum, ex-
tent of the mesopleural sulcus, length of
the metanotum, size of the seventh meta-
somal sternum and hypopygium, all in
the male) were not included here at any
stage. The system of coding for the vari-

Phylogeny and Classification of the Aculeate Hymenoptera

593

OLis States is as for the investigation of the
Aculeata. The names of taxa are those
used in the final chissification resulting
from consideration of these data. (The
final cladogram appears after the figures
of characters, as Fig. 92.)

Characters Utilized for Derivation
OF the Final Cladogram

Fl. Ocelli. Primitively, the ocelli are well-
developed or at least readily distinguish-
able. Fl.l — The ocelli are completely ob-
literated and indistinguishable.

Presence of ocelli is considered primi-
tive because this is the condition found
generally throughout the Aculeata includ-
ing the taxa most closely related to the
Mutillidae.

Loss of the ocelli in apterous forms is
common throughout the Insecta. Such
loss is characteristic of all Mutillidae ex-
cept some Myrmosinae (ocelli present in
species of Myrmosa) and a few species
of Pseitdophotopsis (Pseudophotopsidi-
nae). Thus state Fl.l has been derived
within the Myrmosinae and at least twice
within the remaining Mutillidae, once
within Pseitdophotopsis and once on the
line leading to all other Mutillidae.
Nevertheless, the derived state appears
only once on the tree, on internode 2-3.
In view of its multiple derivations, state
Fl.l cannot be considered strong evidence
for the holophyletic nature of the Mutilli-
dae above Pseitdophotopsis, but indeed the
presence of the primitive condition in
many species of Myrmosinae and Pseudo-
photopsidinae serves to emphasize the
basal position of those taxa.

F2. Eye form. Primitively, the compound
eye is somewhat flattened in profile, merg-
ing smoothly with the surrounding cuticle,
and is also broadly oval in outline. F2.1 —
The compound eye is highly convex in
profile, strongly differentiated from the
surrounding cuticle and approximately

hemispherical, being more or less circular
in outline.

A flattened, ovate eye is considered
primitive because this is the condition in
most Aculeata and in particular those taxa
most closely related to the Mutillidae (viz.,
Fedtschenkiinae, Anthoboscinae).

Although there is a tendency for the
eye to be circular in a few groups where
it is also reduced in size (e.g., Rhopalomit-
tilla, Nanotvatilla), the specialized hemi-
spherical form of the eye has apparently
evolved only once, in the Sphaeropthal-
mini, and is characteristic of almost all
members of this tribe. In a few related
genera (e.g., Eitspinolia, T allium, Atillum,
Hoplocrates) the eye has become some-
what flattened once more although re-
taining some of the other characteristics
of the tribe such as the basically circular
shape. Despite this apparent partial re-
versal within the Sphaeropthalmini, state
F2.1 provides strong evidence for the holo-
phyletic origin of this taxon.

F3. Eye pubescence. Primitively, the com-
pound eye is pubescent, bearing setae set
into minute pores between the ommatidia.
The setae may be prominent at relatively
low magnification or they may be essen-
tially indistinguishable although the pores
may be discerned. F3.1 — The compound
eye is completely glabrous and all pores
are lost.

A pubescent eye is considered primitive
because most groups of Aculeata have eyes
which are either obviously pubescent or
which have well-developed pores bearing
minute setae. In particular this is true of
the Fedtschenkiinae and most Tiphiidae,
the taxa most closely related to the Mu-
tillidae.

Loss of pubescence and pores in the
compound eye has apparently occurred at
least twice, in the Pseudophotopsidinae
and on internode 3-4, so that the only
groups with the eye prominently pubes-
cent or with pores are the Myrmosinae

594

The University of Kansas Science Bulletin

and Ticoplinae. Within the latter group
Areotilla, Nanomiitilla and Ticopla all
have the eyes with dense and quite long
setae (as do the Myrmosinae), whereas
Stnicromynnilla has the setae reduced but
still possesses a few pores in most species.
The presence of the primitive state in the
Ticoplinae is thus significant in empha-
sizing the relatively basal position of this
taxon, rather than the derived state being
useful in defining holophyletic groups.

F4. Maxillary palpus. Primitively, the
maxillary palpus consists of six segments.
F4.1 — The maxillary palpus comprises
only two segments.

A six-segmented maxillary palpus is
considered primitive because this is the
condition in most Aculeata, including the
taxa most closely related to the myrmosid-
mutillid group, and also in practically all
members of this group.

Reduction of the maxillary palpus in
the female to two segments has occurred
in only one genus, Rhopalomittilla, and so
serves to emphasize the distinctness of the
Rhopalomutillinae rather than to establish
any higher groupings.

F5. Labial palpus. Primitively, the labial
palpus consists of four segments. F5.1 —
The labial palpus is reduced to two seg-
ments.

A four-segmented labial palpus is con-
sidered primitive because this is the con-
dition in almost all Aculeata, including
those taxa most closely related to the Mu-
tillidae, and also in most members of this
family.

Reduction of the labial palpus to two
segments in the female has occurred once
only, in the Rhopalomutillinae, thus not
providing any essential information on
higher level groups.

F6. For 771 of mesosoma (Fig. 68). Prim-
itively, the mesosoma has an approxi-
mately constant width although subtle
variations may involve depression of the

68

Figs. 68-69. Characters of Mutillidae. 68, meso-
soma of female, dorsal view, showing primitive and
derived states of form, with position of second spi-
racle arrowed (F6 based on Areotilla: F6.1 on
Sphaeropthalma: ¥6.2 on Rliopalomttttlla; F6.3 on
Ephiita); 69, prothorax, lateral view, showing derived
state of pronotal pit (F8.I, M24.I based on Psetido-
photopsis).

mesopleuron and/or lateral expansion of
the propodeum. Thus the mesosoma is
more or less parallel-sided, narrower at
the mesopleural level than elsewhere or
broader posteriorly than anteriorly. F6.1
— The mesopleuron is protuberant at
about half its length, predominantly as a
result of broad development of the meso-
pleural ridge which is dorsally separated

Phylogeny and Classification of the Aculeate Hymenoptera

595

from the prothoracic spiracle. The meso-
soma thus broadens behind the protho-
racic spiracle and then narrows from a
point some distance anterior to the second
mesosomal spiracle, the metapleural-pro-
podeal region being narrower than the
pronotLim. F6.2 — The metapleuron and
lateral face of the propodeum are some-
what depressed and the mesopleuron is
very slightly protuberant. The mesosoma
thus broadens gently and evenly from the
anterolateral angle of the pronotum to a
point at the approximate level of the sec-
ond mesosomal spiracle, behind which it
is abruptly constricted. F6.3 — The meso-
soma is oval with the mesopleuron very
slightly protuberant although the meso-
pleural ridge is undeveloped. The meso-
soma is thus evenly broadened from the
anterolateral region of the pronotum to
the mesosomal midpoint and then evenly
narrowed to the posterior region of the
propodeum which is about as wide as the
pronotum.

An approximately parallel-sided meso-
soma is considered primitive because this
is the condition in the Myrmosinae which
are considered to occupy a basal branch
on the cladogram on the basis of other
characters. It is also the condition in some
of the less-derived members of the Ticopli-
nae such as Areotilla and Nanotnittilla,
this being a group with many other char-
acters in the primitive states. The primi-
tive state also includes some variations in
mesosomal form that occur in various
genera but do not characterize particular
groups. The particular derived states con-
sidered seem to involve the most distinct
modifications, each having originated in-
dependently.

Some of the variations included in the
primitive type have occurred sporadically.
For example, the mesopleuron is some-
what depressed in Pseudophotopsis (Pseu-
dophotopsidinae), Sqiiamitlotilla (Myrmil-
linae) and Aiireotilla (Mutillini). amongst

others; the propodeum is broadened in
Smicroiyiynnilla (Ticoplinae), Labido-
tnil/a (Myrmillinae) and some Timulla
(Mutillini). State F6.1, by contrast, is
characteristic of a single group, the Sphae-
ropthalminae, having arisen on internode
5-6. It is thus a useful state establishing
this group as holophyletic. Unfortunately,
the strength of this state is slightly lessened
by the fact that the mesopleuron is some-
what similarly expanded in Odontomutilla
(Mutillini) in which, however, the pro-
podeum is not markedly narrower than
the pronotum. State F6.1 is also slightly
modified in the Pseudomethocina, a sub-
group of the Sphaeropthalmini, mainly by
a slight dorsal flattening and anterior dis-
placement of the mesopleural ridge.

State F6.2 is present only in Rhopalo-
7niitilla and state F6.3 is characteristic of
the various members of the Ephutini only.
These states are thus of minimal use in
establishing higher groups. Actually the
various forms of the mesosoma in the fe-
male are ditliicult to describe and are more
distinct than might be indicated by the
particular designation of derived states.
Thus the Pseudophotopsidinae, Myrmilli-
nae and Mutillini also have somewhat
characteristic forms of the mesosoma al-
though these are impossible to describe
and code adequately.

F7. Pro-mesonotal suture. Primitively, the
suture between the pronotum and meso-
notum is freely articulating and is approx-
imately straight. F7.1 — The pro-mesonotal
suture, although distinct, is fused and im-
movable and is also strongly curved so
that the posterior margin of the pronotum
is concave. F7.1.1 — The pro-mesonotal su-
ture is obliterated and indistinguishable
(at least dorsally) or is barely indicated
by a slight variation in sculpturing, when
it is seen to be strongly curved.

An articulating pro-mesonotal suture
is considered primitive because this is the
condition in other Aculeata and almost

596

The University of Kansas Science Bulletin

all male mutillids. An approximately
straight posterior margin to the pronotum
is also considered primitive because this
is the situation in those taxa most closely
related to the Mutillidae (viz., Fedt-
schenkiinae, Anthoboscinae).

Fusion of the pro-mesonotal suture has
occurred in the Aculeata only in the Mu-
tillidae and is characteristic of the entire
family except for the Myrmosinae. Strong
concavity of the posterior margin of the
pronotum is found in a number of the
generally more derived taxa in the Acu-
leata and this condition is also character-
istic of almost all the Mutillidae. Thus,
state F7.1 arose once only, on internode
1-2, and is a very strong indicator of the
holophyletic nature of the six higher sub-
families of the Mutillidae. Further ob-
literation of the pro-mesonotal suture
(F7.1.1) has occurred on internode 2-3,
thus providing further strong evidence of
the holophyletic origin of the remainder
of the family and emphasizing the low
position of the Pseudophotopsidinae
within the family. Although the position
of the suture is sometimes discernible
because of slight variations in sculpturing
(e.g., some species of Dasylabris, Neomu-
tilla), this condition is obviously more
advanced than that in all species of
Pseudophotopsis and cannot be confused
with it, despite the fact that a few pseudo-
photopsidines have the fusion more de-
veloped than others.

F8. Pronotal pit (Fig. 69). Primitively,
the pronotum is more or less evenly sculp-
tured without any specialized pitlike
structures. F8.1 — The pronotum has a
shallow pit at the ventral angle, which is
filled with setae forming a structure simi-
lar to that on the first metasomal tergum
in Paratiphia or the second metasomal
tergum in some species of Odontomittilla,
so that it may be analogous to the felt
lines of the second metasomal segment.
A simple pronotum is considered prim-

itive because this is the condition in prac-
tically all Aculeata, including the taxa
most closely related to the mutillids, and
also in most of the Mutillidae.

The development of specialized ven-
tral pits on the pronotum has occurred in
only one genus, Pseudophotopsis, so that
this is a strong character separating the
Pseudophotopsidinae from the remaining
Mutillidae but it is not of any use in

71

Figs. 70-71. Characters of Mutillidae. 70, mesosoma
of female, dorsal view, showing primitive and derived
states of length of pronotum, with position of second
spiracle arrowed (F9 based on Myrmosa; F9.1 on
LabidomiUa); 71, mesosoma of female, lateral view,
showing primitive and derived states of meso-meta-
pleural suture (FIO based on Myrmilla: FlO.l on
Smicromyrme) .

Phylogeny and Classification of the Aculeate Hymenoptera

597

establishing holophyletic groups at the
hiijher levels.

F9. Length of pronotum (Fig. 70). Prim-
itively, the dorsolateral margin of the pro-
notum is about as long as the distance
between the prothoracic and propodeal
spiracles. F9.1 — The dorsolateral margin
of the pronotum is much shorter than
(about two-thirds the length of) the dis-
tance between the prothoracic and pro-
podeal spiracles.

A long pronotum (laterally) is con-
sidered primitive because the pronotum
in Fedtschenkja and most of the gen-
eralized Aculeata such as the Anthobosci-
nae is quite long and well-developed.
Shortening of the pronotum has occurred
in various of the more derived groups of
the Aculeata.

Shortening of the pronotum in the
female has apparently occurred only once
in the Mutillidae, on internode 4-5, and
is thus characteristic of most members of
the family, excepting only the four most
basal subfamilies. This is thus quite a
good indication of the holophyletic origin
of the three higher subfamilies. Its
strength is, however, perhaps diminished
by the fact that a superficially similar
derived state has occurred within the Myr-
mosinae (e.g., in Myrmosnla), although
here the apparent reduction in size of the
[ironotum appears to have resulted from
a posterior displacement of the propodeal
spiracle. The similarity to state F9.1 may
thus be more apparent than real in this
case.

This measure for describing the lateral
shortening of the pronotum was chosen
for convenience, but it cannot express the
complex changes in proportion which oc-
cur in the mesosoma as a whole (such as
have apparently occurred in Myrmosula).
The change which has taken place on in-
ternode 4-5 is actually more distinct than
might be thought from the description
alone, and consideration of the mesosoma

as a whole strengthens the use of this
character.

FIO. Meso-metapleural suture (Fig. 71).
Primitively, the meso-metapleural suture
passes from a point just posterior to the mid-
coxa dorsally directly to the second mesoso-
mal spiracle, and is approximately straight.
This suture is thus also separated from
the dorsoventral mesopleural ridge. In
some cases the meso-metapleural suture is
slightly anteriorly curved and approaches
the mesopleural ridge, usually if this ridge
is dorsally separated from the prothoracic
spiracle and thus somewhat posteriorly
shifted. FlO.l — The meso-metapleural su-
ture passes anterodorsally from the mid-
coxa and becomes continuous with the
mesopleural ridge which is dorsally con-
tiguous with the pronotal spiracle. The
dorsal section of the meso-metapleural su-
ture (between the mesopleural ridge and
the second mesosomal spiracle) is at an
acute angle to the ventral section. In many
instances the dorsal section of the meso-
metapleural suture and the ventral portion
of the mesopleural ridge are obliterated so
that there is an apparently continuous su-
ture extending from a point just posterior
to the mid-coxa anterodorsally to the pro-
thoracic spiracle. [These modifications
were first noted by Reid (1941) who con-
sidered the mesopleural ridge to be the
prepectal suture. His interpretation of the
evolutionary sequence was also slightly
different since he considered the meso-
pleural ridge to be primitively separated
from the prothoracic spiracle.]

An approximately straight meso-meta-
pleural suture is considered primitive be-
cause this is the condition in most other
members of the Aculeata, including those
taxa most closely related to the Mutillidae,
and also in mutillid males, especially those
of the more generalized subfamilies (but
see character M26).

Strong angulation of the meso-meta-
pleural suture is characteristic of all genera

598

The University of Kansas Science Bulleti

N

of Mutillini except for OdontonuttiUa in
which this state is approached but where
the mesopleural ridge is separated from
the prothoracic spiracle dorsally, this ap-
parently secondary condition obscuring the
situation. State FlO.l is thus shown on
the cladogram on the branch leading to
the Mutillini despite the fact that it is
not definitely present in OdontonuttiUa
(which falls into the Mutillini by virtue
of a combination of different characters).
This state is also considered to be good
evidence for the holophyletic derivation
of the Mutillini, the significance of which
is increased by the fact that this is one of
the largest tribes of the family.

Fll. Mesosternal teeth (Fig. 72). Primi-
tively, the mesosternum bears a simple
transverse carina anterior to each mid-
coxa. Fll.l — The mesosternum has a pair
of toothlike projections, one anterior to
each mid-coxa, being elaborations of the
primitive carinae. Fll. 2 — The mesoster-
num is simple, without any protuberances
or carinae.

A mesosternum with simple carinae is
considered primitive because this is the
condition in the Fedtschenkiinae and is
the precursor to the state in the Tiphiidae,
these being the taxa most closely related
to the Mutillidae.

The presence of small dentate projec-
tions on the mesosternum anterior to the
coxae in females (Fll.l) is characteristic
only of Pseiidophotopsis, Since this is an
autapomorphic state, it does not provide
any data on higher groupings. By con-
trast, loss of the carinae (F11.2) is char-
acteristic of the females of Myrmosinae
and also of the remaining Mutillidae, the
latter derivation having occurred on inter-
node 2-3. Since this state has occurred
twice, it is not a strong indicator on which
to base higher groups. Furthermore, a
state somewhat similar to the primitive
one has been redeveloped in some of the
more highly derived members of the

Figs. 72-74. Characters of Mutillidae. 72, posterior
region of mesosternum, ventral view, showing prim-
itive and derived states of mesosternal teeth (Fll,
M25 based on Mynnosa, $\ Fll.l, M25.1 on Pseti-
dophotopsis. $; F11.2, M25.2 on Areotilla, $);
73, posterior region of metasternum, ventral view,
showing primitive and derived states of metasternal
process (F13, M28 based on Myrmosa, $ ; F13.1,
M28.1 on Dasyhihris, $, and Hoplocrates, $, top to
bottom; 74, hind coxa, showing primitive and
derived states (F14, M29 based on Pseiidophotopsis,
S: F14.1, M29.1 on Myrmosa, $ ; F14.2, M29.2
composite) .

Sphaeropthalmina (e.g., Ascetotilla and
some other genera in the ''Ephittomorpha
complex"), further weakening this state.

F12. Contiguity of mid-coxae. Primitively,
the mid-coxae are contiguous mesally.
F12.1 — The mid-coxae are slightly sepa-
rated and do not contact each other along
the midventral line.

Contiguous mid-coxae are considered
primitive because this is the condition in
the Myrmosinae and Pseudophotopsidinae,
which are the taxa considered to be the
most basal on the cladogram on the basis
of other characters.

Phylogeny and Classification of the Aculeate Hymenoptera

599

Separated mid-coxae are characteristic
of the entire MutiUidae except for Myr-
mosinae and Pseudophotopsidinae, so that
state F12.1 provides good evidence for the
holophyletic nature of the Mutihidae
above internode l-}>. The strength of this
character is, however, somewhat dimin-
ished by the fact that the middle coxae
show various degrees of contiguity or
separation throughout the Aculeata.

F13. Metasternal process (Fig. 73). Prim-
itively, the metasternum is simple and
flattened anteromesal to the hind coxa.
F13.1 — The metasternum bears a process
anteromesal to each hind coxa. This may
be a transverse carina, or the mesal ex-
tremity of this carina may be produced as
a tooth which may be fused with the pro-
cess of the opposite side to form a single
mesal projection on the metasternum.

A simple metasternum is considered
primitive because this is the condition in
the Sapygidae, the sister group of the Mu-
tiUidae. Although there is slight develop-
ment of a mesal longitudinal ridge in
Fedtschenl{ia, this is developed anterior to
the position of the protuberances involved
in state F13.1 and is furthermore obviously
not a paired structure. A simple meta-
sternum is also present in the Myrmosinae,
which forms a basal branch of the clado-
gram on the basis of other characters.

A dentate metasternum is characteristic
of the entire family MutiUidae except for
the Myrmosinae. Since this state is appar-
ently unique in the Aculeata, it provides
very good evidence of the holophyletic
nature of the group above internode 1-2.
This may, however, not be quite as strong
as at first considered since the metaster-
num is apparently somewhat plastic else-
where in the Aculeata and has been mod-
ified in various ways.

F14. Hind coxal tubercle (Fig. 74). Prim-
itively, each hind coxa bears a carinate
tubercle dorsally. F14.1 — Each hind coxa

bears a lamellate process dorsally. F14.2
— Each hind coxa is simple, without any
dorsal tubercle or lamella.

The presence of a carinate tubercle on
the hind coxa is regarded as primitive be-
cause this is the condition in the Fedt-
schenkiinae, which is the group most
closely related to the MutiUidae, and also
in most of the basal members of the Mu-
tiUidae (so considered in the light of other
characters).

Elaboration of the hind coxal carina
into a lamellate process (F14.1) is uniquely
characteristic of the Myrmosinae and is
thus an important character state differ-
entiating this group. Loss of the coxal
tubercle is often dilftcult to establish with
certainty since the tubercle is never very
highly developed even when obviously
present. Nevertheless, a hind coxal tuber-
cle is definitely present in Pseiidophotopsis,
Areotilla and Smicromyrjnilla. The small
size of Nanomutilla makes determination
very ditftcult, but a tubercle does appear
to be present in this genus and in Rhopalo-
miitilla also. Thus state F14.2 appears to
have evolved once only, on internode 4-5.
In view of the dil^culties involved in
determining the condition of the hind
coxa and since it seems likely that such
insignificant structures may have been lost
on various occasions, the presence of state
F14.2 should probably not be regarded as
of great significance in establishing the
holophyly of the MutiUidae above inter-
node 4-5. Instead, the presence of the
primitive state in three subfamilies serves
to emphasize the basal position of these.

F15. Tarsal claws (Fig. 75). Primitively,
each tarsal claw bears a sharp tooth about
halfway along the ventral margin. F15.1
— Each tarsal claw is simple, the ventral
tooth having been lost.

A toothed tarsal claw is considered
primitive because this is the condition in
most Aculeata and in particular those taxa
most closely related to the MutiUidae, a

600

The University of Kansas Science Bulletin

Figs. 75-76. Characters of Mutillidae. 75, claw,
showing primitive and derived states (F15, M30
based on Pseiidophotopsis, 9 : F15.1, M30.1 on
Miitilla. 2; M30.1.1 on Rhopalomtitilla. $); 76,
base of metasoma, dorsal view, showing primitive
and derived states of form of first segment (F17,
M37 composite; F17.2, M37.1 based on Ephtita, 9 ;
F17.1 on Rhopalotniitilhi, 9 ).

single tooth being present in Fedtschenki-
inae and many Tipiiiidae.

Simple tarsal claws in the female are
characteristic of the entire subfamily Myr-
mosinae and all other Mutillidae except
for the Pseudophotopsidinae, having been
derived in the latter instance on internode
2-3. State F15.1 has thus arisen on two
occasions within the myrmosid-mutillid
group, and also elsewhere in the Aculeata;
its presence is thus not particularly good
evidence for the association of most of the
Mutillidae into a single holophyletic
group. Instead, the presence of the prim-
itive condition in Pseitdophotopsis again
serves to emphasize the basal position of
this taxon.

F16. Arolium. Primitively, the arolium is
well-developed, forming a definite padlike
structure between the tarsal claws and be-
ing distinct under magnifications of 50 X
or less. F16.1 — The arolium is obliterated

(or at least extremely reduced) so that
no such structure is distinguishable, even
at magnifications of lOOX.

A well-developed arolium is considered
primitive because this is the condition in
most Aculeata, including the Fedtschenki-
inae and Anthoboscinae, those groups
most closely related to the Mutillidae.

Complete loss of arolia has occurred
on a single occasion, this condition being
characteristic of the Myrmosinae. Within
the other Mutillidae there is some varia-
tion in the degree of development of the
arolia. These are most reduced (but never-
theless still distinct) in, for example, Pseit-
dophotopsis (Pseudophotopsidinae) and
Reedotrnttilla (Sphaeropthalmina).

F17. Fortn of first metasomal segrnent
(Fig. 76). Primitively, the first metasomal
segment (especially the tergum) is grad-
ually and evenly broadened posteriorly
and merges more or less smoothly in con-
tour with the second. Although it may
be somewhat constricted apically and nar-
rower than the second, the first segment
is never predominantly cylindrical (paral-
lel-sided). F17.1— The first metasomal
segment (notably the tergum) is much
enlarged and almost parallel-sided, approx-
imately as broad as the second segment
and more than half its length. F17.2 — The
first metasomal segment, and especially the
tergum, forms a constricted cylinder
which is much narrower than and less
than one quarter the length of the second
segment.

A flaring first metasomal segment is
considered primitive because this is the
most widespread form amongst the Acu-
leata and in particular those taxa most
closely related to the Mutillidae,

Both derived states are uniquely char-
acteristic of single taxa on the cladogram.
F17.1 is present only in the Rhopalomu-
tillinae and F17.2 characterizes the tribe
Ephutini. Since these are autapomorphic
states, they do not provide useful informa-

Phylogeny and Classification of the Aculeate Hymenoptera

601

tion on the grouping of the higher taxa
but merely emphasize the speciahzations
of the particuhir taxa involved,

F1(S. Base of first metasomal tergum (Fig.
77). Primitively, the first metasomal ter-
gum is simple basally without any lateral
protuberances. F18.1 — The first meta-
somal tergum bears a pair of protuber-
ances, one on each side, at the base, form-
ing "auricles" which tend to cup the apex
of the propodeum.

A simple first metasomal tergum is
considered primitive because this is the
condition in most Aculeata, and in par-

. F19.1 ,
M38.1

Figs. 77-79. Characters of Mutillidae. 77, articula-
tion of meso- and metasoma of female, dorsal view,
showing primitive and derived states of base of first
metasomal tergum (F18 based on Myrmosa; F18.1
on Psciidophotopsis); 78, seta from first metasomal
tergum, showing derived state of pubescence (F19.1,
M38.1 based on Cystomutilla, Dasymiitilla, Reedo-
mtitilla, Bothriomutilla, left to right); 79, anterior
region of metasoma, lateral view, showing primitive
and derived states of tergal felt line (F20, M39
based on Areotilla, $ ; F20.I, M39.1 on Odonto-
mittilla, $, and Smiciomyime, $, top to bottom).

ticular in those taxa which are most closely
related to the Mutillidae (viz., Fedt-
schenkiinae, Anthoboscinae) .

The development of auricles has ap-
parently occurred only once. These struc-
tures are present in all members of the
Mutillidae except for the Myrmosinae, so
that state FlS.l was apparently derived on
internode 1-2. The degree of development
of the auricles varies, but they are smallest
in the groups below internode 4-5, except
for Psciidophotopsis where they are quite
large. (This may in part be a result of
the generally large body size of these spe-
cies; the ancestral form may have been
relatively small, if "Cope's Rule" is ap-
plicable—Stanley, 1973.)

F19. Pubescence of first metasomal ter-
gum (Fig. 78). Primitively, all the pubes-
cence is composed of simple, smooth setae.
F19.1 — Some erect setae at the base of the
first metasomal tergum are plumose or
subplumose, bearing fine branches. Much
of the pubescence elsewhere on the body
may also be plumose.

Simple pubescence is considered prim-
itive because this is the condition in most
Aculeata, including those groups most
closely related to the Mutillidae.

Development of plumose pubescence
has apparently occurred on a single occa-
sion, being characteristic of the entire tribe
Sphaeropthalmini. A few scattered genera
within this group appear to have lost the
plumosity, however (e.g., Euspinolia, Atil-
Uim, Hoplocrates, Neomiitilla, Cephalo-
mutilla, Traiimatomnti/la). Despite these
few apparent reversals, the presence of
state F19.1 is considered strong evidence
for the holophyletic origin of the tribe
Sphaeropthalmini. Although some species
of Stenomiitilla (Dasylabrini) have setae
with trifurcate apices, their condition is
unlike that in the Sphaeropthalmini and
thus does not diminish the significance of
this state.

F20. Tergal felt line (Fig. 79). Primi-

602

The University of Kansas Science Bulletin

lively, the second metasomal tergum is
simple, without any development of spe-
cialized lateral felt lines or other secretory
structures. F20.1 — The second metasomal
tergum bears a specialized felt line with
recumbent setae and secretory pores, lat-
erally on each side.

The absence of felt lines is considered
primitive because this is the condition in
most Aculeata, and in particular in the
Sapygidae and Tiphiidae, those groups
most closely related to the Mutillidae. Felt
lines of similar form occur elsewhere only
in the Bradynobaenidae, although some
primitive bees (various CoUetidae) possess
foveae on the second metasomal tertrum
in a similar position, and these may be
analogous to felt lines. The detailed anat-
omy of the felt lines in Mutillidae and
Chyphotinae (Bradynobaenidae) has re-
cently been elucidated by Debolt (1973).

A tendency toward development of felt
lines on the second metasomal tergum ap-
pears to be characteristic of the family
Mutillidae except for the Myrmosinae, al-
though the actual development of such
lines has apparently occurred on two occa-
sions, once in Pseitdophoiopsis and again
on internode 4-5. The presence of the felt
line in a relatively basal taxon (Pseudo-
photopsidinae) as well as most higher
members, indicates that its development
is probably a general tendency in the fam-
ily (at least above Myrmosinae). Since a
tergal felt line has secondarily been lost
in some higher genera (e.g., Stenomiitilla,
many Ephiita), it may be that the absence
of the line is secondary in Ticoplinae and
Rhopalomutillinae also. In both these
groups there are, however, no traces of
even rudimentary development of a felt
line. It is thus more parsimonious to con-
sider state F20.1 to have been fully ex-
pressed on two occasions rather than for it
to have been developed on internode 1-2
and then to have been lost on two subse-
quent occasions (in the Ticoplinae and

Rhopalomutillinae). Nevertheless, the "ten-
dency toward" development of a tergal
felt line may be visualized as having been
uniquely derived on internode 1-2 (as in
the above investigation of Aculeata), this
being a fairly strong state uniting most
Mutillidae. (The development of felt lines
seems to be correlated with the derived
states of characters F53r, F54r, M79r and
M80r, which apparently do not occur in
the Myrmosinae, so that the tendency to-
ward development of felt lines is con-
sidered to have been established after the
divergence of the Myrmosinae.)

M21. Eye form. Primitively, the com-
pound eye in the male is somewhat flat-
tened in profile, merging smoothly with
the surrounding cuticle. M21.1 — The com-
pound eye is highly convex in profile,
strongly differentiated from the surround-
ing cuticle and approximately hemispher-
ical.

A flattened eye is considered primitive
because this is the condition in most Acu-
leata and in particular those taxa most
closely related to the Mutillidae (viz.,
Fedtschcnkiinae, Anthoboscinae).

The specialized hemispherical form of
the eye in the male parallels the develop-
ment of a similar eye form in the female
(F2.1) quite closely. Thus, all members
of the Sphaeropthalmini have hemispher-
ical eyes in the male also, except that
CystomatiUu has this state somewhat less
well-developed than in most of the other
genera. In addition, the group including
EuspinoUa, TaUitun, Atillum and Hoplo-
crates also has the eyes less protuberant
than other Sphaeropthalmini, as in the
females. In contrast to the females, the
males of Tricholabiodes (Dasylabrini)
also have well-developed and protuberant
eyes, although they are less nearly circular
than in the Sphaeropthalmini. This means
that if state M21.1 evolved once only,
then the Sphaeropthalmini and Tricho-
labiodes share a common ancestor not also

PHYLOGENY AND ClASSIIICATION OF THE AcULEATE HyMENOPTERA

603

shared with the remaining Dasylabrini.
Alternatively, state M21.1 may have arisen
twice, once in the ancestral Sphaeropthal-
mini and again in Tricholabiodes. Al-
though there are no good characters indi-
cating holophyly of the Dasylabrini, it
seems that forms similar to Tricholabiodes
(all the species of which are nocturnal
and highly specialized) would probably
have been too specialized to have given
rise to the Sphaeropthalmini. Two sepa-
rate derivations of state M21.1 thus seem
the more probable. Independent develop-
ment of this condition in Tricholabiodes
was probably associated with their noc-
turnal habit since other nocturnal Aculeata
often have the eyes larger and more pro-
tuberant than do their relatives (e.g.,
Brachycistidinae in the Tiphiidae, Mega-
lopta in the halictid bees, and even Pseu-
dophotopsis and Eremornyrme elsewhere
in the Mutillidae to a lesser extent). De-
spite the uncertainty regarding the exact
pattern of evolution of this character, state
M21.1 provides quite good evidence for
the holoph}'letic origin of the Sphaerop-
thalmini.

M22. Eye shape (Fig. (SO). Primitively,
the compound eye is broadly oval with the
inner margin sinuately concave. M22.1 —
The compound eye is subcircular with the
inner margin approximately evenly con-
vex. M22.2 — The compound eye is broadly
oval with the inner margin strongly in-
cised and emarginate, the angle of the
emargination at its apex being less than
90".

A broadly oval eye with sinuate inner
margin is considered primitive because
this is the condition in most Aculeata and
in particular in those groups (such as
Fedtschenkiinae and most Tiphiidae)
which are most closely related to the
Mutillidae.

An approximately circular compound
eye with a convex inner margin in the
male is characteristic only of the Sphaerop-

81

M22.1

Figs. 80-81. Characters o£ Mutillidae. 80, head of
male, anterior view, showing primitive and derived
states of eye shape (M22 based on SmicTomyrmilla;
M22.1 on Ascetotilla; M22.2 on Rhopalomutilla);
81, meso- and metapleura and propodeum of male,
lateral view, showing primitive and derived states
of meso-metapleural suture (M26 based on Psetido-
photopsis; M26.1 on Artotilla: M26.2 on Si/iuiDiii/o-
tilla).

thalmini, so that state M22.1 apparently
arose only once on the cladogram. How-
ever, a few genera (such as Eiispinolia
and Talliiim) have eyes which are almost
circular but with the inner margins very
slightly sinuate. A few members of the
Dasylabrini (such as Apteromiitilla,
Brachymutilla and Dasylabroides) also
have the inner margins of the eyes convex
although the eyes are not as nearly circular
as in the Sphaeropthalmini. Actually,
there seems to be a tendency toward re-
duction in the size of the eyes in many of
the higher Mutillidae and especially in
the Sphaeropthalminae, which sometimes
makes accurate determination of the oc-
currence of state M22.1 difficult. Never-
theless, this does seem to be a fairly good
indicator of the holophyletic nature of the
Sphaeropthalmini.

604

The University of Kansas Science Bulletin

Emargination of the eye (M22.2) is
characteristic of the entire subfamily Mu-
tilUnae and thus arose on internode 7-8.
A similar condition is, however, present
in Rhopalomutillinae. Areotilla shows a
marked tendency toward development of
this state also, with Smicrotnvrtnilhi and
Nanomiitilla (all three Ticoplinae) having
the inner margin slightly more deeply
sinuate than in Pseudophotopsis. Thus
the presence of state M22.2 in the Mutilli-
nae is not as useful an indication of the
association of the members of this sub-
family as could be wished, although it is
obvious that its occurrences in the Rho-
palomutillinae and Areotilla were inde-
pendent of its origin in the Mutillinae.

M23. Eye pubescence. This character
shows the same pattern of evolution and
occurrence as in the female (F3), with the
derived state (M23.1) having occurred in
all members of the Mutillidae except the
Myrmosinae and Ticoplinae. It thus pro-
vides no additional information on higher
groupings.

M24. Pronotal pit (Fig. 69). This char-
acter shows the same pattern of occurrence
as in the female (F8), with the derived
state (M24.1) being characteristic of the
Pseudophotopsidinae. This condition was
one of the main characters used by Schu-
ster (1950) to associate the sexes of this
group.

M25. Mesosternal teeth (Fig. 72). The
states of this character are as for the fe-
male (Fll), although their distribution is
slightly different. The elaboration of the
mesosternal carinae into dentate projec-
tions in the male (M25.1) coincides with
a similar development in the female
(Fll.l), being found only in Pseudopho-
topsis. By contrast, the obliteration of the
carinae (M25.2) has occurred in all other
Mutillidae except for most Myrmosinae,
unlike the situation in females. Thus,
simple carinae are developed in the males

of Myrmosa but they have been lost in
Mvi'nwsula (the condition in Protomutilla
is unknown). State M25.2 thus appears
only once on the cladogram, on internode
2-3. Nevertheless, its parallel occurrence
within the Myrmosinae tends to weaken
it, so that it should not be considered a
particularly good indicator of the holophy-
letic nature of the Mutillidae above inter-
node 2-3. It appears that there is never any
redevelopment of mesosternal carinae or
teeth in the males of the more hii^hlv de-
rived Mutillidae similar to that in the
females of Ascetotilla and other genera re-
lated to Ephutomorpha. A few genera in
the Sphaeropthahna complex (such as
Odontophotopsis) bear mesosternal teeth
of varying types, but these are placed an-
terior to the position of the primitive
carinae and are obviously not an expres-
sion of this character.

M26. Meso-metapleural suture (Fig. 81).
Primitively, the meso-metapleural suture
is approximately straight. M26.1 — The
meso-metapleural suture is curved pos-
teriorly so that the hind margin of the
mesopleuron is convex. M26.2 — The meso-
metapleural suture is sinuate so that the
hind margin of the mesopleuron is con-
cave over at least the ventral half and is
convex dorsally; the metapleuron thus ap-
pears to be expanded anteriorly below the
endophragmal pit.

An approximately straight meso-meta-
pleural suture is considered primitive be-
cause this is the condition in those Acu-
leata most closely related to the Mutillidae
(viz., Fedtschenkiinae, Anthoboscinae)
and also in those groups considered on
the basis of other characters to be the
most basal on the cladogram (i.e., Myr-
mosinae, Pseudophotopsidinae).

A posteriorly curved meso-metapleural
suture (M26.1) is found only in Areotilla,
Nanomutilla and Smicrotnyrmilla so that
this forms a very good characteristic estab-
lishing the Ticoplinae as a holophyletic

Phylogeny and Classii'ication of the Aculeate Hymenoptera

605

group (the condition in Ticopla is, how-
ever, unknown to nie). A sinuate meso-
nietapleural suture (M26.2) is character-
istic of the entire MutiHidae except for the
Myrmosinae, Pseudophotopsidinae and
TicopHnae. This state apparently arose
uniquely on internode 3-4 and is thus a
very useful indicator of the holophyletic
origin of the four higher suhfamilies.

M27. Meso-metapleural "bridge" (Fig.
82). Primitively, the meso- and meta-
pleura are closely associated but not fused
at any point. The anterior margin of the
metapleuron is simple ventrally with only
a slight protuberance approaching the
hind margin of the mesopleuron but not
contacting it. M27.1 — The metapleuron
ventrallv bears a marked tubercle which

M27.1

M3I

M31.1.1

M31.I.2

84

M33

M33.1.1

M33.1.2

M33.1.2.1

Figs. 82-84. Characters of MutiHidae. 82, ventral
region of meso-metapleural suture of male, lateral
view, anterior to left, showing primitive and derived
states of meso-metapleural "bridge" (M27 based on
Myrmosa; M27.1 on Sqiiamnlotilla); 83, left tegula
of male, dorsal view, showing primitive and derived
states (Mil based on Myrmosa; Mil. 1.1 on Odonto-
mtitilla; Mi 1.1. 2 on Ephiita): 84, pterostigma of
forewing, showing primitive and derived states (Mii
based on Pseudophotopsis; Mii. 1.1 on Viereckia:
Mii. 1.2 on Dolicliomntilla: Mii. 1.2.1 on Ephiita).

is fused with a similar protuberance on
the mesopleuron, so that a "bridge" is
formed between these pleura which are
additionally fused for some distance over
their ventral halves.

A simple metapleuron not fused to the
mesopleuron is considered primitive be-
cause this is the condition in most Acu-
leata and in particular in the Fedtschcnki-
inae and Tiphiidae, those taxa most closely
related to the MutiHidae. The develop-
ment of a slight protuberance at the
anteroventral angle of the metapleuron
has not occurred in these related taxa but
is actually another derived state associating
the myrmosid-mutillid complex, and is
thus considered primitive for the complex.

The development of a meso-metapleu-
ral "bridge" is characteristic of all Mutil-
lidae except for the Myrmosinae, Pseudo-
photopsidinae and TicopHnae, in which
there is merely a small tubercle ventrally
on the metapleuron. Thus state M27.1 ap-
parently arose on internode 3-4, strongly
associating most of the MutiHidae into a
holophyletic group.

M28. Metasternal process (Fig. 73). The
evolution and occurrences of the various
states of this character are the same as in
the female (F13). Thus state M28.1 is
characteristic of the entire MutiHidae ex-
cept for the Myrmosinae, apparently hav-
ing arisen on internode 1-2.

M29. Hind coxal tubercle (Fig. 74). The
pattern of modifications of this character
is the same as for the female (F14), so
that state M29.1 is characteristic of the
Myrmosinae, and M29.2 has apparently
arisen on internode 4-5. As in the female,
a tubercle is never very well-developed in
the male, except in Myrmosinae, so that
determination of the state involved is often
difficult, reducing the significance of this
character.

M30. Tarsal claws (Fig. 75). Primitively,
each tarsal claw bears a sharp tooth about

606

The University of Kansas Science Bulletin

halfway along the ventral margin. M30.1
— Each tarsal claw is simple, the ventral
tooth having been lost. M30.1.1— Each
tarsal claw is broadened into a dentate
lamella basally but is simple apically, so
that it appears cleft.

A tarsal claw with a single distal ven-
tral tooth is considered primitive because
most aculeates have toothed claws and
this is the condition in the Fedtschenkiinae
and many Tiphiidae, the groups most
closely related to the Mutillidae. The sub-
sequent development of a basal lamella is
considered to be derived from the simple
condition because this state is unlike that
in most other Aculeata and also because
it appears in only one sex of a single genus
within the Mutillidae, a genus which
seems on the basis of other characters to
have arisen on the cladogram above the
point of derivation of state M30.1.

Loss of the tooth on the tarsal claws of
the male does not quite coincide with this
condition in the female (F15.1). Thus,
the males of Myrmosinae have armed
claws whereas the females do not. State
M30.1 has thus apparently arisen only
once, on internode 2-3, being character-
istic of all Mutillidae except for Myrmosi-
nae, Pseudophotopsidinae and Rhopalo-
mutillinae. The last group is uniquely
characterized by possession of state
M30.1.1. Thus state M30.1 is useful as
an indicator of the holophyletic nature of
the Mutillidae above Pseudophotopsidinae
while M30.1.1 merely serves to emphasize
the distinctness of the Rhopalomutillinae
and does not aid in establishing higher
groupings. This interpretation of the
origin of the condition in the male of
Rhopalomittilla differs from that of Schu-
ster (1947) who considered it to be prim-
itive and comparable to that in Pseudo-
photo p sis.

M31. Tegiila (Fig. 83). Primitively, each
tegula is an evenly convex, subcircular
sclerite with its hind margin not attain-

inii the level of the transscutal suture.
M31.1 — Each tegula is posteriorly pro-
duced so that its hind margin exceeds
the level of the transscutal suture, the
tegula being subovate. M31.1.1— The hind
margin of the elongate tegula is reflexed,
forming a posterior upcurved rim. M31.1.2
— The elongate tegula is longitudinally
angulate basally so that it has two distinct
surfaces approximately perpendicular to
one another, at least anteriorly.

A short, evenly convex tegula is con-
sidered primitive because this is the con-
dition in most Aculeata, including those
groups most closely related to the Mutilli-
dae (viz., Fedtschenkiinae, Anthobosci-
nae).

Elongation of the tegula (M31.1) is
characteristic of the entire subfamily Mu-
tillinae and has apparently arisen on the
cladogram only once, on internode 7-8.
Elongate tegulae are found in scattered
genera elsewhere, however, but in these
their form is generally not identical to
that in the Mutillinae. Thus Smicromyr-
milla has tegulae which reach the trans-
scutal suture but do not exceed it, and
Nanonuitilla and Areotilla have tegulae
which do exceed this line but which are
mesallv curved posteriorly, those in Areo-
tilla in particular being very elongate and
more or less reniform, a condition unlike
that in the Mutillinae. (These modifica-
tions do, however, indicate that there is
a tendency toward some type of tegular
elongation in the Ticoplinae as a whole,
but this does not appear on the cladogram.
It does not fall into any of the designated
states since it is rather indefinite.) A few
of the more highly derived members of
the Sphaeropthalmina, like Bothriomu-
tilhi, also have the tegulae posteriorly pro-
duced to or slightly beyond the level of
the transscutal suture, but in these this
appears to be partly the result of anterior
displacement of this suture laterally, and
the tegulae are not as elongate as in the

Phylogeny and Classification of the Aculeate Hymenoptera

607

Mutillinae. The presence of state M31.1
is thus considered quite good evidence for
the holophyletic origin of the MutiUinae.

State M31.1.1 is present only in the
trihe MutiUini, although a few of the more
derived members of the Sphaeropthalmina
such as Bothrio7ni!tilla exhibit a similar
tendency toward reflexion of the posterior
margin of the tegula. Also, a few genera
of MutiUini such as Ctenotilla and Mime-
comutilla show a secondary reversal of
this state with the hind margin of the
tegula not upturned but forming a more
or less extensive flat area. Despite these
few anomalies, state M31.1.1 appears to be
a strong indicator of the holophyletic
nature of the tribe MutiUini.

State M31.1.2 is uniquely characteristic
of the tribe Ephutini and is not ap-
proached elsewhere in the Mutillidae, thus
providing good evidence of the holophy-
letic nature of this tribe but not proving
useful in the establishment of any higher
groupings.

M32. Extent of venation of forewing.
Primitively, the venation of the forewing
attains the distal margin of the wing
membrane or ends only a very short dis-
tance from it. M32.1 — The venation of
the forewing is obliterated apically so that
the longitudinal veins do not reach the
distal margin of the membrane but end
a considerable distance from it; although
there may be dark lines which do almost
attain the margin, these are merely pig-
mentary and not differentiated cuticular
structures.

Venation attaining the apical margin
(jf the forewing is considered primitive
because this is the condition in those
groups most closely related to the Mutil-
lidae [viz., Sapygidae (although venation
slightly retracted in Fedtschen\ia) and
Anthoboscinae].

Retraction of the venation of the fore-
wing to end a considerable distance from
the apical margin of the wing is charac-

teristic of the entire family Mutillidae ex-
cept for the Myrmosinae, and thus appar-
ently arose on internode 1-2. Although
this is a imique derivation here, a similar
condition has occurred in various other
groups of Aculeata, including some Tiphi-
idae, thus weakening this state somewhat
as an indicator of the holophyletic nature
of most of the Mutillidae. Instead its
absence serves to emphasize the basal posi-
tion of the Myrmosinae.

M33. Pterostigma (Fig. 84). Primitively,
the pterostigma is entirely heavily sclero-
tized and thus not cell-like. M33.1 — The
pterostigma has the sclerotization reduced,
somewhat more so anteriorly than pos-
teriorly, so that it appears as a small cell
bounded basally by vein SC and with
vein R much heavier than vein C. M33.1.1
— The pterostigma is apparently formed
entirely by the relatively heavy vein R,
the free section of vein SC being lost.
M33.1.2 — The pterostigma is entirely de-
sclerotized, with all the bounding veins
of approximately equal width. M33. 1.2.1
— The pterostigma is absent as a result of
the loss of the free section of vein R (or
its fusion with vein SC) eliminating any
cell.

A completely sclerotized pterostigma is
considered primitive because this is the
condition in most Aculeata, including
those taxa most closely related to the
Mutillidae (e.g., Fedtschenkiinae, Antho-
boscinae).

A pterostigma with the sclerotization
reduced anteriorly (M33.1) is not found
in any extant members. Since all the taxa
subtended by internode 5-7 show some
degree of pterostigmal desclerotization, all
with conditions which are logically deriv-
able from a state such as M33.1, this state
may have arisen uniquely on this inter-
node. State M33.1 is thus apparently a
good indicator of the holophyletic nature
of the group comprised of the Myrmillinae
and Mutillinae, despite the fact that this

608

The University of Kansas Science Bulletin

state is hypothetical. There is actually a
tendency toward a similar development
in Protoplwtopsis (Sphaeropthalmina) al-
though in this genus the reduction o£
sclerotization is more even, veins R and
C being approximately equally heavy, both
heavier than any of the other veins.

Loss of vein SC is characteristic of the
entire subfamily Myrmillinae, although
this state (M33.1.1) is not very obvious in
some members where the reduction of
sclerotization is not very marked (e.g.,
Vierecl{ia) and the loss of vein SC merely
results in a basal emargination of the
pterostigma. Since this state is uniquely
characteristic of all members of the sub-
family, it is useful in establishing the holo-
phyletic nature of the Myrmillinae but
does not aid further in higher groupings.

Complete desclerotization of the ptero-
stigma (M33.1.2) is found only in the
Mutillinae and is thus a strong indicator
establishing the holophyly of the group,
having arisen on internode 7-8. Some
slight uncertainty is introduced by the fact
that the Ephutini is characterized by com-
plete loss of the pterostigma, although this
condition seems logically more readily
derivable from a state such as M33.1.2
rather than from the primitive condition.
Loss of the pterostigma in the Ephutini
(M33.1.2.1) has apparently occurred as a
result of the elimination of its apical re-
gion, whether by loss of vein R or by its
fusion with vein SC cannot be established.
In any case it is apparent that vein SC is
retained since the "crossvein" joins vein
SC+R just distal to the weak point in
vein SC+R. This point is at the base of
the pterostigma, just proximal to vein SC,
in other mutillids. This loss of the ptero-
stigma provides a strong indication of the
holophyletic nature of the Ephutini. A
superficially similar loss of the pterostigma
has occurred in Odontomiitilla (Mutillina)
but in this case it is apparently the result
of the obliteration of the free section of

vein SC since the "crossvein" joins vein
"SC+R" at a point some distance distal
to the weak point in vein "SC+R."

M34. Foieiving cell IS (Fig. 85). Primi-
tively, cell IS (second submarginal) is
sessile anteriorly, sharing a section of vein
S with cell R (marginal). M34.1— Cell IS
is petiolate anteriorly, not reaching cell
R (crossvein r-s reaches vein S distal to
the junction of vein S and crossvein Is-m).
A sessile cell IS is regarded as primi-
tive because this is the condition in most
Aculeata, including those taxa most closely
related to the Mutillidae (e.g., Fedt-
schenkiinae, Anthoboscinae).

A petiolate cell IS is found only in the
Ticoplinae where it is characteristic of all
the genera (except perhaps for Ticopla
where cell IS is apparently lost). This
state thus provides very good evidence of
the holophyletic nature of this subfamily
but does not aid further in higher level
groupings. Elsewhere, state M34.1 is ap-
proached within the Myrmosinae (Myr-
mosa has cell IS anteriorly acute and

Figs. 85-86. Characters of Mutillidae. 85, forewing
of male, showing derived state of cell IS (M34.1
based on Areotilla)\ 86, hind wing of male, showing
primitive and derived states of jugal lobe (M35
based on Myrtnosa; M35.1 on Areotilla).

Phylogeny and Classification of the Aculeate Hymenoptera

609

barely reaching cell R although not ac-
tually petiolate).

M35. Jugcil lobe (Fig. 86). Primitively,
the hind wing bears a small but well-dif-
ferentiated jugal lobe basally. M35.1 — The
jugal lobe is completely lost so that the
hind wing has a smooth posterior margin.

Presence of a jugal lobe is considered
primitive because this is the condition in
most Aculeata, including the Fedtschenki-
inae and Anthoboscinae, i.e., those taxa
most closely related to the Mutillidae.

The complete absence of a jugal lobe
in the hind wing is characteristic of the
entire Mutillidae except for the Myrmo-
sinae and Pseudophotopsidinae. This state
has thus apparently arisen on internode
2-3 and provides strong evidence for the
holophyletic nature of the higher Mutil-
lidae, despite the fact that a similar condi-
tion has occurred elsewhere in the Acu-
leata (e.g., in Bethyloidea).

M36. Propodeal disc (Fig. 87). Primi-
tively, the propodeal disc is fairly heavily
and evenly sculptured, often reticulately
so. M36.1 — The disc of the propodeum
bears four subparallel longitudinal carinae
linked posteriorly by a zig-zag transverse
carina.

An evenly sculptured propodeal disc
is considered primitive because this is the
condition in those Aculeata most closely
related to the Mutillidae (viz., Fedtschenk-
iinae, Anthoboscinae) and in most mem-
bers of the Mutillidae.

A characteristic pattern of carinae on
the disc of the propodeum is most highly
developed in Nanomutilla (and Ticopla)
and slightly less so in Areotilla. In Smic-
romyrmilla the fully winged species have
similar carinae on the anterior region of
the disc but the apterous species have the
sculpturing modified so that the carinae
are lost. Although a few species of Rho-
palomiitilla show a superficially similar
condition (differing in detail, however),
this pattern of sculpturing is considered

87

88

Figs. 87-88. Characters of Mutillidae. 87, posterior
region of inesosoma of male, dorsal view, showing
derived state of propodeal disc (M36. 1 based on
KiinomntiUa); 88, posterior region of male genitalia,
lateral view, showing primitive and derived states
of gonostylus (M40 based on Psciidophotopsis; M40.1
on Nanottitttilla: M-IU.1.1 on Piotophotopsis: M40.1.2
on Antcnnotillii) .

to be present in these genera only, and
thus is useful in associating them into the
holophyletic subfamily Ticoplinae. (Nagy,
1970, however, cited this as a character in
which Ticopla was similar to some Bethy-
lidae, although the pattern differs in de-
tail in the groups involved.)

M37. Form of first metasomal segment
(Fig. 76). Primitively, the first metasomal
segment (especially the tergum) is evenly
expanded from base to apex and is only
slightly constricted apically so that it is
not highly differentiated from the rest of
the metasoma. Although it may be some-
what constricted apically and narrower
than the second, the first segment is never
predominantly cylindrical (parallel-sided).
M37.1 — The first metasomal segment (no-
tably the tergum) is approximately cylin-
drical and short, being highly differenti-
ated from the remainder of the metasoma.
An evenly dilated first metasomal seg-
ment is considered primitive because this
is the condition in most Aculeata, includ-
ing those taxa most closely related to the

610

The University of Kansas Science Bulletin

Mutillidae (e.g., Fedtschenkiinae, Antho-
boscinae), and also in most mutillids.

The development of a cylindrical,
parallel-sided first metasomal segment in
the male follows the same pattern as a
similar development in the female (F17.2),
being characteristic only of the Ephutini
and providing a strong indication of the
holophyly of that tribe but not providing
additional information above that supplied
by the female.

M3(S. Pubescence on first metasomal tcr-
gum (Fig. 78). The development of
plumose pubescence in the males (M38.1)
exactly parallels that in the females
(F19.1), being characteristic of the entire
Sphaeropthalmini except for a few genera
in which plumosity has apparently been
lost secondarily. This strengthens the im-
portance of this state in establishing the
Sphaeropthalmini as a holophyletic group,
but does not provide more information on
groupings than that given by females.

M39. Tergal felt line (Fig. 79). The pat-
tern of development of a tergal felt line
in the male is exactly the same as that in
the female (F20), the derived state being
present in the Pseudophotopsidinae and
above internode 4-5 except for a few spe-
cies. The tendency toward expression of
state M39.1 may thus be considered to
have been established on internode 1-2,
although actual expression has occurred
on at least two occasions. This state does
not provide any additional information
on higher groupings beyond that provided
by females, but does serve to strengthen
the conclusions based on that in the female.

M40. Gonostyhts (Fig. 88). Primitively,
each gonostylus is about twice as long as
high, being lamellate with a rounded apex.
M40.1 — Each gonostylus is at least three
times as long as its basal height, being
approximately straight and tapered so that
the apex is acute. M40.1.1 — Each acute
gonostylus is upcurved at the apex.

M40.1.2 — Each acute gonostylus is down-
curved apically.

A broad, rounded gonostylus is con-
sidered primitive because this is the condi-
tion in various of the more generalized
groups of the Aculeata (e.g., Plumariidae,
Sierolomorphidae) and especially in those
taxa most closely related to the Mutillidae,
such as Fedtschenkiinae and Anthobos-
cinae.

The development of an acute gono-
stylus (M40.1) has apparently occurred on
a single occasion on the cladogram, on
internode 2-3, since this condition is char-
acteristic of all Mutillidae except for Myr-
mosinae and Pseudophotopsidinae. There
are, however, a few instances in which
somewhat of a reversal has occurred, so
that in Brachymntilla (Dasylabrini), Atil-
lum and Hoplocrates (both Pseudometho-
cina) the gonostyli are fairly broad and
their apices are somewhat rounded, a simi-
lar but less marked tendency being shown
in Hoplomittilla, Myrmilloides (both Pseu-
domethocina) and Ctenotilla (Smicromyr-
mina). Although state M40.1 is best
placed on internode 2-3, Nanomiitilla
(Ticoplinae) has gonostyli which are nar-
rowed but with apices which are some-
what blunt. In sum, the gonostyli in
Nanomutilla appear to be more similar
to those of the higher Mutillidae than to
Pseudophotopsis, however. Thus the pos-
session of narrow, acute gonostyli is quite
a good characteristic associating the Mu-
tillidae above internode 2-3 into a holo-
phyletic group. In the Myrmosinae, how-
ever, a somewhat similar development has
taken place in some members of the
genus Myrmosa, although Myrmosula has
gonostyli of the primitive type.

Dorsal curvature of the gonostylar apex
(M40.1.1) is characteristic of the entire
subfamily Sphaeropthalminae and is pres-
ent in all members of that taxon except
for a few specialized genera in which the
gonostylus has been further modified and

Phylogeny and Classification of the Aculeate Hvmenoptera

611

the apex is more or less straight (e.g., Hop-
loiniitilla, Myrmilloides, Atillitm, Hoplo-
crates, Lomachaeta). Although the apical
curvature of the gonostylus may often be
rather subtle, this seems to be a useful
state for associating the Sphaeropthalmi-
iiac in a holophyletic group, having arisen
on internode 5-6.

Ventral curvature of the apex of the
gonostylus (M40.1.2) is characteristic of
the entire branch bearing both the Myr-
millinae and Mutillinae although the de-
gree of curvature is somewhat reduced in
Ctetwtilla (Smicromyrmina). This is thus
a useful state establishing that these two
subfamilies form a holophyletic group.
Although Nanomittilla in the Ticoplinae
has the gonostylus approximately straight,
both Areotilla and Smicromyrmilla have
the gonostyli strongly ventrally curved but
with the conformation of the gonostyli in
other respects unlike that of the myrmil-
line-mutilline group. This occurrence thus
almost certainly represents an independent
but somewhat similar modification of the
gonostylus and does not materially weaken
the significance of state M40.1.2 on inter-
node 5-7.

M41. Gonapophysis IX (Fig. 89). Prim-
itively, each gonapophysis IX has the apex
much produced dorsally and posteriorly
into a marked lobe and bears a tooth about
halfway along the ventral margin. M41.1
—Each gonapophysis IX has the dorsal
apical lobe much reduced or lost and bears
a tooth on the apical half of the ventral
margin.

A gonapophysis IX with a large dorsal
lobe apically is considered primitive be-
cause this is the condition in the Fedt-
schenkiinae, the group most closely related
to the Mutillidae, and also in the Myr-
mosinae which forms a basal branch of
the cladogram on the basis of other char-
acters.

The form of the gonapophysis IX is
somewhat more complex than can be con-

FiGS. 89-91. Characters of Mutillidae. 89, right
gonapophysis IX of male, lateral view, anterior to
right, showing primitive and derived states (M41
based on Myrmosa nigriceps: M41.1 on Areotilla
sp.); 90, right gonapophysis IX of male, lateral view,
anterior to right, showing derived state of gonapo-
physeal spines (M42.1 based on Psetidophotopsis
■conlimia); 91, right volsella, mesal view, anterior
to left, showing primitive and derived states of
digitus (M43 based on Dasymiitilla nignpes; M43.I
on Areotilla sp.).

cisely described, but, basically, each has lost
the dorsal lobe or has it greatly reduced
in all taxa except the Myrmosinae, so
that state M41.1 has apparently evolved
once, on internode 1-2. Actually the form
of the dorsal lobe in the Fedtschenkiinae
is slightly different from that in the Myr-
mosinae where it often seems to be more
strongly developed distally. Within the
Myrmosinae its form differs also, being
a large lamella often bearing long mar-
ginal setae in Myrmosa; it is non-setose in

612

The University of Kansas Science Bulletin

Mynnositla but with a large apical ventral
tooth, so that its condition appears to be
more similar to that in both the Fedt-
schenkiinae and other Mutillidae.

M42. Gonapophyseal spines (Fig. 90).
Primitively, each gonapophysis IX does
not bear any articulating spines but is
merely armed with one or more simple
cuticular teeth ventrally. M42.1 — Each
gonapophysis IX bears stout spines articu-
lating at the base along the distal ventral
margin in addition to simple cuticular
teeth.

Simply toothed gonapophyses IX are
considered primitive because this is the
condition in the group most closely related
to the Mutillidae (i.e., Fedtschenkiinae)
and also in almost all mutillids.

A gonapophysis IX with stout, articu-
lating spines is found only in Pseiidopho-
topsis. This is thus a good state establish-
ing the Pseudophotopsidinae as a highly
differentiated taxon, but it does not pro-
vide any information on higher groupings.
Although such stout spines are unique to
Pseiidophotopsis, they may actually be
modified setae such as are present in simi-
lar positions on the gonapophysis IX in
most Mutillidae.

M43. Digitus (Fig. 91). Primitively, the
volsella of the male genitalia comprises
both a cuspis and a digitus. M43.1 — The
volsella consists of only the cuspis, the
digitus having been obliterated.

A volsella with both digitus and cuspis
is considered primitive because this is the
condition in most Hymenoptera, includ-
ing those taxa most closely related to the
Mutillidae. "

A digitus is present in practically all
Mutillidae although it is often somewhat
reduced and apparently not movable rela-
tive to the cuspis. In both the Ticoplinae
and Rhopalomutillinae, however, the digi-
tus has been completely lost, apparently
independently in these two subfamilies.

the volsella otherwise appearing quite dif-
ferent in these taxa. Nevertheless, the
presence of state M43.1 is quite a useful
characteristic associating at least three
genera in the Ticoplinae (the condition in
Ticopla is unknown to me).

Characters Rejected for Derivation of
THE Final Cladogram

The following characters were con-
sidered but were rejected for the construc-
tion of the final cladogram, for various
reasons. Some showed a high incidence
of parallel or convergent derivations which
lessened their usefulness. Others could
not be described unequivocally in terms
of one or a few states which could be
readily distinguished. Instead they varied
gradually, making it impossible to include
them in the formal scheme. Other char-
acters were considered and used early in
the study before it had been established
that the Mutillidae as then constituted was
actually a polyphyletic group. When the
Typhoctinae, Apterogyninae and Chypho-
tinae (Bradynobaenidae) had been re-
moved from the Mutillidae, these char-
acters became invariant, even after the
Myrmosinae had been added to the study.
Further characters were found to vary
only within the final tribal or other divi-
sions which were used as the basic taxa
in the cladogram, and thus did not pro-
vide any information on higher groupings
at the tribal level or above.

Of course most of these characters do
provide useful information which was
utilized mainly early in the investigation
to obtain indications of probable group-
ings. Many will also be found useful in
establishing groups within the final taxa
used in the cladogram. However, the
limitations of each have to be borne in
mind at each stage.

Each character is briefly described be-
low and the reasons for its rejection are
given. The numbering is continuous with

Phylogeny and Classification of the Aculeate Hymenoptera

613

that for the characters utiHzed, but the
suffix "r" uidicates rejection.

F44r. Absence (primitive) or presence
(derived) of laterally flattened, lanceolate
setae in the female.

This character varies only within the
tribe Sphaeropthalmini in which those
genera in the Pseudomethocina possess
the derived state, as do other genera scat-
tered in the Sphaeropthalmina. This char-
acter is thus variable within the group,
provides no information on higher group-
ings at the levels involved in the clado-
gram and is also somewhat equivocal in
that the degree of modification of the
setae varies, making the determination of
setae as lanceolate or not difficult in many
cases,

F45r. Compound eye large (primitive) or
small (derived).

Although relatively large eyes are re-
tained in most members of the Mutillinae
as well as the Pseudophotopsidinae and
Myrmosinae, most members of the Sphae-
ropthalminae have the eyes reduced in
size, as do many Myrmillinae, the Rho-
palomutillinae and some Ticoplinae. There
have been many parallel and convergent
derivations of small eyes, and it is impos-
sible to define the two states unequivocally.

F46r. Compound eye with ommatidia
individually convex and differentiated
(primitive) or with entire surface smooth
and ommatidia undifferentiated (derived).

The derived state is found only within
the Sphaeropthalmini (where it is devel-
oped in many of the genera of the Sphae-
ropthalmina and a few of those in the
Pseudomethocina) and in Tricholabiodes
in the Dasylabrini. Although the two
states can be defined reasonably unequiv-
ocally, the derived state has occurred (or
has been reversed) on many parallel oc-
casions, thus not delimiting any groupings.

F47r. Antennal socket simple (primitive)

or with dorsomesal antennal tubercle
(derived).

After removal of the Typhoctinae,
Apterogyninae and Chyphotinae from the
Mutillidae, all members of the myrmosid-
mutillid complex possess the derived state.

F48r. Scape approximately straight and
relatively short (less than 1.5 times as long
as thick) (primitive) or sigmoid and more
than twice as long as thick (derived).

All members of the myrmosid-mutillid
complex have the derived state, the primi-
tive scape being found in some of the
genera that were excluded from the
Mutillidae.

F49r. First flagellar segment subequal in
length to the second (primitive) or more
than twice the length of the second
(derived).

Apart from a few members of the
Myrmillinae, the derived state is found
only in various members of the Sphaerop-
thalmini, where it is most commonly de-
veloped in the Pseudomethocina. This
character shows numerous parallel and
convergent derivations and also is incapa-
ble of unequivocal formulation.

F50r. Antenna with 12 (primitive) or 13
(derived) segments.

Only two genera possess the derived
state {Atillum and Hoplocmtes in the
Pseudomethocina) so that this is of no
use in establishing groups at the tribal
(or even the subtribal) level.

F51r. Absence (primitive) or presence
(derived) of a ventral excision forming a
basal tooth on the mandible. .

Although most members show the
primitive state, the derived state appears
in scattered genera throughout the Mutil-
lidae so that many parallelisms and con-
vergences are involved.

F52r. Mesosomal pleura approximately
flat or convex (primitive) or markedly
concave (derived).

614

The University of Kansas Science Bulletin

Although the derived state is partic-
ularly characteristic of the Myrmillinae,
it also occurs in one or more members of
the Myrmosinae, Ticoplinae and Mutilli-
nae and thus shows some parallel and con-
vergent developments. The states are also
difficult to define and identify unequiv-
ocally.

F53r. Lateral face of pronotum dorsally
without (primitive) or with (derived) a
small tubercle bearing fine setae and per-
haps secretory pores.

Although the derived state appears to
be characteristic of the entire Mutillidae
except for the Myrmosinae, an unequivo-
cal decision on the state involved, espe-
cially in small species, is often impossible
because of the small size and minimal
differentiation of this structure. The tu-
bercle may be analogous to the felt line
of the second metasomal segment.

F54r. Anterior face of pronotum dorsally
without (primitive) or with (derived) a
short transverse ridge bearing fine setae
and perhaps secretory pores.

As for the previous character, the de-
rived state is apparently characteristic of
the entire Mutillidae expect for the Myr-
mosinae, but certainty is impossible. This
structure may also be analogous to the felt
line of the second metasomal segment.

F55r. Pronotum with posteroventral mar-
gin distinct so that pronotum is well-dif-
ferentiated from mesopleuron (primitive)
or with posteroventral margin obliterated
so that pronotum and mesopleuron are
not differentiated (derived).

The primitive state is generally present
in the Pseudophotopsidinae, Mutillini and
many Sphaeropthalmini; the derived state
is found in the other groups, generally
differing in details in the different taxa.
This means that many of the higher taxa
can be characterized by detailed formula-
tions of the type of derived state involved
in each. Nevertheless, it is difficult to do

this unequivocally. These states also
would provide no information on supra-
tribal groupings. Furthermore, there is
much variation in the degree of expres-
sion of these derived states, especially in
the Sphaeropthalmini.

F56r. Mesopleuron dorsally protuberant
(primitive) or depressed (derived).

Although the derived state is generally
characteristic of the Mutillini (except for
Odontomiitilla) and Pseudophotopsidinae,
it is also present in a few other genera so
that it has occurred in parallel or con-
vergently on various occasions. Further-
more, the range of variation present pre-
cludes accurate delimitation of these states.

F57r. Mesopleural ridge with ventral
(often carinate) section anterior to (prim-
itive) or dorsal to midpoint of (derived)
mid-coxae.

Although the derived state tends to be
characteristic of the Ticoplinae, unequivo-
cal determination of the state is often dif-
ficult because of intermediates and also be-
cause of lack of development of the ridge.

F58r. Ventral section of mesopleural ridge
not (primitive) or sharply (derived)
carinate.

The derived state is found in most
members of the Ticoplinae, Myrmillinae
and Mutillini, and sporadically elsewhere,
so that it shows many instances of parallel
or convergent development. Intermediates
also preclude unequivocal assignation of
various members to one or the other state.

F59r. Meso-metapleural suture distinct
(primitive) or obliterated and indistin-
guishable (derived).

Since the meso- and metapleura are
fused in all members, there is a general
tendency for the meso-metapleural suture
to become obliterated, so that the derived
state is expressed in many members inde-
pendently in the various subfamilies and
tribes involved. Furthermore, unequivocal
categorization of the degree of distinctness

Phylogeny and Classification of the Aculeate Hymenoptera

615

of the suture is impossible because of con-
tiiuious variation from a quite distinct
state to one in which the suture is com-
pletely indistinguishable.

F60r. Metapleural-propodeal suture dis-
tinct (primitive) or obliterated (derived).
As for character F59r, there is a gen-
eral tendency for this suture to become
obliterated so that the derived state shows
many parallel and convergent occurrences,
and also cannot be characterized unet^uiv-
ocally.

F6lr. Calcar on front tibia with an elon-
gate, narrow blade (primitive) or with a
short, broad blade (derived).

Although all Mutillidae except Myr-
mosinae, Ticoplinae and Rhopalomutilli-
nae tend to have the derived state, there
is actually continuous variation so that
definite decisions on which state is present
are impossible in many instances.

F62r. Pectinal spines on fore tarsi barely
developed (primitive) or highly elaborated
(derived).

Although highly developed pectinal
spines are mainly characteristic of many
members of the Sphaeropthalminae and
Mutillini, and also Pseudophotopsidinae,
there is much variation within these
groups so that it is impossible to decide
unequivocally which state is involved in
many cases. The derived condition has
also arisen independently in various
groups.

F63r. Mesotibia with two or more longi-
tudinal rows of spines (primitive) or with
only one or no rows of spines (derived).
Although degree of spininess is vaguely
correlated with tribal groupings, the de-
rived state has arisen on many indepen-
dent occasions. There is also essentially
continuous variation in the character so
that meaningful states are impossible to
define.

F64r. Metatibia with two or more longi-

tudinal rows of spines (primitive) or with
one or no rows of spines (derived).

As with character F63r, the less spiny
tibiae are mainly characteristic of some
Sphaeropthalmini, Ephutini, Myrmillinae,
Rhopalomutillinae and Ticoplinae, but ex-
treme variability and impossibility of pre-
cise definition preclude the use of this
character.

F65r. First metasomal segment sessile and
not apically constricted (primitive) or pet-
iolate, narrowed and apically constricted
(derived).

Although the derived state tends to be
restricted to various members of the
Sphaeropthalminae, it has arisen on nu-
merous occasions independently within
the subfamily and is, furthermore, difficult
to describe unequivocally since there is
considerable variation in form. The de-
rived state is also present in a less extreme
form in many members of the Smicro-
myrmina.

F66r. First metasomal tergum extending
over entire length of segment (primitive)
or absent on anterior half of segment
which is thus comprised of the sternum
only (derived).

The derived state occurs in the Chy-
photinae (Bradynobaenidae), so that with
removal of this taxon from the Mutillidae,
this character becomes invariant in the
myrmosid-mutillid complex.

F67r. Second metasomal sternum without
(primitive) or with (derived) a secretory
felt line on each side.

The derived state occurs in scattered
genera throughout the Mutillidae except
for the Myrmosinae, Rhopalomutillinae,
Sphaeropthalmini and Mutillinae, and
thus may reflect a general tendency for
the development of felt lines (whether
tergal or sternal) in the Mutillidae (above
the Myrmosinae), or may simply show
numerous independent derivations.

F68r. Mesal stridulitrum on third meta-

616

The University of Kansas Science Bulletin

somal tergum absent (primitive) or pres-
ent (derived).

After removal o£ the Typhoctinae,
Apterogyninae and Chyphotinae, this
character is invariant in the myrmosid-
mutilhd complex, being present only in
the derived state.

M69r. This character is the same as F44r,
but for the male, and the same comments
apply.

M70r. Head in anterior view tending to
be vaguely triangular with the clypeal
margin fairly short (primitive) or more
rounded with the clypeal margin longer
(derived) (see Fig. 80).

Although there is a striking difference
in the shape of the head, the derived state
being characteristic of all taxa except the
Ticoplinae and less obviously the Myr-
mosinae and Pseudophotopsidinae, and
thus providing useful information estab-
lishing the holophyly of all Mutillidae
above the three basal subfamilies, it is
impossible to formulate the character states
unequivocally so that they can be ob-
jectively applied.

M71r. Compound eye large (primitive)
or small (derived).

Although this character is similar to
F45r, and it was rejected for the same
reasons, the distribution of the derived
state is somewhat different. In the male
small eyes occur in some Dasylabrini,
most Sphaeropthalmini and most Myr-
millinae.

M72r. This character is the same as F46r
and the same comments apply.

M73r. Antennal scrobe dorsally unarmed
(primitive) or with a tooth or transverse
carina (derived).

Although the derived state tends to be
present in most members of the Sphaerop-
thalminae, it also occurs sporadically in
almost all the other subfamilies of Mutil-
lidae. Unequivocal decisions on which

state is present are also often impossible,
mainly because of the heavy sculpturing
of many species.

M74r. This character is the same as F47r
and the same comments apply.

M75r. This character is the same as F48r
and the same comments apply.

M76r. First flagellar segment subequal to
the second (primitive) or less than half
the length of the second (derived).

The derived state is characteristic of
most members of the Ticoplinae and also
some Dasylabrini and a few Sphaeropthal-
mini and Mutillini, but cannot be une-
quivocally determined because of contin-
uous variation between the two states.

M77r. This character is the same as F51r,
but the derived state is more commonly
present in the male, appearing in many
members scattered throughout the Mutil-
lidae.

M78r. Hind margin of pronotum shal-
lowly arcuate (primitive) or deeply con-
cave or angulate (derived).

Although the derived state tends to
be characteristic of the Rhopalomutillinae,
Myrmillinae and Mutillinae, scattered
members of the other three higher sub-
families of Mutillidae also show it. The
Myrmosinae have a more primitive con-
dition than most other Mutillidae for this
character. In addition, continuous varia-
tion precludes unequivocal decisions on
which state is involved in many instances.

M79r. This character is the same as F53r
and the same comments apply.

M80r. This character is the same as F54r
and the same comments apply.

MSlr. Mesoscutum with the parapsidal
sutures and the parapsidal furrows both
present (primitive) or with one (or both)
of these paired "sutures" obliterated (vari-
ous derived states).

Although most members show the

Phylogeny and Classii-ication of the Aculeate Hymenoptera

617

[irimiiivc slate, some scattered members
of the Sphaeropthalmiiii have lost the
parapsidal sutures, as has Smicromyrmilla.
The [xirapsidal furrows have been lost
onlv rarely, e.g., in Stenomittilla. Both
"sutures" are never lost simultaneously, it
appears. Apart from the limitations
caused by scattered occurrence, it is often
difficult to ascertain the state, both because
the detrree of reduction of the "suture"
varies and because it may be obscured by
the sculpturing.

M82r. Axilla smoothly rounded postero-
laterally (primitive) or produced into an
obtuse or acute tooth (derived).

Although most members have the
primitive state, various scattered genera,
especially in the Sphaeropthalminae, show
the derived state, which is furthermore
difficult to designate unequivocally be-
cause of continuous variation.

M83r. Meso-metapleural suture fused for
only a short distance ventrally (primitive)
or fused for half its length or more
(derived).

This character is associated with M27
in that only those members possessing
state M27.1 have the meso-metapleural su-
ture at all fused. This excludes the Myr-
mosinae, Pseudophotopsidinae and Ticop-
linae froin consideration. Most other
mutillids possess the derived state of M83r,
but various scattered members of the
Sphaeropthalminae in particular have the
primitive condition. Since there is also
continuous variation it is impossible to
designate the states unequivocally.

M84r. Metasternal process bidentate (prim-
itive) or acute (derived).

This character is related to M28 in
that it only applies to members possessing
state M28.1. The Myrmosinae are thus
excluded. The derived state of M84r has
occurred on various occasions, being pres-
ent in Nanomiitillu (Ticoplinae), Rho-
palomutillinae, some scattered Sphaerop-

thalmini and Ephutini. Although the
details of modification dilTer among the
dilTerent groups involved, the differences
are often subtle and difHcult to designate
unequivocally.

M85r. Metapleural-propodeal suture below
the endophragmal pit not evident (primi-
tive) or marked by a secondarily devel-
oped ridge (derived).

The primitive state is present in the
Myrmosinae, Pseudophotopsidinae and
Ticoplinae, and also in various scattered
members of the Sphaeropthalminae, so
that the derived condition must have
arisen on numerous independent occa-
sions. Because of essentially continuous
variation and obscuration by sculpturing,
unequivocal determination of the state
represented is often impossible.

M86r. Tegula with inner margin approx-
imately straight (primitive) or markedly
concave as a result of posteromesal expan-
sion of the tegula (derived).

Although most members possess the
primitive state, a very few scattered genera
(e.g., Areotilla, Odontomiitilla, Ctenotilla,
Dolichomutilla) show the derived state,
the definition of which is also somewhat
ecjuivocal.

M87r. The forewing has three (primitive)
or two (derived) or even one (further
derived) submarginal cells (i.e., cell 2S or
both 2S and IS obliterated in the two
derived states, respectively).

Many members scattered throughout
the Mutillidae possess the first derived
state (within some genera, e.g., Dasylabris,
both the primitive and first derived states
are found). Relatively few members pos-
sess the second derived state, but these are
also scattered (e.g., Ticopla, Atillitm).
Thus both derived states have arisen on
numerous independent occasions.

^r. This character is the same as F61r
and the same comments apply.

618

The University of Kansas Science Bulletin

M89r. An arolium is distinctly present
(primitive) or is absent (derived).

This character is comparable to F16
but shows a different distribution o£ states.
After the removal of Typhoctinae, Chy-
photinae and Apterogyninae, all members
of the myrmosid-mutillid complex show
the primitive state for character M89r.
This character is thus invariant despite the
fact that the females of Myrmosinae pos-
sess the comparable derived state (F16.1).

M90r. This character is the same as F65r
except that in the primitive state the first
metasomal segment is not as broad as in
the female. Similar comments apply as
to F65r, except that the derived state of
M90r is present in the Rhopalomutillinae
also. Furthermore, a different derived
condition in which the first metasomal
segment is apparently broadened seems
characteristic of the Mutillina and to a
lesser extent of the Ticoplinae and some
Myrmillinae.

M91r. This character is the same as F66r
and similar comments apply. It appears,
however, that the derived state of this
character is actually not present in the
males of at least some Chyphotinae, but
has been misinterpreted in the past.

M92r. Second metasomal sternum with-
out (primitive) or with (derived) a secre-
tory felt line on each side.

This character is the same as F67r but
shows a different distribution in that the
derived state is much more commonly
present in the male, being found in scat-
tered members of all taxa of the Mutillidae
except for Myrmosinae, Rhopalomutillinae
and Ephutini. (In addition some species
of Dasymittilla and Traiimatomutillu bear
midventral felt lines on the second ster-
num.)

M93r. This character is the same as F68r
and the same comments apply.

M94r. Hypopygium simple (primitive)

or with an upcurved aculeate spine (de-
rived).

After removal of the Apterogyninae
and Chyphotinae, all members of the
myrmosid-mutillid complex possess the
primitive state, although Myrmosa and
Rlwpalotvttilla have the hypopygium
modified i i various ways.

M95r. Base of gonostylus approximately
vertical (i.e., at right angles to longi-
tudinal axis of genitalia) (primitive) or
oblique (i.e., originating dorsally at a
point more proximal than the ventral
point of origin) (derived).

Although the derived condition is the
more common, the primitive state occurs
in the Pseudophotopsidinae and many
scattered members of the Sphaeropthal-
minae so that there must have been nu-
merous independent derivations. Also, be-
cause of continuous variation, unequivocal
designation of the states is impossible.

M96r. Gonapophyses IX symmetrical
(primitive) or with the right gonapophysis
larger and more elaborate than the left
(derived).

The derived state occurs in only a few
members of the Smicromyrmina (viz.,
"Lobotilla," "Timitlla," "Trogaspidia")
and is thus only of use at the generic level.

Discussion

Taxonomic Conclusions

The final cladogram (Fig. 92) includes
nine taxa, each of which is internally
homogeneous for the various states of the
particular characters considered (except
for the few instances detailed in the above
section describing these characters). As
for the study of the Aculeata, values of
DF, DE, DC and DT (various weighted
measures of distinctness) were calculated
(Tables IV and V). Although various
characters occurring in both sexes were
considered separately in cladogram con-

Phylogeny and Classification of the Aculeate Hymenoptera

619

Sphaeropthalmini

Ephutini

Dasylabnni

Rhopalomutillinae

Pseudophotopsidinae

Myrmosinae

Fig. 92. Cladogram of nine taxa of Mutilliclae;
lengths of heavy lines are proportional to the dis-
tinctness measures (DC). Mutillini (comprising Mu-
tillina and Smicromyrmina) + Ephutini = Mutilli-
nae; Dasylabrini + Sphaeropthalmmi (comprising
Pseudomcthocina and Sphaeropthalmina) = Sphacr-
opthalminae.

struction, they were combined and coded
as to occurrence in the same fashion as
for the aculeates for calculation of these
values (i.e., a derived state occurring in
only one sex on a particular internode was
scored as 0.5 if it occurred in both sexes
elsewhere on the tree). It should be noted
that only the characters considered the
most significant in deriving the cladogram
have been included here, unlike for the
Aculeata where all the characters con-
sidered were used. What influence this
may have on the conclusions is unclear,
although "poor" characters might be ex-
pected to occur at random on the tree and
thus not affect the relative proportions of

"distinctness" for the various internodes.
Certainly, the intuitive impressions gained
during both investigations were supported
by the values derived in each case.

The highest value for taxonomic dis-
tinctness (DT; Table V) between ad-
jacent taxa is that for pseudophotopsidines
and ticoplines, despite the fact that the
myrmosines have commonly been ex-
cluded from the Mutillidae and so might
be expected to show the greatest distinct-
ness. The fact that the pseudophotopsi-
dines are more distinct from the higher
mutillids (minimum, 136) than from the
myrmosines (108) strengthens the inclu-
sion of the myrmosines in the Mutillidae.
The ticoplines are also quite distinct from
the rhopalomutillines (79) which in turn
are slightlv more distinct from the closest
higher taxa (97 to myrmillines, 91 to dasy-
labrines). This suggests that the four
basal taxa at least should each be recog-
nized as distinct at the subfamily level.
The myrmillines are somewhat less dis-
tinct from neighbouring taxa but approxi-
mately equally so (60 to both mutillines
and dasylabrines). The two lowest values
are those between the mutillines and ephu-
tines (50) and the dasylabrines and sphae-
ropthalmines (37) indicating that these
pairs should perhaps not be recognized at
the subfamily level. The intermediate po-
sition of the myrmillines between two
complexes which are highly distinct
(minimum value 106), mutillines to dasy-
labrines) suggests that the three groups
containing myrmillines, mutillines and
dasylabrines each be recognized at the
subfamily level.

The cladogram is thus interpreted to
encompass one family, the Mutillidae, con-
sisting of seven subfamilies, two of which
comprise two tribes each. Two of these
tribes (in different subfamilies) are in
turn divided into two subtribes each, in
both instances mainly on the basis of
characters excluded in cladogram construe-

620

The University of Kansas Science Bulletin

Table IV. Occurrences (Z^i) and numbers (k) of derived states, number of species subtended (S)
and various distinctness measures (DF, DE, DC) for the internodes and final branches of the cladogram

of MutilHdae (Fig. 92).

Internode

2i

k

S^

DF

DE

DC

1-Myrmos.

3.0

4

50

11.04

0.83

9.16

1-2

5.0

5

4850

84.65

1.00

84.65

2-Psdpht.

5.0

5

50

18.40

0.80

14.72

2-3

7.0

7

4800

118.09

0.90

106.28

3-Ticopl.

4.0

4

75

16.88

0.88

14.85

3-4

3.0

3

4725

50.34

0.83

41.78

4-Rhopal.

7.0

7

50

25.76

0.86

22.15

4-5

3.0

3

4675

50.16

0.83

41.63

5-6

2.0

2

2475

27.06

1.00

27.06

6-Dasyla.

0.0

0

550

0.00

0.00

0.00

6-Sphaer.

3.0

3

1925

37.32

1.00

37.32

5-7

2.0

2

2200

26.00

1.00

26.00

7-Myrmil.

1.0

1

400

7.37

1.00

7.37

7-8

3.0

3

1800

36.48

0.83

30.28

8-Mutiln.

2.0

2

1500

22.90

1.00

22.90

8-Ephutn.

4.0

4

300

26.76

1.00

26.76

* Based on the sum of the estimated number of species in each genus (including those yet to be
described). Calculations utilize '■^S.

tion. The genera and subgenera included
in these various taxa are Hsted in Table III.

Mutillidae Latreille, 1802

The Mutillidae as a whole is charac-
terized by some quite strong synapomor-
phies, as outlined in the section on the
Aculeata, above. The most significant of
these are probably the development of a
complex basal angulation of the scape,
fusion of the moderately large prepectus
to the mesepisternum, form of the modi-

Table V. Taxonomic distinctness for the taxa of
Mutillidae.

•3'

<?■

.*

/

/ / .^
^ / / /

2,5 ,3. / / # /

264 185 79 ^ ,v /^

^ ^ S .^

310

231

125

91

•Q''

f

343

269

163

128

37

i'

•N

317

233

132

97

60

98

363

284

177

150

106

144

60 4V

/

366

287

181

154

no

147

64 50

♦

fied mesosoma of the female (which is
invariably apterous) and the development
of a mesal stridulitrum on the third meta-
somal tergum.

Despite past attribution of the name
Mutillidae to a wide variety of authors,
it appears that Latrielle's treatment was
the first in which a group based on a
genus included in this taxon was allocated
a rank at the family group level.

Myrmosinae Fox, 1894

With the suite of characters in use
here, the Myrmosinae has only five char-
acters in the derived state, three of these
being unique to this subfamily. In two
there is parallel development of a derived
condition in the Myrmosinae and also
within the remaining Mutillidae, viz., un-
armed claws and an unarmed mesoster-
num, both in the female only. Of the
other three characters, the development of
a dorsal lamella on the hind coxa in both
sexes is probably the most significant. The
remaining Mutillidae are distinguished
from the Myrmosinae by the presence of
six characters in the uniquely derived
state. Of these, two involve the develop-

Phylogunv and Classification of the Aculeate Hymenoptera

621

mcnt of an armed meta stern urn in both
males and females, considered a single
character for calculation of distinctness.
The most useful character states uniting
these higher taxa are the fusion of the
pronotum with the mesothorax in the
female and the form of the gonapophysis
IX in the male. Although the absence of
felt lines in the Myrmosinae has been con-
sidered to exclude them from the Mutilli-
dae (e.g., Krombein, 1940), the develop-
ment of felt lines in the higher subfamilies
is somewhat irregular, so that this is not
as fundamental a character as has been
considered in the past.

The Myrmosinae is a relatively homo-
geneous group without any obvious sub-
divisions so that there does not seem to
be any reason to subdivide it, at least not
at this time. Protomiitilla is perhaps the
most anomalous member, and it may later
be found that this genus is distinct enough
to warrant the erection of a separate supra-
generic taxon to contain it. Until the
male is discovered, however, any such de-
cision would be premature. Although
Ghesquiere (1951) placed the genera
Obenbergerella Strand (= Alien us Brid-
well) and Alienisciis Benoit as the tribe
Obenbergerellini in the subfamily Myr-
mosinae of the Tiphiidae, Krombein
(1957) included these genera in the Ami-
seginae, placed by him in the Chrysididae
(Bethyloidea). Nagy (1969c) considered
this group to fall in the Cleptidae. The fe-
males have 13-segmented antennae, only
four exposed metasomal segments and
only one spur on the hind tibia, among
other characters. These characteristics ex-
clude the group from the Myrmosinae but
not from the Cleptidae or Chrysididae
(sensii lata ).

Handlirsch (1925) cites Ashmead
(1896a) as the first to base a taxon of the
family group on the genus Mynnosa. It
appears, however, that Fox should be
cited as the author since the paper in

which he established the tribe Myrmosini
was published before that of Ashmead.

Pseudophotopsidinae Bischoflf, 1920

The Pseudophotopsidinae is distin-
guished from the higher Mutillidae (those
above internode 2-3) by a relatively large
number of characters. The most impor-
tant of these are probably the retention of
ocelli (or their rudiments) in the female
in many species, retention of a distinct
suture between the pronotum and meso-
notum in the female, retention of a
straight meso-metapleural suture in the
male, retention of teeth on the claws in
both sexes, retention of a jugal lobe in the
hind wing of the male, retention of a
rounded gonostylus in the male (all prim-
itive states), development of a felted pro-
notal pit and mesosternal teeth in both
sexes and development of articulating
spines on the gonapophysis IX in the male
in the Pseudophotopsidinae.

Although Schuster (1950) suggested
that Krombein (1940) should be con-
sidered the author for this group, he ap-
parently overlooked the fact that Bischof?,
in his 1920-21 monograph, had based a
tribe on the genus Pseudophotopsis, in the
section of that work published in 1920.

Ticoplinae Nagy, 1970

The Ticoplinae differ from the higher
groups of Mutillidae (i.e., those subtended
by internode 3-4) by a number of char-
acters, of which the following are probably
the most significant: retention of setae
and minute pores in the compound eye
of both sexes, retention of an articulating
meso-metapleural suture in the male (i.e.,
no development of a ventral bridgelike
fusion) (both primitive characters al-
though only the latter is shared with both
the Myrmosinae and Pseudophotopsidi-
nae), development of a posteriorly convex
mesopleural margin in the male and de-
velopment of a petiolate cell IS in the
forewing of the male. In addition the

622

The University of Kansas Science Bulletin

form of the head (especially in the male)
tends to be quite different from the other
Mutillidae where the oral fossa appears to
have been laterally expanded (less mark-
edly so in the Pseudophotopsidinae, and
barely so in the Myrmosinae, however).
In general facies the females of the Tico-
plinae appear to be more similar to the
Myrmosinae than do any of the other
higher Mutillidae. This is especially true
of Nanomiitilla which is uncannily similar
to Protomiitilla, whereas Areotilla is more
similar to some species of Myrmosa. Sivi-
cromynnilla is more highly modified in
many respects and so shows less obvious
similarity to the Myrmosinae.

In 1970 Nagy described the genus
Ticopla (comprising two species) from
two male specimens, both from the Jordan
region. He made this genus the type of
a new subfamily of his family Heterogyni-
dae. (It appears that Heterogyna Nagy
is actually a member of the Plumariidae
— especially on the basis of wing vena-
tion— and it has been treated as such in
the above investigation of the Aculeata.)
Although it has been impossible for me
to examine Nagy's specimens, it seems
certain from his figure and description
that Ticopla is very similar to Naiwtna-
tillu. Ticopla has the wing venation more
reduced, however, and the genitalia may
be somewhat dilTerent althcjugh Nagy's
description of these is not detailed enough
for certainty. In particular, these genera
are similar in various characteristics cited
by Nagy as indicating relationships of
Ticopla to the Bethylidae rather than the
Mutillidae, such as the position of the
compound eye (very short malar space),
setae on the eye, acute posterolateral angle
and pattern of sculpturing of the propo-
deum, and the prominent setae on the
major wing veins. It thus seems that
Ticopla is in actual fact a member of the
Mutillidae, closely related to Nanomittilla.
On this basis, the subfamily including

Nanomittilla (and thus Ticopla) must be
designated the Ticoplinae since Nagy has
already based the name of a taxon of the
family group on that genus. It may even
be that a species of Ticopla is actually the
male of Nanomiitilla vaiicheri (Tournier),
females of which have been collected in
the same region (Andre, 1902). If this
should prove to be true, then Ticopla
would have to be synonymized with
Nanomittilla. although the subfamily
name would be unchanged.

Within the Ticoplinae there are no
particularly marked groupings (other than
that of Nanomittilla and Ticopla already
mentioned), so that tribal divisions are
unwarranted. The most distinct genus is
probably Smicromyr?nilla which possesses
sternal felt lines as well as a variously
armed mesosoma. The female in particu-
lar has a superficial similarity to some
members of the Myrmillinae, especially in
mesosomal form, as has been noted by
Suarez (1965). The males of Smicromyr-
milla also show a complete gradation from
fully winged forms to species without any
trace of sutures on the mesosomal dorsum,
a condition found elsewhere only in
Brachymiitilla (Dasylabrini). Neverthe-
less, the least-modified members (espe-
cially the males) are reasonably similar to
Nanotniitilla and were placed in this ge-
nus by Bischoff (1920-21). This place-
ment may, however, have been due in
considerable part to the fact that Bischoff
had seen neither the type species (N.
vaitcheri) nor the only other species placed
in Nanomittilla by Andre (the author of
the genus) up to that time, but appar-
ently based his generic allocation m.ainly
on the presence of a petiolate cell IS,
which is now seen to be a subfamily char-
acteristic. Although Nonveiller (1973)
recognized the confusion then existing be-
tween Nanomittilla and Smicromyrmilla,
he had apparently not seen any male speci-
mens of Nfl/7o/?;;///7A/.

PllVLOGliNY AND CLASSIFICATION OI- THE AcULEATE HyMENOPTERA

623

Rhopalomutillinae Schuster, 1949

The RhopalomutilHnae difTcr from the
higher MutiHidae (i.e., those siihtended
hy internode 4-5) in the retention of a
relatively long pronottim in the female,
the absence of any felt lines and the re-
tention of a small metacoxal tubercle in
both sexes (all primitive states) and from
all other members of the family in the
reduced maxillary and labial palpi of the
female, the form of the mesosoma and
first metasomal segment in the female
(although a somewhat similar but less ex-
treme development of the metasoma has
occurred in the Sphaeropthalmini in the
complex of species related to "Ephutomor-
pha" addenda), and the development of
a basal lamella on the tarsal claws in the
male. In addition, the hypopygium of the
male is much modified, being reduced
and often with complex protuberances, a
condition unlike that in other Mutillidae
(the hypopygium is also reduced but dif-
ferently modified in many Myrmosinae).

Although Schuster in 1947 considered
Rhopalomiitilla to be of doubtful position,
in or near the Sphaeropthalminae, he did
not designate any family level group to
contain it at that time. However, in his
1949 treatment, Schuster used the term
"Pseudophotopsidinae-Sphaeropthalminae-
Mutillinae-Rhopalomutillinae complex"
when referring to the taxon here con-
sidered to comprise the Mutillidae (except
for the Myrmosinae). He also gave a few
characters (pp. 123, 125) differentiating
the Rhopalomutillinae from other groups.
This is apparently the first instance in
which a family level group name was
based on Rhopalornutilla.

Sphaeropthalminae Schuster, 1949 (1903)

The Sphaeropthalminae differ from
the two higher subfamilies (subtended
by internode 5-7) in their retention of a
completely sclerotized pterostigma and
also in the form of the mesosoma in the

female and the direction of curvature of
the gonostylus in the male.

In his classification of the Mutillidae,
Ashmead (1903-4) based two tribes on
genera which are included in the present
concept of the Sphaeropthalminae. These
were the Photopsidini (in his subfamily
Mutillinae) and the Sphaerophthalmini
(sic) (in his Ephutinae). Although the
name "Photopsidinae" was used by Brad-
ley & Bec]uaert (192(S) and by Schuster
in the first paragraph of his 1947 paper,
Schuster in the same paper later spe-
cifically designated this group as the
Sphaerophthalminae (sic), because ''Pho-
topsis is not generically distinct from
Sphaewphthahna'' (sic). [The confusion
in the spelling of the generic name has
arisen because Blake's (1871) original
spelling, Sphaeropthalma, was an incorrect
transliteration from the Greek — aipaipa a
ball, and o(/)^aA^o? an eye — which Blake
corrected in 1886 to Sphaerophthalma.
This corrected spelling was used uni-
formly from then on until Schuster (1949)
reverted to the original spelling as the
basis for the name "Sphaeropthalminae."
The "corrected" (1886) form is an un-
justified emendation under the provisions
of Articles 32(a) & 33(a) of the Interna-
tional Code of Zoological Nomenclature
(1964), since Blake's change was obviously
intentional, appearing consistently in
many places in his 1886 paper. As such,
Sphaerophthalma Blake, 1886 must be con-
sidered a junior objective synonym of
Sphaeropthalma Blake, 1871, rather than
merely a misspelling.] This change in the
name has won general acceptance and was
used by Krombein (1951) in his catalog
of the Mutillidae of America North of
Mexico, with the additional change of
spelling resulting from Schuster's (1949)
correction of the spelling of the type
genus. The appropriate name to be con-
served under Article 40 of the Interna-
tional Code of Zoological Nomenclature

624

The University of Kansas Science Bulletin

(1964) is thus "Sphaeropthalminae," which
is considered to date from 1903 when it
was in effect first used by Ashmead. (The
citation o£ author and date, as in the
heading to this section, follows Recom-
mendation 40A of the Code.)

Dasylabrini Invrea, 1964

The Sphaeropthalminae contains two
distinct groups, one of which (Sphaerop-
thalmini) is recognized by a few quite
distinct derived states. The Dasylabrini,
however, does not possess any particular
derived characteristics which would indi-
cate it as a holophyletic group. It may
thus be that it is actually paraphyletic,
representing those members of the Sphae-
ropthalminae which do not fall in the
holophyletic tribe Sphaeropthalmini. Since
homogeneous paraphyletic groups fulfill
the criteria of monophyly generally re-
quired of named taxa (Tuomikoski, 1967;
Ashlock, 1971), recognition of the Dasy-
labrini as a named taxon is not illegiti-
mate. Within the Dasylabrini there are
no particular subgroups recognizable, so
that the entire taxon is best regarded as a
single tribe. Although Skorikov (1935)
seems to have been the first to propose a
family level group name based on one of
the genera (Dasylabris) included here, he
did not characterize the group. Appar-
ently Invrea (1964) was the first to fulfill
all the requirements of the International
Code of Zoological Nomenclature (1964).

Sphaeropthalmini Schuster, 1949 (1903)

The tribe Sphaeropthalmini is distin-
guished from other members of the sub-
family by some unique derived states in
both sexes. These are the form of the eye
which is approximately hemispherical and
often smooth and polished (particularly
in the male) and the development of plu-
mose pubescence. These characteristics
unite a large and varied group of genera
which Schuster (1947) considered to com-
prise at least five taxa, each apparently at

the tribal level, although this was not
explicitly stated. In 1958 Schuster revised
one of these groups and Krombein (1967)
designated it a tribe, so that his "Sphaerop-
thalmini" is more restricted than the
group included here under that name.

When the members of the Sphaerop-
thalmini were examined more closely,
using many of the characters which were
rejected for the study as a whole, no very
distinct groupings were recognizable. The
most distinct subgroup is that including
Pseiidomethoca and related taxa, mainly
based on the form of the female meso-
soma and the tendency for both sexes to
have the metasoma sessile. The remaining
members are associated by a tendency to-
ward development of a petiolate metasoma
in the male and sometimes in the female,
but do not fall into any further clear sub-
groups. There is instead a continuum
from those genera with fewest derived
states to those with many. Although no
final decisions can now be made on the
evolutionary relationships among these
genera, since additional investigations and
clarification of generic limits are needed,
it is nevertheless clear that the genera
nearest the base of the line are those like
Cystomittilla, Photomorplnts and Proto-
photopsis, with the moderately advanced ||
ones being similar to Sphaeropthahna,
Lomachaeta and Dasymutilla, and the
most highly derived ones being some
members of the ''Ephittomorpha com-
plex," such as Ascetotilla and the group
of "Ephittomorpha" paradisiaca.

At this point it seems best to consider
the tribe Sphaeropthalmini to consist of
two more or less equal subdivisions which
may be given the status of subtribes. The
group with a generally sessile metasoma
is the Pseudomethocina. [Although Schu-
ster (1947) characterized this group, he
referred to it as the "Pseudomethocine
complex." Suarez (1962) was apparently
the first to Latinize the name (as Pseudo-

PllVLOGtNV AND ClASSIMCATION Ol- THli AcULEATIi HyMENOPTERA

625

methocini), luil lie did nol accompany il
with aiiv dilTcrentiating characters, nor
has this been done since. It thus seems
that the name should formally be con-
sidered to date from the present paper.]
The group with the metasoma tending to
be petiolate (in the male at least) is the
Sphaeropthalmina Schuster, 1949 (1903).
Additional and more detailed investiga-
tions of this tribe may cjuiie likely indi-
cate that fewer or more such subtribal
divisions are warranted.

Myrmillinae Bischoflf, 1920

The Myrmillinae differ from the Mu-
tillinae in the retention of a simple, short
tegula and broadly oval eye with shallowly
sinuate inner margin in the male (both
primitive states) and also in the form of
the pterostigma in the male. The form
of the mesosoma in the female is also
different, with the pleura in the Myrmil-
linae tending to be evenly concave and
rather smooth (except that the meso-
pleural ridge tends to form a strong carina
or lamella ventrally just anterodorsal to
the mid-coxa), and the pronotal-meso-
pleural suture is essentially obliterated.
Female Myrmillinae also tend to have the
head rather heavy, often roughly quadrate
with the gena somewhat swollen and with
the mandible very strong and broad
apically. Although the females of Myr-
millinae are sometimes difficult to allocate
unequivocally to this subfamily without
prior experience, having few good features
to distinguish them from Mutillini, the
males are quite distinct.

Since this family is confined to the
Old World, its members were apparently
not included by Schuster (1947) in his
general scheme. Interestingly, Skorikov
(1927, 1935) placed this subfamily in the
Myrmosidae. Bischoff (1920-21), appar-
ently the first to base the name of a
group at the family level on that of one
of the genera included here in the Myr-

millinae, derived the myrmillines as a
separate branch from a myrmosid an-
cestor but included them in his Mutillinae
and not the Myrmosinae!

Within the Myrmillinae there are no
definite subgroups of genera which could
be designated as tribes. It is of some in-
terest that this subfamily includes the
highest proportion of species with brachyp-
terous or apterous males of all the sub-
families. Various genera contain both
fully winged and wingless species (e.g.,
Mynnilla, Labidomilla) although the
range of variation in these is not as great
as in Smicromyrmilla (Ticoplinae) where
the most advanced forms have in addition
lost all traces of sutures dorsally on the
mesosoma.

Mutillinae Latreille, 1802

The Mutillinae is characterized by de-
velopment in the male of a posteriorly
produced tegula, deeply and sharply emar-
ginate eye and a completely membranous
pterostigma. (The pterostigma is further
entirely lost in some members.)

Mutillini Latreille, 1H02

This tribe is distinguished by the de-
velopment of a recurved posterior margin
of the tegula in almost all genera and, in
the female, by the dorsal depression of the
mesopleural ridge and modifications of
the meso-metapleural suture (except for
Odontomiitilla).

The tribe Mutillini itself consists of a
large number of genera which fall into
two reasonably distinct groups. The
smaller group consists of a few genera
with males which tend to have the meso-
soma somewhat compact and the meta-
soma completely sessile, with the first
tergum somewhat broadened. The fe-
males also have the first metasomal ter-
gum broadened and often almost disclike.
This group may be designated the sub-
tribe Mutillina Latreille, 1802. The other,
larger, group has males which are gen-

626

The University of Kansas Science Bulletin

erally slightly more slender, with the first
metasomal tergum almost campaniform,
narrower and weakly demarcated from
the second. The females also have the
first metasomal segment narrower and
sometimes almost petiolate. This second
group is the subtribe Smicromyrmina
Bischoflf, 1920. Actually this subtribe in-
cludes two of Bischoffs (1920-21) tribes,
his Smicromyrmini and Trogaspidiini. In
contrast to Bischoflf's treatment, Bradley
& Bequaert (1923) placed many of the
genera included in both "tribes" under
the genus Smicromyrme. Mickel (1933)
once more considered Smicromyrme and
Trogaspidia to be genetically distinct (al-
though he placed the latter as a subgenus
of Timulla). Since that time various au-
thors of major works have treated Tro-
gaspidia either as a subgenus of Timidla
(e.g., Mickel, 1935; Krombein, 1971) or
as a distinct genus (e.g., Olsoufieff, 1938;
Chen, 1957). Furthermore, Krombein
(1972) returned to Bischoff's (1920-21) con-
cepts of tribal classification, but changed
the name of the Trogaspidiini to the
Timullini (a procedure disallowed under
Article 40 of the International Code of
Zoological Nomenclature, 1964). In view
of the confusion surrounding the name
Trogaspidia it thus seems best to use
Smicromyrme (which is also the older
name) as the type genus for this subtribe.

Ephutini Ashmead, 1903

The Ephutini is a very distinct group,
characterized by the peculiar cylindrical
form of the first metasomal segment in
both sexes, the form of the mesosoma in
the female and the tegula in the male, as
well as the loss of the pterostigma.

There are no distinct subgroups within
this tribe so that no subdivisions can be
recognized, although much work needs to
be done to clarify generic limits. It ap-
pears that Ashmead (1903-4) was the first
to base a family-level name on the genus

Epintta, first doing so adequately in the
section of his paper published in 1903.

The morphological investigation thus
indicates that the family Mutillidae may
be considered to be composed of seven
subfamilies, five of these being monotypic;
the other two subfamilies each contain
two tribes and, in each, one of the tribes
comprises two subtribes. This arrange-
ment should be found to be useful and
stable provided that it is based on an
accurate interpretation of the patterns of
evolution within the family. Some indi-
cation of the probable accuracy of a clado-
gram may often be obtained when ex-
ternal data are superimposed on it and
their fit to the scheme is judged. Such
data may be biological (information on
life histories, hosts, etc.) or distributional,
and these two types of data will be evalu-
ated below.

Life Histories and Host Relationships

Unfortunately, rather little work has
been done on the life histories of the Mu-
tillidae. Detailed, although nevertheless
incomplete, information is available for
Chrestomutilla glossinae (Lamborn, 1915,
1916; Heaversedge, 1968, 1969a & b,
1970) (Dasylabrini), Dasymittilla bioculata
(Mickel, 1924; Cottrell, in Brothers, 1972)
and Sphaeropthalma (Photopsis) spp.
(Ferguson, 1962) (Sphaeropthalmina),
and Pseudomethoca frigida (Brothers,

1972) (Pseudomethocina) in the Sphaerop-
thalminae, and also for Smicromyrme
rnfipes (Crevecoeur, 1930; Marechal, 1930)
(Smicromyrmina) and Mittilla eitropaea
(Hoflfer, 1886; Jordan, 1935; Pouvreau,

1973) (Mutillina) in the Mutillini. Rea-
sonably detailed information also exists
for Myrmositla parvula in the Myrmosi-
nae, although this is as yet unpublished
(Brothers, in prep.). There are only brief
observations on mating, feeding and other
details for the subfamilies Pseudophotop-
sidinae, Rhopalomutillinae and Myrmilli-

PllYLOGENY AND CLASSIFICATION OF THE AcULEATE HyMENOPTERA

627

nae, as well as the Ephutini. There is es-
sentially no information on the Ticoplinae.
In view of the paucity of data, it is
impossible to ascertain whether life history
information supports the cladogram or
not. The time and method involved in
mating seem to dilTer according to taxo-
nomic unit. Sphaeropthalmini such as
Dasymittilla (Cottrell, in Brothers, 1972;
Linsley, MacSwain & Smith, 1955), Plw-
topsis (Ferguson, 1962; Salman, in Mickel,
193(S) and Pseiidomethoca (Brothers, 1972)
mate on the substrate, intromission lasting
a few seconds. The males of Smicromyr-
mina (Mutillini) such as Timitlla (Lins-
ley, 1960; Sheldon, 1970), Smicromyrme
(Bertkau, 1S84; Crevecoeur, 1930; Pagden,
1934) and Siilcotilla [specimens in British
Museum (Natural History) collected by
Risbec] often transport the female in
flight before settling and mating. They
may even mate in flight, the male clasping
the female with his mandibles and legs.
In the Rhopalomutillinae also, the male
transports the female (Bridwell, 1917;
Pagden, 1938), but in this case the female
is supported entirely by the attachment of
the male genitalia and by modifications
of the apical sterna of the male. In Rho-
palomittilla tongaana at least, the relative
positions of the male and female are simi-
lar to those figured by Evans (1969a) for
Apenesia (Bethylidae) (and thus unlike
most Thynninae), although the size dif-
ference between the sexes is greater in
Rhopalomiitilla (pers. obs.). In the in-
stance observed, the male was visiting the
flowers of Zizy pints with the female at-
tached passively to his metasomal apex.
In some Myrmosinae, also, the male trans-
ports the female during copulation, with
the female below the male and venter to
venter, the only support of the female
being provided by the genitalia of the
male {Myrmosa; Townes, in Pate, 1947a;
Krombein, 1956). It is obvious that
phoretic copulation is most likely to occur

in species where the male is appreciably
larger and stronger than the female. This
condition occurs most markedly in the
Rhopalomutillinae, many but not all Smi-
cromvrmina and some Myrmosinae, as
well as s(jme species in other groups.
Thus, information on mating may eluci-
date phyletic relationships within the
higher groups but will probably not be
particularly useful in confirming (or re-
futing) the present cladogram based on
adult morphology.

Other biological data which have been
found to be useful in other parasitic
groups, are host relationships. If the para-
sites tend to be at all host specific, they
should show a pattern of evolution which
is compatible with that of the host organ-
isms, unless there have been transfers to
unrelated hosts. Such data have, for ex-
ample, proved to be useful in studies of
mammalian ectoparasites (see Traub, 1972,
for a discussion of the interrelationships
of evolutionary patterns and geographic
distributions of fleas and mammals).
From the available host records, it ap-
pears that ncj parallels can be drawn be-
tween the evolution of the Mutillidae and
that of their hosts. Members of this fam-
ily were almost certainly originally para-
sitoids of ground-nesting Hymenoptera, as
are the Fedtschenkiinae (Bohart & Schu-
ster, 1972). However, host specificity is
apparently often not strict [e.g., Steno-
miitilla argentata parasitizes Eumenidae
(Vespoidea), Megachilidae (Sphecoidea)
and Clythrinae (Coleoptera; Chrysomeli-
dae) (Giner Man', 1944); Pseitdomethoca
jrigida attacks at least eight species of
halictine bee and perhaps a eumenid wasp
(Brothers, 1972) |, and transfers to en-
tirely unrelated hosts (even in different
orders) have apparently occurred sporadi-
cally. Thus, Coleoptera and Diptera are
utilized by a few members of the Dasy-
labrini and Smicromyrmina, Lepidoptera
only by Stenomiitilla (Dasylabrini) and

628

The University of Kansas Science Bulletin

at least some Odontophotopsis may para-
sitize cockroach oothecae (Mickel, 1928,
1974; Seyrig, 1936). As a result, it ap-
pears that host information is of no
use in confirming or refuting the clado-
gram. Instead, members of the Mutillidae
seem to be capable of parasitizing a wide
variety of hosts, the one requirement per-
haps being that the stage of the host at-
tacked be in the form of a more or less
immobile "package" sealed off from the
environment, whether in a closed cell or a
hard cocoon, puparium or ootheca.

Geographic Distribution

Since the Mutillidae comprises exclu-
sively species with low vagility as a result
of the winglessness of the females, it might
be expected that the geographic distribu-
tions of the various subtaxa would reflect
their places (and times) of origin when
considered in the light of the theory of
plate tectonics and continental drift, if
the group is of appropriate age. This
should be more clearcut than for a group
with high vagility which could more
easily cross water gaps. In fact it appears
that the cladogram of the Mutillidae is
consistent with the data on geographic
distribution, and is thus supported (or at
least not refuted) by it. The distribution
of the more primitive subfamily (Fedt-
schenkiinae) of the sister group to the
Mutillidae (the Sapygidae) is also of im-
portance in this regard, since it provides
a clue as to the area in which the ancestor
of the Mutillidae arose. [The more de-
rived Sapyginae are essentially cosmo-
politan— absent only from the Australian
region (Pate, 1947c) — and thus do not
provide any useful data. | Although the
cladogram is not being used here as evi-
dence for former land connections, as was
the purpose of Hennig (1966b) in his
survey of the Diptera fauna of New Zea-
land, the rigorous principles which he
outlined in that paper still apply.

Reconstructions of Pangaea and the
sequence of events leading to fragmenta-
tion and rearrangement of the compo-
nents, together with estimates of the times
involved, have been attempted by various
authors and are being steadily refined.
The following sequence has been derived
from information presented by Dietz &
Holden (1970), Axelrod & Raven (1972),
Fooden (1972), Raven & Axelrod (1972)
and Heirtzler et al. (1973). Most of these
authors have based their concepts on data
derived from a wide variety of sources, in
large part geophysical. This sequence
should, nevertheless, be regarded merely
as the best estimate derivable from their
data.

About 200 million years (m.y.) ago
(during the Triassic) there was apparently
a single land mass, Pangaea, surrounded
by the universal ocean, Panthalassa, an
arm of which formed the Tethys Sea
separating Eurasia and Africa. The first
rift occurred in the west and separated
North America from South America and
Africa, resulting in two continental areas,
Laurasia and Gondwana, with tenuous
contact maintained across the Gibraltan
area (180 m.y., late Triassic — early Juras-
sic). About 150 m.y. ago (middle Juras-
sic), the Indian Ocean first opened from
the east, separating East Gondwana (con-
sisting of Australia-New Guinea, New
Zealand and Antarctica) from West
Gondwana (India, Madagascar, Africa
and South America) although contact was
maintained between Antarctica and South
America. The next rift occurred within
West Gondwana when South America
split off from Africa (110 m.y., mid-Cre-
taceous), although these two continents
remained reasonably close to each other
for perhaps another 20 m.y. because their
relative movement in the north resulted
from a shear fault (similar to the San
Andreas Fault in California). The next
separation was that of New Zealand from

Phylogeny and Classification of the Aculeate Hymenoptera

629

West Antarctica (80 m.y., late Cretaceous),
thus splitting East Gondwana. Shortly
thereafter (70 m.y.) India separated from
Africa and Madagascar and began its mi-
gration north-eastward. About 60 m.y. ago
(late Cretaceous — Paleocene) Madagascar
split off from Africa and, as a result of
rotational movements, the Arabian section
of the African plate came into contact
with Eurasia, cutting off the Mediter-
ranean Sea from the western extremity of
the Tethys. Not long after this (50 m.y.,
Eocene) North America and Greenland
finally separated from Europe, although
this rift had gradually been extending
northwards since its initiation more than
80 million years before. At about the same
time, Australia-New Guinea split off from
East Antarctica and moved northward,
contacting the eastern region of the Asian
plate about 20 m.y. ago (Miocene). India
collided with Asia about 15 m.y. ago and
North and South America were linked
only about 10 m.y. ago (early Pliocene).
In addition, Antarctica only moved to its
polar position within the last 40 m.y., after
Australia had separated from it, its climate
up to that time having been much milder
than presently. Also, New Caledonia re-
mained linked to (or at least close to)
Australia for a short time after New Zea-
land had separated from the rest of East
Gondwana.

The present disjunct Holarctic distribu-
tion of Fedtschenkja (Fig. 93; Guiglia,
1972) probably represents the relicts of a
more widespread group which was un-
doubtedly of Laurasian origin. (The
Sapyginae have apparently radiated from
this area, perhaps relatively recently since
they are not found in the Australian re-
gion.) The Myrmosinae (Fig. 94) are
also confined to the Holarctic region al-
though the single modern representative
of the derived genus Protomiitilla occurs
in the Oriental region. Fossil specimens
of Protomiitilla are, however, from Baltic

amber (Bischoff, 1915), so that members
of this group were probably present
throughout Eurasia. Apparently both the
Fedtschenkiinae and the Myrmosinae
originated on Laurasia and dispersed
throughout this area after the splitting of
Pangaea into Laurasia and Gondwana.
Even after this break, however, there was
still contact (at least intermittently) across
the region of Gibraltar, which may ex-
plain the presence of Myrmosa in North
Africa (although this is perhaps more
likely a later development).

By contrast, the Pseudophotopsidinae
(Fig. 95) is presently distributed across
the northern section of Africa, Arabia and
into southwestern Asia. The Ticoplinae
(Fig. 96) is predominantly African al-
though apparently absent from the Sa-
haran area (perhaps a secondary develop-
ment because of ecological changes), and
also occurs in India and Spain. The Rho-
palomutillinae (Fig. 97) is again mainly
tropical African but with representatives
in the Oriental region.

Since these three are the most basal
groups of Mutillidae (apart from the Myr-
mosinae), it seems that the center of origin
for this section of the family was the
African plate, and probably its northern
section. This likely resulted from the in-
troduction of an ancestral form across the
Gibraltan region at about the time that
the final break occurred between Laurasia
and Gondwana. The ancestral form gave
rise to the Pseudophotopsidinae which ap-
parently spread to the east and eventually
across the Arabian area and on to the
Eurasian plate after contact was reestab-
lished there. The most generalized species
of Pseiidophotopsis is probably P. continua
(ocelli and dorsal mesosomal sutures in
the female the least reduced) which is
widespread in North Africa. The more
eastern species are generally more highly
derived (e.g., P. syriaca), which supports
this pattern of dispersal.

630

The University of Kansas Science Bulletin

Phylog£Ny and Classification of the Aculeate Hymenoptera

631

3

>>

s

3

J5
D.
a

bo
O

V

O

632

The University of Kansas Science Bulletin

Phylogeny and Classification of the Aculeate Hymenoptera

633

u
C

GO
ON

c

Q.

O

ON

3
O

V)

u

X)

3

C

o

3

x:

J5
Q.
«

M
O
u

o

ON

634

The University of Kansas Science Bulletin

The ancestral Ticoplinae probably
arose somewhat farther south and east on
the African plate. This group then may
have diversified and spread on to the
Indian plate while it was still in contact
with Africa or at least close to it. At least
one member apparently crossed the Gi-
braltan area to Europe, probably much
more recently. Although it is ditScult to
state which genus of Ticoplinae is the
most generalized, it seems clear that the
most specialized morphologically is Smi-
cromyrmilla, which is also the most wide-
spread at present, and the one which
occurs in both Spain and India, as well as
Africa. (Some Indian species have the
most highly derived males of the subfam-
ily since they are completely apterous and
lack any traces of sutures on the meso-
somal dorsum.) This is not inconsistent
with the concept of diversification of the
group from an African center.

A similar sequence may have led to
the origin of the Rhopalomutillinae, which
seem to be adapted to more tropical con-
ditions than the Ticoplinae. Rhopalomu-
tillinae also probably spread from the
African to the Indian plates, and when
India came into contact with the Eurasian
plate there was apparently additional di-
versification and dispersal of this subfam-
ily to the east into Indo-China and further
to Borneo and Java. From morphological
considerations it appears that the most
generalized species of Rhopalomittilla oc-
cur in Africa (e.g., R. tongaana; least
modified hypopygium in the male) and
some of the more highly derived are the
Asian species (e.g., R. oceanica), a distri-
bution not inconsistent with the above
ideas.

These three basal subfamilies were
probably confined to the east African re-
gion, at least initially, since none of them
occurs in the New World. This may have
resulted from the presence of epiconti-
nental seas over much of northern and

western Africa during the Cretaceous
(Cracraft, 1973). The absence of the
Ticoplinae and Rhopalomutillinae in
Madagascar is somewhat puzzling and re-
quires further elucidation. These groups
may, however, have become extinct there,
perhaps as a result of competition from
the more advanced subfamilies. Until
more data have been gathered on these
groups, further speculation is pointless.

The subfamily Sphaeropthalminae is
cosmopolitan in distribution (Fig. 99).
However, the more generalized tribe
(Dasylabrini) is confined to the Ethiopian,
southern Palearctic and Oriental regions
(with one species of Stenomiitilla on the
Solomons in the Australasian region). The
most widespread genera are Dasylabris
and Sterjotnittilla, both of which extend
over approximately the entire range of the
subfamily. The most derived genera are
probably Apteromiitilla and Brachymutilla
(both with apterous males), both of which
are confined to the southern tip of Africa.
This distribution is not inconsistent with
an origin for the group in north-eastern
Africa with the spread of two vigorous
genera into Eurasia, either across Arabia
when contact was established there or by
transport on the Indian plate, or both.
Diversification and dispersal also appar-
ently occurred toward the south and west
across Africa.

The tribe Sphaeropthalmini (Fig. 99)
occurs entirely in the New World and the
Australasian region, apart from the genus
Cystom nulla (two species) which now oc-
curs in the Mediterranean region and
Japan, a somewhat enigmatic situation.
Except for Cystomiitilla (which may ac-
tually be misclassified as a result of paral-
lel development of certain characters),
this distribution may have resulted from
a single introduction of an ancestral
sphaeropthalmine from Africa into South
America at about the time that these two
continents separated. After becoming es-

Phylogeny and Classification of the Aculeate Hymenoptera

635

3

J2

C/3

'-4-*

o

^
^

c
o

O

636

The University of Kansas Science Bulletin

tablished in South America, the Sphaerop-
thalmini apparently radiated extensively
and rapidly, establishing two basic lines at
this time (viz., Pseudomethocina and
Sphaeropthalmina). Since Australia and
South America were linked via Antarctica
until much more recently (50 m.y. ago),
most o£ the Australian fauna (predomi-
nantly Sphaeropthalmina, very few Pseu-
domethocina and MutiUini) could easily
have been derived from the South Amer-
ican. In fact the most highly derived
members of the Sphaeropthalmina seem
to be various members of the ''Ephuto-
tnorpha complex," especially those on New
Guinea. (This type of progression is an
important factor in establishing probable
routes of dispersal; Hennig, 1966b.)
Furthermore, there are various Australian
species which share some highly charac-
teristic derived features with a few South
American genera (e.g., the flangelike ex-
pansions of the antennal tubercles in the
females of Scaptodactyla), and the females
of the South American (southern Chile
and Argentina) genus Neomutilla are re-
markably like some "Ep/intotnorpha" spe-
cies in general facies. It may also be
significant that there are many highly
metallic species (often blue or green) in
the Australasian region (especially New
Guinea), whereas the only other metallic
Sphaeropthalminae occur in the southern
section of South America (e.g., females of
Neomutilla and Diniorphotnittilla suavis-
sima). (Marked similarities have been
found between the Australian and south-
ern South American representatives of
various insect groups — see Brundin, 1966;
O'Brien, 1971.) That the introduction of
mutillids into Australia by this route was
probably relatively recent, is indicated by
their absence in New Zealand, so that this
dispersal probably occurred less than 80
m.y. ago. Furthermore, there are very few
species (although at least two) on New
Caledonia, which probably split from Aus-

tralia soon after the New Zealand — Ant-
arctica separation. The North American
sphaeropthalmine fauna is not as rich as
that of South America and was apparently
derived from it by a few introductions,
probably when contact between the two
continents was established relatively re-
cently (10 m.y. ago).

The subfamily Myrmillinae (Fig. 98)
is distributed across the Ethiopian, most
of the Palearctic and the Oriental regions.
The richest representation is in Africa with
fewer genera in the peripheral areas and
only two species in Madagascar. (These
last are species of Pygomilla, and not
Ctenotilla as was indicated by Krombein,
1972.) In some respects the Indian fauna
is more similar to that of Africa than that
of Europe. The genus Sqiiatnulotilla, for
example, is present in Africa, India and
farther east in the Oriental region. This
may merely reflect ecological similarities,
but perhaps indicates spread of this genus,
at least, on to the Indian plate from Africa
and subsequent dispersal from India on
to Asia. The most highly derived species
of Sqitamiilotilla also tend to be those
which occur the farthest east (e.g., S.
byblis, with extreme development of the
ventral lamella on the mesopleural ridge
in the female, in the Philippines). Again,
it appears that this subfamily originated
in north-eastern Africa with subsequent
dispersal, probably on to the Indian plate
as well as more direct movement into
Eurasia across the Arabian region.

The subfamily Mutillinae is worldwide
in distribution, like the Sphaeropthalmi-
nae. Unlike the Sphaeropthalminae, how-
ever, this subfamily is richest in the Old
World, except for the Australian region.
The two tribes of Mutillinae also have
essentially complementary distributions
(Fig. 100). The Mutillini is entirely an
Old World group except for one genus
(Titnidla), which is cosmopolitan. This
genus shows signs of recent rapid specia-

Phylogeny and Classification of the Aculeate Hymenoptera

637

638

The University of Kansas Science Bulletin

tion and so may represent a more recent
introduction into the New World, perhaps
via the Beringian connection. [A problem
is that the species of Timiilla with the
most highly complex male genitalia (in
terms of asymmetry and development of
the endophallus) are African and Indian,
but various species from eastern Asia and
most New World species have the geni-
talia more nearly symmetrical (Peterson,
//; lift.). The latter would thus intuitively
seem to be the less highly derived forms,
which would suggest dispersal from the
New World through Asia to Africa. Ad-
ditional work is definitely necessary to
clarify the complexities of this genus and
its relatives before these problems can be
solved.] The Mutillini is most richly rep-
resented in Africa, and only two species
{Odontomntilla aiistralica and Timulla
cookj) extend into northern Australia.
The spread into the Australasian region
is thus obviously very recent and has prob-
ably occurred since the Australian and
Asian plates came into contact. There are
in addition two genera of Mutillini in the
New World, ChaetotiUa from Argentina
and Physetopoda from Haiti, each known
from only one specimen and not seen by
me (I could not find the types in the Paris
Museum), so that comment on their dis-
tribution and origin is impossible at this
time.

The tribe Ephutini (Fig. 100) is en-
tirely New World in distribution and has
its greatest diversity in northern South
America. It seems likely that this group
also arose from a single introduction of
the ancestral form from Africa to South
America at about the time that these two
continents separated. There was later dis-
persal from South America to North
America, probably when these continents
came into contact, but apparently no dis-
persal to Australia. This may have been
because the ephutines seem generally to

be adapted to more tropical conditions
than are many sphaeropthalmines.

It thus appears (Fig. 101) that the
Mutillidae probably arose in Laurasia and
diversified on the northern and eastern
part of the African plate less than 180
million years ago, from where dispersal
occurred on numerous occasions to India
and Eurasia. Two introductions from
Africa into South America probably gave
rise to almost the entire New World fauna
and also to most of the Australasian fauna
by a few introductions across Antarctica
from South America less than 80 million

vears ago.

Since the cladogram can be applied in
this way to explain the distribution of
Mutillidae in accordance with the se-
quences of events involved in the breakup
of the continents (derived from other evi-
dence), it appears to be supported (or at
least not contradicted) by the distribu-
tional data. The likelihood that the
cladogram reflects the true phylogeny of
the Mutillidae is thus enhanced.

GENERAL CONCLUSIONS

The entire investigation of the Acu-
leata, including the section on the Mutil-
lidae, has led to the proposed classification
presented in Tables VI and VII. Since it
has not been possible to examine each
taxon with equal thoroughness, the lower
levels remain to be clarified by subsequent
workers in the majority of instances.
Furthermore, it is quite possible that the
conclusions drawn at the higher levels
may in some cases prove to be unwar-
ranted when more complete representa-
tion is obtained for various groups. Cer-
tainly, some of the relationships shown
on the cladogram (such as the proximity
of Scoliidae to Vespidae and their remote-
ness from Tiphiidae) are unexpected.
These definitely warrant more intensive
examination. In fact, even if the only re-
sult of this investigation is the stimulation

Phylogeny and Classification of the Aculeate Hymenoptera

639

.2

sc

■ - c

u >

3 r3
^ .15

'J 3

O

640

The University of Kansas Science Bulletin

Table VI. Proposed classification of tiie Hymenoptera Aculeata.

Superfamily

Family

Subfamily"

BETHYLOIDEA

Plumariidae

Bethylidae

Scolebythidae

Cleptidac

Chrysididae

Loboscelidiidae

Dryinidae

Sclerogibbidae

Embolemidae

SPHECOIDEA

(Sphcciformes)

(Apiformcs)

Ampulicidae

Sphecidae

Larridae

Mellinidae

Pemphredonidae

Astatidae

Philanthidae

Nyssonidae

Colletidae

Halictidae

Oxaeidae

Andrenidae

Mclittidac

Fidcliidae

Megachilidae

Anthophoridae

Apidae

VESPOIDEA
(Vespiformes)

(Formiciformes)

Tiphiidae

Sapygidae

MutiUidae"

Sierolomorphidae

Rhopalosomatidae

Pompilidae

Bradynobaenidae

Scoliidae

Masaridae

Eumenidae

Vespidae

Formicidae

Anthoboscinae

Thynninae

Myzininae

Methochinae

Tiphiinae

Brachycistidinae

Fedtschenkiinae

Sapyginae

Typhoctinae"
Chyphotinae
Apterogyninae
Bradynobaeninae

° Subfamilies indicated only for taxa previously considered to comprise the tiphioid-mutilloid group, and

the Sapygidae.

'' For details of subfamily divisions sec Table VII.

'' Comprises two tribes, Eotillini and Typhoctini.

Phylogeny and Classification of the Aculeate Hymenoptera

641

Tablk VII. Proposed classification of the Mutillidae.

Subfamily

Tribe

Subtribe

Myrmosinac

Pscudophotopsidinae

Ticoplinae

Rhopalomutillinae

Sphacropthalminae

Myrmillinae
Mutillinae

Dasylabrini
Sphaeropthalmini

Mutillini
Ephutini

Pseudomethocina
Sphaeropthaimina

Mutillina
Smicromyrmina

of critical studies which may refute the
conclusions presented here, then it will
have accomplished its purpose.

Those conclusions which are most at
variance with the current ideas on the
classification of the Hymenoptera Acu-
leata, may be summarized as follows:

a) The Aculeata should be considered
as comprising only three superfamilies
(Bethyloidea, Vespoidea, Sphecoidea) as
opposed to the previously commonly ac-
cepted seven.

b) The Scolebythidae and Plumariidae
should be included with seven other fam-
iles in the Bethyloidea.

c) The Vespoidea should be divided
into two informal groups, the Vespiformes
(with 11 families) and the Formiciformes
(1 family).

d) The Sphecoidea should be divided
into two informal groups, the Spheci-
formes (with 8 families) and the Api-
formes (9 families).

e) The Myrmosinae and Bradynobae-
ninae should be removed from the Tiphi-
idae.

f) The "Eotillinae," "Typhoctinae"
and "Apterogyninae" ("Chyphotini" and
"Apterogynini") should be removed from
the Mutillidae.

g) The family Bradynobaenidae should
be recognized as comprising the subfam-
ilies Typhoctinae, Chyphotinae, Aptero-
gyninae and Bradynobaeninae.

h) The subfamily Typhoctinae should

be recognized as comprising two tribes,
the Typhoctini and Eotillini.

i) The family Mutillidae should be
divided into seven subfamilies, the Myr-
mosinae, Pscudophotopsidinae, Ticoplinae,
Rhopalomutillinae, Myrmillinae, Mutilli-
nae and Sphaeropthalminae.

i) The subfamily Mutillinae should be
recognized as comprising two tribes, the
Mutillini and Ephutini.

k) The tribe Mutillini should be recog-
nized to comprise two subtribes, the Mu-
tillina and Smicromyrmina.

1) The subfamily Sphaeropthalminae
should be divided into two tribes, the
Dasylabrini and Sphaeropthalmini.

m) The tribe Sphaeropthalmini should
be recognized as comprising two subtribes,
the Pseudomethocina and Sphaeropthai-
mina.

LITERATURE CITED

(For conyenience, a few recent titles are included
eyen though it has been impractical to incorporate
their content fully into the body of the text.)

Allen, H. W. 1972. A monographic study of the
subfamily Tiphiinae (Hymenoptera: Tiphi-
idae) of South America. Smithson. Contrib.
Zool. 113:i-iv, 1-76.

Allen, H. W. and H. A. Jaynes. 1930. Contribu-
tion to the taxonomy of Asiatic wasps of the
genus Tiphia (Scoliidae). Proc. United States
Nat. Mas. 76(17) :1-105, pi. 1-4.

Allen, H. W. and K. V. Krombein. 1964. A re-
vision of Neotiphia Malloch and Krombeinia
Pate (Hymenoptera: Tiphiidae). Trans.
American Ent. Soc. 89:211-275, pi. 19-26.

Andre, E. 1899-1903. Les MutiUides. Spec. Hym.
d'Europe & d'Algerie 8:1-479, pi. 1-15.

642

The University of Kansas Science Bulletin

. 1902. Hymenoptera. Fam. Mutillidae.

Genera Insectorum 11:1-77, pi. 1-3.

ASHLOCK, P. D. 1971. Monophylv and associated
terms. Syst. Zool. 20:63-69.'

. 1972. Monophylv again. Svst. Zool. 21:

430-438.

AsHLOcK, P. D. AND J. D. Lattix. 1963. Stridula-
tory mechanisms in the Lygaeidae, with a
new American genus of Orsillinae (Hcmip-
tera: Heteroptera). Ann. Ent. Soc. America
56:6^3-703.

AsHMEAD, W. H. 1896a. Descriptions of new para-
sitic Hymenoptera. Trans. American Ent.
Soc. 23:179-234.

. 1896b. The phyiogeny oi the Hymenop-
tera. Proc. Ent. Soc. Washington 3:323-336.

. 1900, 1903-1904. Classification of the fos-

sorial, predaceous and parasitic wasps, or the
superfamily Vespoidea [part]. Canadian Ent.
32:145-149; 35:3-8, 39-44, 95-107, 155-158,
199-205, 303-310, 323-332; 36:5-9.

AxELROD, D. 1. AND P. H. Raven. 1972. Evolution-
ary biogeography viewed from plate tectonic
theory. In Challenging Biological Problems
(Behnke, J. H., ed.),^ A.I.B.S. 25th Anni-
versary Volume. Oxford University Press, New
York. pp. 218-236.

Bertkau, P. 1884. Die Begattung von Mutilla
epinppitim. Biol. Zentralblatt 3:722-724.

Betrem, ]. G. 1928. Monographic dcr Indo-Aus-
tralischen Scoliiden (Hym. Acul.) mit zoo-
geographischen Betrachtungen. Treubia 9
'(Suppl.):l-388, pi. 1-5.

BiscHOFF, H. 1915. Bernsteinhymenopteren. Schrift.
Phys.-Oekon. Ges., Konigsberg 56:139-144.

. 1920-1921. Monographic dcr Mutiliiden

Afrikas. Arch. Naturg. (A)86:l-830, pi. 1-7.

. 1927. Biologic dcr Hymenoptcrcn. Biolo-

gische Studienbiicher, V. Springer, Berlin, pp.
i-viii, l-5'*«.

Blake, C. A. 1871. Synopsis of the Mutillidae of
North America. Trans. American Ent. Soc.
3:217-265.

. 1886. Monograph of the Mutillidae of

North America. Trans. American Ent. Soc.
13:179-286.

BoRNER, C. 1919. Stammcsgcschichte dcr Hautfliig-
ler. Biol. Zentralblatt 39:145-186.

BoHART, R. M. AND A. S. Menke. 1963. A re-
classification of the Sphecinac with a revision
of the nearctic species of the tribes Sccliphro-
nini and Sphecini (Hymenoptera, Sphccidae).
Univ. California Pubis. Ent. 30(2) :i-v, 91-181.

BoHART, R. M. AND R. O. Schuster. 1972. A host
record for Fedtschenl^ia (Hymenoptera: Sapy-
gidae). Pan-Pacific Ent. 48:149.

Bradley, J. C. 1950. The most primitive Scoliidae.
Eos, Tomo extraord.:427-437.

. 1955. Notes on the synonymy, distribution

and affinities of the subfamily Fedtschenkiinae
of the Sapvgidae (Hvmenoptera). Ent. News
66:230-233.

. 1958. The phyiogeny of the Hymenoptera.

Proc. X. Int. Congr. Ent., Montreal (1956)
1:265-269.

. 1972. Notes on the distribution of the

genus Plumaritis (Hvmenoptera, Plumariidae).
EnL News 83:135-139.

Bradley, J. C. and J. Bequaert. 1923. Studies in
African Mutillidae. Rev. Zool. Africainc 11:
211-258.

AND . 1928. A synopsis of the Mutil-
lidae of the Belgian Congo. Bull. American
Mus. Nat. HisL 58:63-122.

Bridwell, [J. C] 1917. [Exhibit of] Rhopalotnii-
tiUa claricornis. Proc. Hawaiian Ent. Soc.
3:260.

. 1920. Miscellaneous notes on Hymenoptera,

2nd paper, with descriptions of new species.
Proc. Hawaiian Ent. Soc. 4:386-403.

Brothers, D. J. 1970. Discovery of the mutillid
subfamily Typhoctinae in South America, with
description of a new species (Hymenoptera).
J. Kansas Ent. Soc. 43:302-308.

. 1972. Biology and immature stages of

Pseudomethoca /. frigida, with notes on other
species (Hvmenoptera: Mutillidae). Univ.
Kansas Sci. Bull. 50:1-38.

. 1974a. The new genus Typhoctoides and

a new species of Eotilla (Hymenoptera:
Bradynobaenidae; Tvphoctinae). J. Kansas
Ent Soc. 47:359-363.'

. 1974b. The first recent species of Proto-

mtitilla (Hvmenoptera: Mutillidae; Mvrmosi-
nae). Psyche 81:268-271.

— ■ . 1974c. The genera f)f Plumariidae, with

description of a new genus and species from
Argentina (H\ menoptera: Bethyloidea). J.
Ent. Soc. Southern Africa 37:351-356.

Brown, W. L., Jr. 1954. Remarks on the internal
phyiogeny and subfamily classification of the
family Formicidae. Insectes Sociaux 1:21-31. I|

. 1969. Ant, paleontology of. McGraw-Hill ''

Yearbook of Science and Technology for
1969. pp. 101-103.

Brown, W. L., Jr. and W. L. Nutting. 1950. Wing
venation and the phyiogeny of the Formicidae
(Hvmenoptera). Trans. American Ent. Soc.
75:'l 13-132, pi. 8-9.

Brown, W. L., Jr. and R. W. Taylor. 1970. Super-
family Formicoitlea. In The Insects of Aus-
tralia. Melbourne University Press, pp. 951-
959.

Browx, W. L., Jr. and E. O. Wilson. 1959. The
search for Notliomyrmccia. Western Aus-
tralian Nat. 7:25-30.

Brues, C. T. 1943. The American species of
Rhopalosoma. Ann. Ent. Soc. America 36:
31U-318, pi. 1. ,

Brundin, L. 1966. Transantarctic relationships and |l
their significance, as evidenced by chironomid
midges; with a monograph of the subfamilies
Podonominae and Aphroteniinae and the aus-
tral Heptagyiae. K. Svenska Vetensakad.
Handl. (4)11:1-472, pi. 1-30.

BuRDicK, D. J. AND M. S. Wasbauer. 1959. Biology
of Mcthoclm calijornica Westwood (Hyme-
noptera: Tiphiidae). Wasmann J. Biol. 17:
75-88.

Burmeister, H. C. C. 1874. Mutillae Argentinae.
Bol. Acad. Nac. Cienc. Exact. Univ. Cordoba
1:461-502, 1 pi.

BuzicKY, A. W. 1941. -A monograph of the genus {I
Chyphotes (Hymenoptera, Mutillidae, Apte-
rogvninae) of North .America. Ent. Ameri-
cana (NS)21:201-250. 1 1

Callan, E. McC. 1939. A note on the breeding of \
Prohethyhts callani Richards (Hymenopt.,
Bethvlidae), an embiopteran parasite. Proc.
R. Ent. Soc. London (B) 8:223-224.

Phylogeny and Classification of the Aculeate Hymenoptera

643

Camix, I. H. AND R. R. SoKAL. 1965. A method
for deducing branching sequences in phylog-
eny. Evolution 19:31 1-326.

Charxley, H. W., Jr. 1973. The value of the
propodeal orifice and the phallic capsule in
vespid taxonomy (Hymenoptera, Vespidae).
Bull. Buffalo Soc. Nat. Sci. 26:i-iv, 1-79.

Chex, C-W. 1957. A revision of the velvety ants
or Mutillidae of China (Hvmenoptera). Quart.
J. Taiwan Mus. 10:135-224, pi. 1-6.

Clausen, C. P. 1940. Entomophagous Insects. Mc-
Graw-Hill, New York. pp. Vx, 1-688.

Clausex, C. p. AND T. R. Gardner. 1932. Biology
of some Japanese and Chosenese grub parasites
(Scoliidae). United States Dept. Agric, Tech.
Bull. 308:1-26.

Clement, S. L. and E. E. Grissell. 1968. Observa-
tions of the nesting habits of Etiparagia scii-
tellaris Cresson (Hymenoptera: Masaridae).
Pan-Pacific Ent. 44:34-37.

Cracraft, J. 1973. Continental drift, paleoclimatol-
ogv, and the evolution and biogeographv of
birds. J. Zool., London 169:455-545.

Creighton, W. S. 1950. The ants of North .Amer-
ica. Bull. Mus. Comp. Zool., Harvard 104:
1-5S5. pi. 1-57.

Crevecoeur, A. 1930. Recherchcs biologiques sur
Smicromyrme (Mntilla) rnfipes F. (Hym.
Mutillidae). Bull. Ann. Soc. Ent. Bclgiquc
70:271-284.

Daly, H. V., Jr. 1955. A comparative study of the
sting of aculeate Hymenoptera. Unpublished
M.A. thesis, Universitv of Kansas, pp. i-iv,
1-67, pi. 1-7.

. 1964. Skeleto-muscular morphogenesis of

the thorax and wings of the honey bee ylpis
niellifera (Hymenoptera: Apidae). Uni\-.
California Pubis Ent 39:1-77.

Debolt, J. W. 1973. Morphology and histology of
the felt line and felt line organ of the mutillid
wasp genera Sphaeropthabna, Dasymiitilla,
Psettdomethoca, and Chypliotes. Ann. Ent.
Soc. America 66:100-108.

DiETZ, R. S. and J. C. HoLDEX. 1970. The breakup
of Pangaca. Sci. American 223(4) :30-41.

Duncan, C. D. 1939. A contribution to the biology
of North American vespine wasps. Stanford
Univ. Publ. (Univ. Ser.) Biol. Sci. 8:1-272.

EicKwoRT, G. C. 1969. A comparative morpho-
logical study and generic revision of the
augochlorine bees (Hvmenoptera: Halictidae).
Univ. Kansas Sci. Bull. 48:325-524.

Evans, H. E. 1953. Comparative ethology and the
systematics of spider wasps. Svst. Zool. 2:
155-172.

. 1958. The evolution of social life in wasps.

Proc. X. Int. Congr. Ent., Montreal (1956)
2:449-457.

. 1959a. Studies on the larvae of digger

wasps (Hymenoptera, Sphecidae). Part V:
Conclusion. Trans. American Ent. Soc. 85:
137-191, pi. 18-24.

. 1959b. The larvae of Pompilidae (Hyme-
noptera). Ann. Ent. Soc. America 52:430-
444.

. 1961. A preliminary review of the nearctic

species of Sierolomorpha (Hvmenoptera).
Breviora 140:1-12.

. 1963. A new family of wasps. Psyche

70:7-16.

. 1964a. A synopsis of the American Bethyli-

dae (Hymenoptera, Aculeata). Bull. Mus.
Comp. Zool., Harvard 132:1-222.

. 1964b. The classification and evolution of

digger wasps as suggested by larval characters
(Hvmenoptera: Sphecoidea). Em. News 75:
225-237.

. 1965. A description of the larva of

Mcthoclia stygia (Say), with notes on other
Tiphiidae (Hymenoptera). Proc. Ent. Soc.
Washington 67:88-95.

. 1966a. The Comparative Ethology and

Evolution of the Sand Wasps. Harvard Uni-
versity Press, Cambridge, Mass. pp. i-xviii,
1-526.

. 1966b. The behavior patterns of solitary

wasps. Ann. Rev. Ent. 11:123-154.

. 1967. Discovery of the female Plntnaritis

(Hvmenoptera, Plumariidae). Psyche 73:229-
237.

. 1969a. Phoretic copulation in Hymenop-
tera. Ent. News 80:113-124.

. 1969b. Three new Cretaceous aculeate

wasps (Hymenoptera). Psyche 76:251-261.

. 1972. The tribe Ctcnoceratini in Australia

(Hvmenoptera: Pompilidae). J. Australian
Ent. Soc. 11:244-252.

Evans, H. E. and C. S. Lin. 1956. Studies on the
larvae of digger wasps (Hymenoptera, Sphe-
cidae). Part I: Sphecinae. Trans. American
EnL Soc. 81:131-153, pi. 1-8.

Evans, H. E. and M. J. West Eberhard. 1970. The
Wasps. University of Michigan Press, Ann
Arbor, pp. i-vi, 1-265.

ExLEY, E. M. 1968. Revision of the genus Brachy-
besma Michener (Apoidea: Colletidae). Aus-
tralian J. Zool. 16:167-201.

Parish, D. J. 1972. The evolutionary implications
of qualitative variation in the grooming be-
haviour of the Hvmenoptera (Insecta). Anim.
Behav. 20:662-676.

Farris, J. S. 1967. The meaning of relationship
and taxonomic procedure. Syst. Zool. 16:
44-51.

. 1969. A successive approximations approach

to character weighting. Svst. Zool. 18:374-
385.

. 1970. Methods for computing Wagner

trees. Syst. Zool. 19:83-92.

Ferguson, W. E. 1962. Biological characteristics of
the mutillid subgenus Photopsis Blake and
their systematic values. Univ. California
Pubis EnL 27:1-91.

Fooden, J. 1972. Breakup of Pangaea and isolation
of relict mammals in Australia, South Amer-
ica, and Madagascar. Science 175:894-898.

Fox, W. J. 1894. A proposed classification of the
fossorial Hymenoptera of North America.
Proc. Acad. Nat. Sci. Philadelphia 1894:
292-307.

. 1899. The North American Mutillidae.

Trans. American Ent. Soc. 25:219-292.

. 1900. Arrangement of the extra- American

species of Mntilla. (Described by Cresson and
Blake) according to my classification of the
genus. EnL News 11:400-401.

Ghesquiere, J. 1951. Synonymies nouvelles pour
la famille des Obenbergerellidae (Hym.
Scolioidea Tiphiidae Mvrmosinae). Bull. Ann.
Soc. Em. Belgique 87.-238-241.

644

The University of Kansas Science Bulletin

GiNER Mari, J. 1944. Himenopteros de Espana.
Fams. Apterogynidae y Mutillidae. Instituto
Espanol de Entomologia, Madrid, pp. 1-124.

Given, B. B. 1954. Evolutionary trends in the
Thynninae (Hymenoptera: Tiphiidae) with
special reference to feeding habits of Australian
species. Trans. R. Ent. Soc. London 105:1-10,
pi. 1-7.

. 1958. Notes on Australian Thynninae. II.

The genera Dimorphothynntis, Rliagigaster
and Eirone. Proc. Linn. Soc. New South
Wales 83:309-326.

. 1959. Notes on Australian Thynninae. IV.

The morphology of Thynnoides itifithorax
Turner with notes on the prepupal larva and
the cocoon. Proc. Linn. Soc. New South
Wales 84:391-399.

GoTWALD, W. H., Jr. 1969. Comparative morpho-
logical studies of the ants, with particular
reference to the mouthparts (Hymenoptera:
Formicidae). Cornell Univ. Agric. Expt. Stn.,
Mem. 408:1-150.

Grandi, G. 1961. Studi di un entomologo sugli
imenotteri superiori. Boll. Inst. Ent. Univ.
Bologna 25:i-xvi, 1-661.

GuiGLiA, D. 1955. Appunti intorno a! gen. Fedt-
schenl{ia Saussure (Hymenoptera: Sapygidae).
Doriana, Genoa 2(55) :l-4.

. 1963. Una nuova specie di Tedtschenkia

della Palestina. Doriana, Genoa 3(134):l-6.

. 1968. Sul genere A-/v=;«<?//a Guiglia (1959).

Descrizione della femmina (Hymenoptera:
Tiphiidae). Mitt. Schweizerischen Ent. Ges.
41:171-174.

. 1969. The history of the peculiar genus

Fedtschenk'ia Saussure (1880) (Hymenop-
tera: Fedtschenkiidae). Israel J. Ent. 4:
339-342.

. 1972. De la distribution geographique du

genre Fedtschenk'ia (Saussure, 1880) (Hy-
menoptera: Fedtschenkiidae). Proc. XIII.
Congr. Ent., Moscow (1968) 1:138-140.

GuRNEY, A. B. 1953. Notes on the biology and
immature stages of a cricket parasite of the
genus Rhopalosoma. Proc. United States Nat.
Mus. 103:19-34, pi. 1.

Hamann, H. H. F. 1960. Stenomcthoca suhgcn.
nov., nigra spec. nov. (Hymenopt., Methoci-
dae) von West-Borneo. Idea 13:35-41.

Hamilton, K. G. A. 1971. The insect wing, part I.
Origin and development of wings from notal
lobes. J. Kansas Ent. Soc. 44:421-433.

. 1972a. The insect wing, part II. Vein

homology and the archetypal insect wing.
J. Kansas EnL Soc. 45:54-58.

. 1972b. The insect wing, part III. Venation

of the orders. J. Kansas Ent. Soc. 45:145-162.

Handlirsch, a. 1925. Uberordnung und Ordnung:
Hymenoptera L. (Hautfliigler). ]n Handbuch
der Entomologie (Schroder, C). Fischer,
Jena. Vol. 3, pp. 712-825.

Haskins, C. p. and E. V. Ensman. 1938. Studies of
certain sociological and physiological features
in the Formicidae. Ann. New York Acad.
Sci. 37:97-162.

Haskins, C. P. and E. F. Haskins. 1950. Notes on
the biology and social behavior of the archaic
ponerine ants of the genera Myrmecia and
Promyrmecia. Ann. Ent. Soc. America 43:
461-491.

Haupt, H. 1938. Zur Kenntnis der Dryinidae I.
(Hymenoptera-Sphccoidea). Zeitschr. Naturw.
92:13-35.

Heaversedge, R. C. 1968. Variation in the size of
insect parasites of puparia of Glossina spp.
Bull. Ent. Res. 58:153-158.

. 1969a. The sex ratio of insect parasites of

Glossina morsilans nricntalis Vanderplank
(Diptera). Arnoldia (Rhodesia) 4(14) :I-4.

. 1969b. Brief note on the reproductive mor-
phology of Mtitilla glossinae Turner (Hyme-
noptera) and the development of its imma-
ture stages. J. Ent. Soc. Southern Africa
32:485-488.

. 1970. Developmental periods of insect

parasites of Glossina morsitans orientalis
Vanderplank (Diptera: Muscidae). J. Ent.
Soc. Southern Africa 33:351-354.

Heirtzler, J. R., J. V. Veevers, H. M. Bolli, A.
N. Carter, P. J. Cook, V. A. Krasheninni-
Kov, B. K. McKnight, F. Proto-Decima, G.
W. Renz, p. T. Robinson, K. Rocker Jr.
AND P. A. Thayer. 1973. Age of the floor
of the eastern Indian Ocean. Science 180:
952-954.

Hennig, W. 1966a. Phylogenetic Systematics (trans-
lated by D. D. Davis and R. Zangerl). Uni-
versity of Illinois Press, Urbana. pp. i-vi,
1-263.

. 1966b. The Diptera fauna of New Zealand

as a problem in systematics and zoogeography
(translated by P. Wygodzinsky). Pacific Ins.
Monog. 9:1-81.

. 1969. Die Stammesgeschichte der Insekten.

Scnckenbergischcn Naturforschenden Gesell-
schaft, Frankfurt am Main. pp. 1-436.

Hermann, H. R., Jr. and M. S. Blum. 1966. The
morphology and histology of the hymenop-
terous poison apparatus. I Paraponera clarata
(Formicidae). Ann. Ent. Soc. America 59:
397-409.

Hinton, H. E., D. F. Gibbs and R. Silberglied.
1969. Stridulatory files as diffraction gratings
in mutillid wasps. J. Insect Phvsiol. 15:549-
552, fig. 1-10.

Hofeer, E. 1886. Zur Biologic der Mtitilla ciiro-
paea L. Zool. Jahrb. 1:677-686.

Houston, T. F. 1970. Discovery of an apparent
male soldier caste in a nest of a halictine bee
(Hymenoptera: Halictidae), with notes on
the nest. Australian J. Zool. 18:345-351.

Hull, D. L. 1970. Contemporary systematic philos-
ophies. Ann. Rev. Ecol. Syst. 1:19-54.

International Code of Zoological Nomenclature.
1964. [Code] adopted by the XV. Interna-
tional Congress of Zoology, London, July
1958. International Trust for Zoological
Nomenclature, London, pp. i-xx, 1-176.

Invrea, F. 1957. Le Apterogyna del Sud Africa
(Hymenoptera-Apterogynidae). Ann. Mus.
Civ. Stor. Nat., Genova 69:257-332.

. 1964. Mutillidae — Myrmosidae. Fauna

d'ltalia 5:i-xii, 1-303.

luGA, V. 1968. La persistance des segments gnathaux
dans le tube labio-maxillaire des Hymenop-
teres Scoliides. Trav. Mus. d'Hist. Nat.
"Grigore Antipa" 9:471-493.

IwATA, K. 1942. Comparative studies on the habits
of solitary wasps. Tenthredo 4:1-146, pi. 1-5.

Phvlogeny and Classification of the Aculeate Hymenoptera

645

. 1950. Hypothetic habit-type trees of the

aculeate Hymenoptera. Kagawa Kcnritsu Noka
Daigaku. Tech. Bull. 1:75-81.

Janvier, H. 1933. Etude biologic]Uc de quelques
Hymenopteres du Chili. Ann. Sci. Nat., Zoo!.
(10)16:209-356.

}oRD.\N, R. 1935. Die Spinncnamcisc, Miitilla eiiro-
pac'ii, ein Bienenschiidling! Deutsche linker
47:421-427.

Kluge, a. G. ,\nd J. S. Farris. 1969. Quantitative
phvletics and the evolution of anurans. Svst.
Zool. 18:1-32.

KxERER, G. 1973. Periodizitiit und Strategic dcr
Schmarotzer einer socialen Schmalbiene,
Evylaeits malachtirns (K.) (Apoidea: Halicti-
dae). Zool. Anz., Leipzig 190:41-63.

Krombeiv, K. V. 1937. Studies in the Tiphiidae
(Hymenoptera Aculeata). I. A review of the
genera of Mvzininae. Ann. Ent. Soc. America
30:26-30.

■ . 1940. Studies in the Tiphiidae (Hymenop-
tera Aculeata). IV. A revision of the Myr-
mosinae of the New World with a discussion
of the Old World species. Trans. American
Ent. Soc. 65:415-465. pi. 24.

. 1951. Superfamily Scolioidea [except

Sapygidae]. \n Hymenoptera of America
North of Mexico — Svnoptic Catalog (Muese-
beck, C. F. W., K. V. Krombein and H. K.
Townes, eds.). United States Dept. Agr.,
Agr. Monog. 2:735-776.

. 1956. Biological and taxonomic notes on

the wasps of Lost River State Park, West
Virginia, with additions to the faunal list.
Proc. Ent. Soc. Washington 58:153-161.

. 1957. A generic review of the Amiseginae,

a group of phasmatid egg parasites, and notes
on the Adelphinae (Hymenoptera, Bethyloi-
dea, Chrysididae). Trans. American Ent. Soc.
82:147-215, pi. 19-22.

. 1967. Superfamily Scolioidea. In Hyme-
noptera of America North of Mexico — Synoptic
Catalog, Second Supplement (Krombein, K.
V. and B. D. Burks, eds.). United States
Dept. Agr., Agr. Monog. 2(Suppl. 2):322-
j4:).

. 1968. Studies in the Tiphiidae, X. Hylo-

mesa, a new genus of myzinine wasp para-
sitic on larvae of longicorn beetles (Hyme-
noptera). Proc. United States Nat. Mus. 124
(3644):l-22, pi. 1.

. 1971. A monograph of the Mutillidae of

New Guinea, Bismarck Archipelago and Solo-
mon Islands. Part I: Mutillinae (Hymenop-
tera Aculeata). Entomological Essays to Com-
memorate the Retirement of Professor K.
Yasumatsu. Tokyo, pp. 25-60.

. 1972. Monograph of the Madagascan Mu-
tillidae (Hymenoptera). Part 1: Myrmillini,
Mutillini and Smicromyrmini. Ann. Mus. R.
Afrique Centrale, Tervurcn (in-8°, Zool.)
199:i-x, 1-61.

Krombein, K. V. and R. M. Schuster. 1957. A
review of the Typhoctinae (Hymenoptera:
Mutillidae). Proc. Ent. Soc. Washington 59:
209-232.

Lamborn, W. .\. 1915. Secf>nd report on Glossina
investigations in Nyasaland. Bull. Ent. Res.
6:249-265.

. 1916. Third report on Glossina investiga-
tions in Nyasaland. Bull Ent. Res. 7:29-50.

Lanham, U. N. 1951. Review of the wing venation
of the higher Hymenoptera (suborder Clisto-
gastra), and speculations on the phylogeny of
the Hymenoptera. Ann. Ent. Soc. America
44:614-628.

Latreille, p. A. 1802. Famille Quinzieme. Mutil-
laires; miitilhviae. In Histoire naturelle, gen-
erale et particuliere des Crustaces et des In-
sectes. Paris. Vol. 3, pp. 347-352.

. 1809. Genera Crustaceorum et Insectorum,

. . . Vol. 4, pp. 117-124. [Not seen by me.]

. 1825. Huitieme ordre: Hymenopteres,

Hymenoptera. Seconde section: Les Porte-
aiguillon, Aculeata. Troisieme famille:
Heterogynes, Hetcrogyna. Seconde tribu:
Mutillaires, Miitillanae. In Families naturelles
de Regne Animal, . . . Paris, pp. 452-453.

. 1829. Le neuvieme ordre des Insectes, celui

des Hymenopteres. Des Heterogynes, des Mu-
tilles. In Le Regne Animal (Cuvier, G.). Vol.
5, pp. 314-316.

Le Quesne, W. J. 1969. A method of selection of
characters in numerical taxonomy. Syst. Zool.
18:201-205.

. 1972. Further studies based on the uniquely

derived character concept. Syst. Zool. 21:281-
288.

Leclercq, J. 1954. Monographic systematique, phy-
logenetique et zoogeographique des Hyme-
nopteres Crabroniens. Lejeunia Press, Liege,
pp. 1-371, maps 1-84.

Linnaeus, C. 1758. Mtitilla. In Systema naturae
. . . 10th edition. Holmiae. Vol. 1, pp.
582-583.

LiNSLEY, E. G. 1960. A fragmentary observation on
the mating behavior of Thntilla (Hymenop-
tera: Mutillidae). Pan-Pacific Ent. 36:36.

LiNSLEY, E. G., J. W. MacSwain and R. F. Smith.
1955. Observations on the mating habits of
Dasymntilla formicalia Rohwer (Hymenop-
tera: Mutillidae). Canadian Ent. 87:41 1-413.

LiNSLEY, E. G. AND C. D. MiCHENER. 1939. A

generic revision of the North American
Nomadidae (Hymenoptera). Trans. American
Ent. Soc. 65:265-305, pi. 15-18.

Maa, T. C. and C. M. Yoshimoto. 1961. Loboscelidi-
idae, a new familv of Hymenoptera. Pacific
Insects 3:523-548. '

Mackerras, I. M. 1970. Evolution and classifica-
tion of the insects. In The Insects of Australia.
Melbourne University Press, pp. 152-167.

Malyshev, S. I. 1968. Genesis of the Hymenoptera
and the Phases of their Evolution (Richards,
O. W. and B. Uvarov, eds.). Methuen, Lon-
don, pp. i-viii, 1-319.

Marechal, p. 1930. Sur trois Hymenopteres se
developpant dans un cocon en mosaique.
Mem. Soc. Ent. Belgique 23:1-23.

Markl, H. 1973. The evolution of stridulatory
communication in ants. Proc. VII. Int. Congr.
Int. Union Study Soc. Ins., London, pp.
258-265.

Matsuda, R. 1965. Morphology and evolution of
the insect head. Mem. American Ent. Inst.
4:i-viii, 1-334.

. 1970. Morphology and evolution of the

insect thorax. Mem. Ent. Soc. Canada 76:
1-431.

646

The University of Kansas Science Bulletin

Mavr, E. 1969. Principles of Systematic Zoology.
McGraw-Hill, New York etc. pp. i-xii, 1-428.

MicHENER, C. D. 1944. Comparative external mor-
phology, phylogeny and a classification of the
bees (Hymenoptera). Bull. American Mus.
Nat. Hist. 82:151-326.

. 1953. Comparative morphological and sys-
tematic studies of bee larvae with a key to
the families of hvmenopterous larvae. Univ.
Kansas Sci. Bull. 35:987-1102.

. 1965. A classification of the bees of the

Australian and South Pacific regions. Bull.
American Mus. Nat. Hist. 130:1-362, pi. 1-15.

. 1970. Diverse approaches to svstematics.

Evol. Biol. 4:1-38.

. 1974. The Social Behavior of the Bees.

Harvard University Press, Cambridge, Mass.
pp. i-xii, 1-404.

MicKEL, C. E. 1924. An analysis of a bimodal
variation in size of the parasite Dasymiitilla
biocidata Cresson (Hymen.: Mutillidae). Ent.
News 35:236-242, pi. 6.

. 1928. Biological and taxonomic investiga-
tions on the mutillid wasps. Bull. United
States Nat. Mus. 143:i-x, 1-351, pi. 1-5.

. 1933. The Mutillidae of Formosa. Ann.

Ent. Soc. America 26:381-423.

. 1935. The mutillid wasps of the islands of

the Pacific Ocean (Hymenoptera; Mutillidae).
Trans. R. Ent. Soc. London 83:177-312.

. 1938. Photopsoid mutillids collected by

Dr. K. A. Salman at Eagle Lake, California
(Hymenoptera). Pan-Pacific Ent. 14:178-185.

. 1967. A review of the mutillid genus

Chyphotes Blake (Hymenoptera; Mutillidae;
Apterogyninae). Trans. American Ent. Soc.
93:125-234, pi. 1-10.

. 1968. The female sex of the Mutillidae

subfamily Eotillinae (Hymenoptera). Psyche
74:234-236.

MicKEL, C. E. AXD K. V. Krombein. 1942.
Glyptometopa Ashmead and related genera
in the Brachycistidinae, with descriptions of
new genera and species (Hymenoptera, Tiphi-
idae). American Midi. Nat. 28:648-679.

Moss, W. W. AND J. A. Hendrickson, Jr. 1973.
Numerical taxonomy. Ann. Rev. Ent. 18:
227-258.

Nagy, C. G. 1967. Systematisches Studium der
Tiphiiden Rumiiniens (Hymenoptera, Tiphi-
idae). Reichenbachia 8:175-204.

. 1969a. A new taxon of the family Heter-

ogynidae Latreille (Hym., Aculeata). Ent.
Mitt. Staatsinst. Zool. Mus. Hamburg 64:7-11.

. 1969b. The taxon of the genus Pscudotiphia

Ashm. (Hymenoptera, Tiphiidae). Reichen-
bachia 12:141-142.

. 1969c. Sur la sous-famille Mesitinae Ber-

land (Hym., Bethylidae). Lucrarile Stat.
Cercet. Mar. 'Trof. loan Borcea" Agigea
3:275-300, pi. 1-19.

. 1970. Further investigations on the heter-

ogynid wasps. Ent. Mitt. Zool. Mus. Ham-
burg 4:83-86.

. 1971. First record of the Old World species

of Sieiolomorplia Ashm. (Hym., Heterogy-
noidea). Reichenbachia 13:247-249.

Nelson, G. J. 1971. Paraphyly and polyphyly: re-
definitions. Syst. Zool. 20:471-472.

Nielsen-, E. T. 1932. Sur les habitudes des Hyme-

nopteres aculeates solitaires I. (Bethylidae,

Scoliidae, Cleptidae, Psammocharidae). Ent.

Meddel. 18:1-57.
Nonveillen, G. 1973. Rccherchcs sur les Mutillides

de I'Afrique (Mutillidae, Hymenoptera). 111.

Remarques concernant le genre Nanomutilla

Andre 1899 (9). Ann. Fac. Sci. Cameroun

15-16:63-73.
O'Brien, C. W. 1971. The biogeography of Chile

through entomofaunal regions. Ent. News

82:197-207.
Oeser, R. 1961. Vergleichend-morphologische Un-

tersuchungen iJber den Ovipositor der Hyme-

nopteren. Mitt. Zool. Mus. Berlin 37:3-119.
Ohno, S. 1973. Ancient linkage groups and frozen

accidents. Namre 244:259-262.
Olsoufieff, G. 1938. Revision systematique des

Mutilles de Madagascar. Bull. Acad. Malgache

(NS)20:171-217,V- 1-2-
Pagden, H. T. 1934. Biological notes on some

Malayan aculeate Hymenoptera I. J. Fed.

Malay States Mus. 17:458-466.
. 1938. On a new species of Rliopahmtitilla

(Hym. Mutillidae) from Java. J. Fed. Malay

States Mus. 18:213-217.
Pate, V. S. L. 1947a. A conspectus of the Tiphiidae,

with particular reference to the nearctic forms

(Hymenoptera Aculeata). J. New York Ent.

Soc. 55:115-146.
. 1947b. A minute on Paramyrmosa Saussure

(Hymenoptera, Tiphiidae). Canadian Ent.

78:196-197.
. 1947c. Neotropical Sapygidae, with a con-
spectus of the family. Acta Zool. Lilloana

4:393-426.
Perez D'Angello, V. 1968. Nueva especie de

Bradvnobaeninae (Hymenoptera, Tiphiidae).

Rev. Chilena EnL 6:127-129.
Perkins, R. C. L. 1908. Some remarkable Australian

Hymenoptera. Proc. Hawaiian Ent. Soc. 2:

27-35.
Ponomarenko, N. G. 1972. Some peculiarities of

development of Dryinidae. Proc. XIH. Int.

Congr. Ent., Moscow (1968) 1:281-282.
Pouvreau, a. 1973. Les ennemis des bourdons.

I — Etude d'une zoocenose: le nid de bourdons.

Apidologie 4:103-148.
Rasnitsyn, a. p. 1969. Proiskhozhdenie i Evo-

lyutsiya Nizshikh Pereponchatokryiykh. Akad.

Nauk SSSR, Trudy PaleonL InsL 123:1-196.
Raven, P. H. and D. 1. Axelrod. 1972. Plate

tectonics and Australasian paleobiogeography.

Science 176:1379-1386.
Reid, J. A. 1939. On the relationship of the hyme-

nopterous genus Olixon and its allies, to the

Pompilidae (Hym.). Proc. R. Ent. Soc. Lon-
don (B)8:95-102.
. 1941. The thorax of the wingless and

short-winged Hymenoptera. Trans. R. Ent.

Soc. London 91:367-446.
. 1942. On the classification of the larvae

of the Vespidae (Hymenoptera). Trans. R.

Ent. Soc. London 92:285-331.
Richards, O. W. 1939. The Bethylidae subfamily

Sclerogibbinae (Hymenoptera). Proc. R. Ent.

Soc. London (B)8:'21 1-223.
. 1956a. An interpretation of the ventral

region of the hymenopterous thorax. Proc.

R. EnL Soc. London (A)31:99-104.

Phylogeny and Classification of the Aculeate Hymenoptera

647

■ . 1956b. Hymenoptera. Introduction and

kev to families. R. Hnt. Soc. London,
Hndbks. Ident. British Ins. 6(l):l-94.

■ . 1962. A Rcvisional Study of the Masarid

Wasps (Hymenoptera, Vespoidea). Trustees
of the British Museum, London, pp. i-vii,
1-294.

. 1971. The thoracic spiracles and some

associated structures in the Hymenoptera and
their significance in classification, especially
of the Aculeata. Entomological Essays to
Commemorate the Retirement of Professor K.
Yasumatsu. Tokyo, pp. 1-13.

RiDSDiLL Smith, T. J. 1970a. The biology of
Hcmithyniiiis hyalinatiis (Hymenoptera: Ti-
phiidae), a parasite on scarabaeid larvae. J.
Australian Ent. Soc. 9:183-195.

. 1970b. The behaviour of Hemithynniis

hyalinatiis (Hymenoptera: Tiphiidae), with
notes on some other Thvnninae. J. Australian
Ent. Soc. 9:196-208.

RiKK, E. F. 1970. Hymenoptera [except Apoidea
and Formicoidea] . In The Insects of Aus-
tralia. Melbourne University Press, pp. 867-
943.

Roberts, R. B. 1973. Nest architecture and imma-
ture stages of the bee Oxaea flarescens and
the status of Oxaeidae (Hymenoptera). J.
Kansas Ent. Soc. 46:437-446.

Robertson, P. L. 1968. A morphological and func-
tional study of the venom apparatus in repre-
sentatives of some major groups of Hymenop-
tera. Australian J. Zool. 16:133-166.

Ross, H. H. 1936. The ancestry and wing venation
of the Hymenoptera. Ann. Ent. Soc. America
29:99-lli.

. 1965. A Textbook of Entomology, 3rd edi-
tion. Wiley, New York etc. pp. i-x, 1-539.

RozEN, J. G., Jr. 1965. The biology and immature
stages of Melitlurga davicornis (Latreille)
and of Sphecodes alhilabris (Kirby) and the
recognition of the Oxaeidae at the family
level (Hymenoptera, Apoidea). American
Mus. Novit. 2224:1-18.

Salter. K. E. W. 1957. Studies on Australian
Thynnidae Shuckard, 1841. (Hymenoptera)
111. An introduction to the comparative mor-
phology of the male. Proc. Linn. Soc. New
South Wales 82:328-351.

Saussure, H. de. 1880. Skolii (Scolidae). hi
Puteshestvie v Turkestan (Fedchenko, A. P.);
No. 16, Vol. 2, Part 5, Section 7. Izvestiya
Imperatorskago Obshchestva Lyubitelei Es-
testvoznaniya, Antropologii i Etnografii 26
(3):l-46, pi. 1-2.

. 1892. Famille des Heterogynes; sous-famille

des Mutillines. In Histoire naturelle des
Hymenopteres. Grandidier, Hist. Phys., Nat.
Polit. Madagascar 20(1) :251-302, pi. 6, 7,
18, 19, 22.

Schuster, R. M. 1947. A revision of the sphaeroph-
thalmine Mutillidae of America north of
Mexico. Ann. Ent. Soc. America 39:692-703.

•. 1949. Contributions toward a monograph

of the Mutillidae of the neotropical region,
III. A key to the subfamilies represented and
descriptions of several new genera (Hymenop-
tera). Ent. Americana (NS)29:59-140.

. 1950. Notes on the Pseudophotopsidinae

(Mutillidae) with description of the female
sex. J. New York Ent. Soc. 58:192-198.

. 1958. A revision of the sphaeropthalmine

Mutillidae of America north of Mexico. II.
Ent. Americana (NS)37:1-130.

Sevrk;, A. 1936. Un mutillide parasite d'un
Lepidoptere: Stenomtitilla jreyi. Livre Jubi-
laire de M. E.-L. Houvier. Paris, pp. 313-316.

Sharov, a. G. 1957. Perbaya nakhodka melovogo
zhalonosnogo pereponchatokrylogo (Hymenop-
tera, Aculeata). Doklady Akad. Nauk SSSR
112:943-944.

Sheldon, J. K. 1970. Sexual dimorphism in the
head structure of Mutillidae Hymenoptera:
a possible behavioral explanation. Ent. News
81:57-61.

SiCHEL, J. AND O. Radoszkovsky. 1869. Essai d'une
monographie des mutilles de I'ancien conti-
nent. Horae Soc. Ent. Rossicae 6:139-309.

Skorikov, a. S. 1927. O podscm. Myrmillini (Myr-
mosidae, Hymenoptera) v Palearktike. Akad.
Nauk SSSR', Ezhegodnik Zool. Muz. 28:33-47.

. 1935. K faune inutillid srednei azii. Akad.

Nauk SSSR, Trudy Tadzhikskoibazy (Zool. i
Parazit.) 5:257-349.

Smith, E. L. 1969. Evolutionary morphology of the
external insect genitalia. I. Origins and rela-
tionships to other appendages. Ann. Ent. Soc.
America 62:1051-1079.

. 1970a. Evolutionary morphology of the

external insect genitalia. 2. Hymenoptera.
Ann. Ent. Soc. America 63:1-27.

. 1970b. Hymenoptera. In Taxonomist's

Glossary of Genitalia in Insects (Tuxen, S.
L., ed.), 2nd edition. Munksgaard, Copen-
hagen, pp. 156-170.

Sneath, p. H. a. and R. R. Sokal. 1973. Nu-
merical Taxonomy. Freeman, San Francisco,
pp. i-xvi, 1-573.

Snodgrass, R. E. 1910. The thorax of the Hyme-
noptera. Proc. United States Nat. Mus. 39:
37-91, pi. 1-16.

. 1935. Principles of Insect Morphology.

McGraw-Hill, New York etc. pp. i-x, 1-667.

. 1941. The male genitalia of Hymenoptera.

Smithson. Misc. Coll. 99(14) :l-86, pi. 1-33.

Spinola, M. 1851. V. MutiUarias. I Mutilla, II
Bradinobeno, VI Chesto. In Historia Fisica
y Politica de Chile (Gay, C), Zoologia 6:
269-287, 305-308.

Spradbery, J. P. 1973. Wasps. Sidgwick & Jack-
son, London, pp. i-.xvi, 1-408, frontis., pi.
1-28.

Stanley, S. M. 1973. An explanation for Cope's
Rule. Evolution 27:1-26.

Stephen, W. P., G. E. Boh.art and P. F. Torchio.
1969. The biology and external morphology
of bees with a synopsis of the genera of
northwestern America. Agric. Expt. Stn.,
Oregon State University, Corvallis. pp. i-ii,
1-140.

Suarez, F. J. 1960. Las Myrmosa de Marruecos
(Hym., Myrmosidae). Hesperis Tamuda 1:
111-141.

. 1962. Datos sobre mutilidos neotropicales.

III. Un nuevo genero de centroamerica y
paises noroccidentales de America del Sur
(Hymenoptera). Eos 38:115-146.

. 1965. Datos preliminares al estudio de los

mutilidos Ibericos (Hymenoptera). Eos 40:
569-586.

648

The University of Kansas Science Bulletin

Thomson, G. C. 1870. XIII. Ofversigt af Sveriges
rofsteklar. Familia Mutillidae. Opuscula
Entomologica, Lundae 2:207-209.

Tobias, V. I. 1965. K poznaniyo semeistva Fedt-
schenkiidae (Hvmenoptera, Sapygoidea). Zool.
Zhurnal 44:706-715.

ToRCHio, P. F. 1970. The ethology of the wasp,
Psetidomasaris edwardsii (Cresson), and a
description of its immature forms (Hymenop-
tera: Vespoidea, Masaridae). Los Angeles
Co. Mus., Contrib. Sci. 202:1-32.

. 1972. Sapyga pttmila Cresson, a parasite

of Megachile rottindata (F.) (Hymenoptera:
Sapygidae; Megachilidae). I: Biology and
description of immature stasjcs. Melanderia
10:1-22.

Traub, R. 1972. The zoogeography of fleas (Si-
phonaptera) as supporting the theory of con-
tinental drift. J. Med. Ent. 9:584-589.

TuoMiKosKi, R. 1967. Notes on some principles of
phylogenetic systematics. Ann. Ent. Fennici
33:137-147.

Turner, R. E. 1910. Hymenoptera. Fam. Thynni-
dae. Genera Insectorum 105:1-62, pi. 1-4.

. 1912. Studies in the fossoriai wasps of the

family Scoliidae: subfamilies Elidinae and
Anthoboscinae. Proc. Zool. Soc. London
1912:696-754, pi. 81-83.

Van Valen, L. 1973. Are categories in different
phyla comparable? Taxon 22:333-373.

Wasbauer, M. S. 1966. Revision of the male wasps
of the genus Bracbycistis in America north of
Mexico. Univ. California Pubis Ent. 43:i-vi,
1-96.

. 1968. Some sex associations in the Brachy-

cistidinae (Hymenoptera: Tiphiidae). Pan-
Pacific Ent. 44:297-299.

. 1971. A sex association in the genus

Brachycistis (Hymenoptera: Tiphiidae). Pan-
Pacific Ent. 47:211-214.

. 1973. Some new taxa in the Myrmosinae

with keys to the females in North America
(Hvmenoptera: Tiphiidae). Pan-Pacific Ent.
49:325-337.

Wesmael, C. 1851. Revue critique des Hymenop-
teres fouisseurs de Belgique. Bull. Acad. R.
Sci. Belgique 18:362-381.

Wheeler, G. C. and J. Wheeler. 1952. The ant
larvae of the subfamily Ponerinae — Part L
American Midi. NaL 48:111-144.

and . 1971. Ant larvae of the sub-
family Myrmeciinae (Hymenoptera: Formici-
dae).' Pan-Pacific EnL 47:245-256.

Williams, F. X. 1919. Philippine wasp studies.
Part 2. Descriptions of new species and life
history studies. Hawaiian Sug. PI. Assoc.
Expt. Stn., Bull. (EnL Ser.) 14:19-186.

. 1927. Etiparagia sciitellaiis Cresson, a

masarid wasp that stores its cells with the
young of a curculionid beetle. Pan-Pacific
EnL 4:38-39.

. 1928. Studies in tropical wasps — their hosts

and associates (with descriptions of new spe-
cies). Hawaiian Sug. PI. Assoc. Expt. Stn.,
Bull. (EnL Ser.)19:i-vi, 1-179.

Wilson, E. O. 1971. The Insect Societies. Belknap

Press, Harvard University, Cambridge, Mass. .
pp. i-xii, 1-548.

Wilson, E. O., F. M. Carpenter and W. L. Brown,
Jr. 1967. The first Mesozoic ants, with the
description of a new subfamily. Psyche 74:
1-19.

Wilson, E. O. and D. J. Farish. 1973. Predatory
behaviour in the ant-like wasp Methocha
stygia (Sav) (Hvmenoptera: Tiphiidae).
Anim. Behav. 21:292-295.

Zavattari, E. 1910. Sulla posizione sistematica del
genere Bradynobaentts Spin. Boll. Mus. Zool.
AnaL Comp. R. Univ. Torino 25(621) :l-5.

^^\ S'AJ/^'Ua-ujrenciJ

I

i

i

I

I

i

.V

I

1

%

%

g

>:« •,•• •:•;

& g :¥

I

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

N'.US. COMR. ZOOL.
LIBr?ARY

AUG 2 7 19/5

HARVARD
UNIVERSI fY

TAXONOMY OF FIVE NEARCTIC

SUBGENERA OF COELIOXYS

(HYMENOPTERA: MEGACHILIDAE)

• • •••

:•:•

ii

*

I

ft:

i

I

«

By
J. R. BAKER

Vol. 50, No. 12, pp. 649-730

August 18, 1975

^<

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out at
the University of Kansas or by University faculty and students. Since its incep-
tion, volumes of the Bulletin have been variously issued as single bound volumes,
as two or three multi-paper parts or as series of individual papers. Issuance is at
irregular intervals, with each volume prior to volume 50 approximately 1000
pages in length. Effective with volume 50, page size has been enlarged, reducing
the length of each volume to about 750 pages.

The supply of all volumes of the Kansas University Quarterly is now ex-
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased at
$20.00 per volume. Where some of these volumes were issued in parts, individual
parts are priced at the rate of 2 cents per page. Current policy, initiated with
volume 46, is to issue individual papers as published. Such separata may be
purchased individually at the rate of 3 cents per page, with a minimum charge
of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at
the rate of $20.00 per volume. All communications regarding exchanges, sales
and subscriptions should be addressed to the Exchange Librarian, University
OF Kansas Libraries, Lawrence, Kansas 66045.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

The International Standard Serial Number of this publication is US ISSN
0022-8850.

Editor
Eugene C. Bovee

Editorial Board

William L. Bloom
Philip W. Hedrick

Rudolf Jander

Harvey Lillywhite

Charles D. Michener

Norman A. Slade

Henry D. Stone

George W. Byers, Chairman

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. 50, No. 12, pp. 649-730 August 18, 1975

Taxonomy of Five Nearctic Subgenera of Coelioxys
(Hymenoptera: Megachilidae)^

By J. R. Baker"
TABLE OF CONTENTS

Ackn()\vkilj;ments 650 Schizocoelioxys 672

Methods 651 Boreocoelioxys 675

Characters 651 Xerococlioxys 702

Keys to Subgenera or Species 654 Coelioxys 719

Synocoelioxys 657 Literature Cited 730

ABSTRACT

The taxonomy of five subgenera of Coelioxys which are restricted primarily to the North
American continent are reviewed. The recognition of a surprising amount of infraspecific
variation in many species allowed the synonymization of 64 names into 25 presumably valid
species. The following species are described as new: C. mitchelli, C. nodis, C. oaxacana, and
C. serricandata. Keys are presented to the subgenera of North American Coelioxys and to
species in the five subgenera treated herein. Distribution maps, host records where known,
season of flight, a discussion of habitat where remarkable, of geographic variation where
noticed and of comparative comments complete the description or redescription of each species
in the fi\e subgenera.

INTRODUCTION

Five of fourteen New World subgenera level. Mitchell provides a thorough diag-

of bees in the genus Coelioxys are almost nosis of each subgenus, however, and it

entirely confined to the North American seems unnecessary to repeat such informa-

continent. This paper concerns these five tion here. Almost none of Mitchell's work

nearctic subgenera and is intended to com- is concerned with taxonomy at the species

plement a review of the New World sub- level.

genera of Coelioxys by Mitchell (1973) Bees of the genus Coelioxys are so dis-
which set up 13 new subgenera. Therefore tinctive that since Latreille (1809) erected
this paper may appear to be shallow in the the genus, not a single bee now considered
diagnostic treatment at the subgeneric to be in Coelioxys has been described in
another genus. So distinctive are these

1. Contribution number 1432 from the Depart- bees, in fact, that Dalla Torre and Friese

rcnce 16045°""°'°^^' '^^ '^"^"""'^ °^ ^'"'^'' ^^^' (1894) set up the subfamily Coelioxinae

t' " ,, ^ , , c . • XT .u based upon the genus Coelioxys. Cockerell

z. Present address: Entomology Extension, North r ^

Carolina State University, Raleigh, 27607. and Robbins (1910) changed the Spelling

649

650

The University of Kansas Science Bulletin

to Coelioxynae. The group was last given
this rank by Hicks (1926). Robertson
(1903), slightly more conservative, erected
the tribe Coelioxyini but was the only and
last (1929) person to use this rank for
Codioxys.

The genus Coelioxys shares certain
characters with the genus Megachile: The
notaulices are linear, arolia between the
claws are lacking, and the basal metasomal
tergum is shallowly concave anteriorly.
Megachile and Chalicodoma are the most
closely related genera. However, Coeli-
oxys dilTers from them in that the meta-
soma has no scopa, the axilla is elongate,
the pronotal tubercle is carinate, and the
mesepisternum has an anterior and a lat-
eral surface separated by a vertical carina.
The metanotum and posterior surface of
the scutellum are perpendicular, and the
fore coxae of both sexes possess spines
which are usually distinct. In general,
Coelioxys species are black or very dark
brown bees with white fasciae on the
metasoma. The metasomas of females
taper posteriorly, and the metasomas of
males are denticulate or spinose apically.
A character separating New World Coeli-
oxys from all other New World mega-
chilids is the presence of ocular hairs
(however, the name Coelioxys does not
mean "hairy eyes" as interpreted by Ste-
phen, Bohart and Torchio (1969:54) but
rather, according to Shuckard (1866), is
derived from two Greek words, \oilia and
oxus, which mean "acute abdomen").

ACKNOWLEDGMENTS

Of the many people who have contributed
to this study, special thanks must be extended
to Dr. Charles D. Michener whose time, con-
structive criticism, advice and encouragement
were frequently made available. Special
thanks are also extended to Dr. T. B. Mitch-
ell who originally encouraged the idea of
working with this group and who made his
personal collection available for my use.

The following people and institutions

loaned specimens and made types available
and are gratefully thanked: W. W. Moss,
Academy of Natural Sciences of Philadel-
phia; J. G. Rozen, American Museum of
Natural History; F. G. Werner, University of
Arizona; E. P. Rouse, University of Arkan-
sas; V. M. Tanner, Brigham Young Univer-
sity; G. R. Else, British Museum (Natural
History); J. J. Pasteels, University Libre de
Bruxelles; P. H. Arnaud, Jr., California
Academy of Sciences; P. D. Hurd, then of
the University of California, Berkeley; R. O.
Schuster, University of California, Davis;
E. I. Schlinger, P. H. Timberlake (personal
collection) University of California, River-
side; A. R. Moldenke, (personal collection)
University of California, Santa Cruz; H. E.
Milliron, Canada Department of Agriculture;
G. Wallace, Carnegie Museum of Natural
History; U. N. Lanham, University of Colo-
rado Museum; G. I. Stage (personal collec-
tion). University of Connecticut; L. L.
Pechuman, Cornell University; M. E. Prokop,
Field Museum of Natural History; H. V.
Weems, Florida Department of Agriculture
and Consumer Services, Division of Plant In-
dustry; W. T. Atyeo, University of Georgia;
H. W. Smith, University of Idaho; W. E.
LaBerge, Illinois Natural History Survey;
D. W. Ribble, University of Indiana, Ko-
komo; C. D. Michener, University of Kansas;
H. D. Blocker, Kansas State University; R. R.
Snelling, Los Angeles County Museum of
Natural History; T. E. Moore, University of
Michigan; K. Mac Arthur, Milwaukee Public
Museum; P. J. Clausen, University of Minne-
sota; B. C. Ratcliffe, University of Nebraska;
T. B. Mitchell (personal collection), North
Carolina State University; E. Cross, North-
western State College, Natchitoches, Louisi-
ana; C. A. Triplehorn, Ohio State Univer-
sity; C. E. Hopla, University of Oklahoma;
W. A. Drew, Oklahoma State University;
R. B. Roberts, Oregon State University; K. C.
Kim, Pennsylvania State University; W. Fer-
guson, San Jose College; H. R. Burke, Texas
A and M University; G. I. Stage, then of the
National Natural History Museum; W. J.
Hanson, Utah State University; J. D. Eves
and M. T. James, Washington State Univer-

Taxonomy of Five Nearctic Subgenera of Coelioxys

651

sity; L. J. Bayer, University of Wisconsin;
R. J. Lavigne, University of Wyoming.

Finally I wish to thank my wife, Faye H.
Uaker, for her exceeding patience and help
during the preparation and typing of this
manuscript.

METHODS

Approximately 8,000 specimens were
examined. Examination of type specimens
was expedited by carrying a synoptic
collection to various museums. Thus
"homotypes" could be selected and later
scrutinized at a more leisurely pace. When
none of the specimens in the synoptic col-
lection agreed (within a reasonable degree)
with a type, notes describing the variation
and photographs were made. Type speci-
mens were run through the appropriate
keys and compared with the appropriate
descriptions so that all descriptions agree
with all type specimens in each synonymy.
After consideration of the general lack of
reliability of the characters of hidden
sterna and genital armature at the species
level, a decision was made not to dissect
out these parts of type specimens. At least
these structures will be intact should a
future systematist discover more stable
characters than I have been able to find
in these hidden segments.

Drawings of adults were made with
the aid of an ocular grid except for the
hidden sterna of males which were cleared
in hot KOH, mounted on slides, and pro-
jected through a microscope for drawing.
Use of an incandescent bulb in the illumi-
nator of the binocular microscope seemed
to be more effective than fluorescent light-
ing in the differentiation of dark brown
areas of the integument.

Literature citations in the synonymies
are restricted to articles in which synony-
mies were made and to biological studies.
Extensive citations are given in a thesis by
Baker (1972).

Characters are numbered throughout

each description in order to facilitate com-
parisons among species. To save space,
references to characters of a species are
deleted when they agree with the descrip-
tion of those characters for the first species
treated in the group. Thus if a number
does not appear in a description, that char-
acter agrees with the description of the
first species in the group. Although this
method may at first seem confusing, the
value in space saved will undoubtedly out-
weigh the value of redundancy.

Flower records are given in a thesis by
Baker (1972).

Host records and season of flight were
taken from specimen labels (or from the
literature, as noted).

Kiichler's (1964) map of potential vege-
tation types of the contiguous United
States was used to try to find some corre-
lation between species distributions and
vegetation types. Many species extend
through so many types that they appear to
be distributed regardless of vegetation type.
Some of the more restricted bees, however,
seem to be associated with a group of
vegetation types such as western conifer-
ous forests or western shrub and grassland.

The distribution of any intraspecific
variants noted was examined to see if
there was a relation to geography. Size,
color (of the legs especially), and crenu-
lations of the sixth tergum in males, were
all found to vary geographically in some
species.

CHARACTERS

The following list enumerates char-
acters found to vary among the species of
the five subgenera of Coelioxys in this
study. Explanatory comments are added
wherever they seem to facilitate under-
standing. The Characters are numbered
to correspond to the numbers used in de-
scriptions (the first number for females,
the second number for males).

(1, 24) Length: measured in a dor-

652

The University of Kansas Science Bulletin

rest of face

lateral ocellus,
ocellar area

gena
vertex

vertex

preoccipital
carina

gena

paraocular carina

hypostonnal area of
the gena

ventral angle

malar area

gena

paraocular carina

dorsal carina of
the genal excavation

genal excavation
(hypostonnal area
of the gena)

preocc ipita I
carina

gen a

dorsal carina of the
genal excavation

preoccipital carina

gena

genal excavation (hypostomal area of the gena)

Figures 1-6. 1, 2, Dorsal and frontal views of head of Coelioxys texaiia female (arrow, angle from which
measurement of ocular hair made), a, ocelloccipital distance; b, interocellar distance; c, ocellocular
distance; d, interantennal distance; e, antenniKular distance; f, ciypeoantennal tlistance; g, lateral
margin of clypeus. 3-6. Lateral views of male Coelioxys heads. 3, C. texana; 4, C. juneraria;
5, C. sodalis; 6, C. edita.

Taxonomy of Five Nearctic Subgenera of Coelioxys

653

sal view by an ocular scale in a binocular
microscope. Lengtb is the distance (in
mm) from the bases of the antennae to
the ti[i of the metasoma. By measuring
from the bases of the antennae, compensa-
tion was made for specimens in which the
normally hypognathous head is turned up
t(i the prognathous position.

(2, 25) Integument: punctation and
color.

(3, 26) Pubescence: color, density,
form. Pubescence of the tarsi is almost
inv.iriably golden.

(4, 27) Ocular hairs: length. Ocular
hairs are longest along the posterior mar-
gin of the compound eye. The measure-
ment for all descriptions was made on the
disc of the eye (Fig. 1) at the middle.

(5, 28) Clypeus: amount and type of
pubescence, texture of integument, and
conformation of the anterior margin.

(6, 29) Clypeoaxtennal Distance and
the Lateral Margin of the Clypeus (/
and g in Fig. 2) : relative lengths.

(7, 30) Paraocular Area (Fig. 2) :
amount and type of pubescence (subject
to wear), and integumental characters
(when the integument is normally ex-
posed in most specimens).

(8, 31) The Rest of the Face (Fig.
2) : pubescence, conformation, and punc-
tation.

(9, 32) The Clypeoantennal, Inter-
antennal, and Antennocular Distances
(/, d, and e in Fig. 2) : relative lengths.

(10, 33) The Ocellar Area (Figs. 1, 2,
roughly equivalent to Mitchell's (1973)
subocellar area) : conformation.

(11, 34) OCELLOCCIPITAL, OcELLOCULAR,

AND Interocellar DISTANCES {u, c, and b

in Figs. 1, 2) : relative lengths.

(12, 35) Vertex: punctation and pu-
bescence.

(13, 36) Gena: width in relation to
the width of the eye (measured at the
broadest part of both structures in lateral
view).

(14, 37) Hypostomal Area of the
Gena: conformation, pubescence, angle of
the posteroventral carina (the ventral an-
gle). The hypostomal area of the gena
varies greatly in pubescence and confor-
mation between sexes among species and
groups. This problem is most acute in
males in which many times the ventral
portion of the gena is deeply excavated. In
Coelioxys texana the paraocular carina al-
most surrounds the eye and more or less
defines the posterior margin of the malar
area (Fig. 3). The gena sweeps from the
vertex to the hypostomal carina without
interruption. In C. funeraria a. remnant of
the paraocular carina remains posterior to
the eye and another carina, the dorsal
carina of the genal excavation, separates
the gena from the genal excavation (Fig.
4). In C. sodalis, no remnant of the para-
ocular carina remains in the area of the
gena (Fig. 5). In C. edita the dorsal ca-
rina of the genal excavation recurves dor-
sally where it merges with preoccipital
carina (Fig. 6). In C. ritfitarsis the con-
dition of the genal excavation approxi-
mates that of C. funeraria.

(15, 38) Mandible: conformation.
Wear may give the mandibles a more
robust appearance.

(16, 39) Scutum: punctation and pu-
bescence.

(17, 40) Mesepisternum: pubescence
and punctation. The mesepisternum is
divided into anterior and posterior sur-
faces by a distinct carina (Fig. 7). The
lateral surface sometimes has minute punc-
tures as well as larger punctures (Fig. 8).
The anterior surface of the mesepisternum
shows little variation throughout the five
groups considered herein.

(18, 41) Scutellum: punctation, pu-
bescence, and conformation of carina sepa-
rating the dorsal and posterior surfaces in
both dorsal and lateral views (Figs. 10
and 14 F).

(19, 42) Axilla: length.

654

The University of Kansas Science Bulletin

(20, 43) Metanotum: pubescence.

(21, 44) Fore Coxal Spine: shape,
size, and the direction of protrusion.

(22, 45) Metasomal Terga: color, pu-
bescence, punctation, and the presence or
absence of graduli and foveae (Fig. 9) . The

su&pieurai s'g'ijm'

Figures 7-9. Various Coelioxys parts. 7, Antero-
lateral view of mesosoma of female C. texana;
8, Portion of integument of mesepisternum show-
ing minute and larger punctures of female C.
texana; 9, Dorsal view of metasoma of male of
C. rtifitarsis. Roman numerals indicate tcrgum
numbers.

shapes of the apices of the sixth and sev-
enth terga are sometimes specifically dis-
tinct (especially in males), and the general
conformation of these segments is impor-
tant at the subgeneric level.

(23, 46) Exposed Metasomal Sterna:
punctation, pubescence, and conformation
of the graduli. The apical margin of the
sixth sternum of females displays a num-

ber of characters useful at the subgeneric
and specific levels.

(47) Hidden Sterna and Genital
Armature of Males: punctation, pubes-
cence, and conformation. These characters
are useful at the subgeneric level but are
of less use, if any, at the species level.

KEYS TO SUBGENERA OR SPECIES

The following keys separate species of
North American Coelioxys into subgenera
or species (for species not included in the
five subgenera covered by this study). All
Coelioxys species for which specimens
were available from Mexico and north-
ward are included.

Distribution is used as a character only
in the two following keys, where species
not otherwise treated in this paper can be
separated by range. Careful taxonomic
studies will have to be done before mor-
phological characters are found to separate
the members of several groups of similar
species, and some species currently sepa-
rated may be found to be synonymous.
For example, the western forms Coelioxys
gilensis and C. deani will probably be
found synonymous when a systematic re-
view is made of that group as will likely
also be the case with C. modesta and C.
scitula, two eastern forms. However, the
morphological differences separating C.
gilensis and C. deani from C. modesta and
C. scitula are subtle, and the decision to
recognize these four names as two or one
species will be a difficult one. Distribution
separates the two groups readily, although
the separation may be an artificial one.

Key to Females of North American
COELIOXYS.

1. Oceliar area (Figs. 1-2) swollen, im-

punctate 2

Oceliar area not swollen or if raised then
closely punctate 8

2. Sixth metasomal sternum broadly

rounded to acute apicaliy, not notched

subapically (Fig. 14A-D)

Synocoelioxys

Taxonomy of Five Nearctic Subgenera of Coelioxys

655

Sixth metasomal sternum apically or
SLibapically notched 3

3. Vertex between lateral ocellus and com-

pound eye completely or almost im-
punctate, shining; scutellum almost im-
punctate medially, with a large,
rounded projection posteriorly; sixth
metasomal tergum blunt or truncate.

4

Vertex moderately to closely punctured;
scutellum variously punctured, but with
no projection; sixth metasomal tergum
usually rounded 5

4. Wings blackish-brown; large bee (13-16

mm); southeastern United States

dolichos Fox

Wings pale brown; medium sized bee
(10-12 mm); Mexico and southward

laevigata Smith, laevis Friese,

tolteca Cresson

5. Scutellum impunctate medially or with

scattered punctures, posterior margin of

dorsal surface elevated 6

Scutellum closely, evenly and deeply
punctured, posterior margin of dorsal
surface not raised 7

6. Only basal metasomal tergum ferrugi-

nous across entire width; scutellum im-
punctate medially; Mexico and south-
ward assHmptionls Schrottky

Basal two or three metasomal terga ferru-
ginous across entire widths; scutellum
punctured medially; northern Mexico
and northward menthae Cockerell

7. Thoracic and metasomal fasciae dense,

conspicuous

slossoni arenicola Crawford

Thoracic and metasomal fasciae narrow,

inconspicuous

slossoni slossoni Viereck

8. Sixth metasomal sternum notched sub-

apically (as in Fig. 21) 9

Sixth metasomal sternum with margin
entire 12

9. Scutellum without medioposterior pro-

jection 10

10. Scutellum with longitudinal, raised, im-

punctate line medially

mextcana Cresson

Scutellum without raised, impunctate
line medially 11

11. Metasomal terga two and three with

graduli complete Boreocoelioxys

Metasomal terga two and three with
graduli incomplete Schizocoelioxys

12. Sixth metasomal tergum and sternum

attenuate apically, sixth sternum with-
out conspicuous, erect subapical setae;

Mexico and southward

zapoteca Cresson

Sixth metasomal tergum and sternum
variable but not conspicuously elongate,
sixth sternum usually with erect sub-
apical setae 13

13. Scutellum with posterior margin in dor-

sal view medially subangulate or with
medioposterior projection; Mexico and

southward 14

Scutellum with posterior margin in dor-
sal view gently rounded to straight ..15

14. Scutum and axilla black

chichimeca Crawford

Scutum and axilla with distinct ferrugi-
nous areas gonaspts Cockerell

15. Axilla with projecting portion as long as

basal portion .... obtiisiventris Crawford

Axilla with projecting portion about half

as long as basal portion 16

16. Scutum and axilla with ferruginous

areas; Mexico and southward

sanguinicolUs Friese

Scutum and axilla black; usually north-
ern Mexico and northward 17

17. Sixth metasomal tergum in lateral view

upturned apically 18

Sixth metasomal tergum in lateral view
tapering horizontally 19

18. Metasoma with postgradular areas evenly

punctured (one to two puncture widths
between punctures) with small punc-
tures; west of Rocky Mountains

deani Cockerell, gilensis Cockerell

Metasoma with postgradular areas less
closely punctured medially (four or five
puncture widths between punctures)
with medium sized punctures; east of

Rocky Mountains

scitula Cresson, modesta Smith

19. Prothoracic tubercle expanded into thin,

plate-like structure Xerocoelioxys

Prothoracic tubercle with strong carina
but not expanded into thin, plate-like
structure Coelioxys

20. Eastern United States

germana Cresson

Mexico and southward

totonaca Cresson

Key to Males of North American
COEUOXyS.

1. Ocellar area impunctate and swollen .. 2

656

The University of Kansas Science Bulletin

Ocellar area closely punctured or not
swollen 9

2. Sixth metasomal tergum with dorsal

spines modified into a crenulate plate

(Fig. 12B-G) Synocoelioxys

Sixth metasomal tergum with two dis-
tinct dorsal spines (sometimes with a
third median spine in specimens from
Mexico and southward 3

3. Metasomal terga three to five with longi-

tudinal, median ridge; vertex with scat-
tered punctures; scutum practically im-
punctate on disc; scutellum impunctate
medially with a large, rounded projec-
tion posteriorly 4

Metasomal terga three to five without
longitudinal, median ridge; vertex vari-
ably punctured; scutum variably punc-
tured; scutellum moderately to closely
punctured, with or without large pro-
jection 5

4. Wings blackish-brown; large bee (10-14

mm); southeastern United States

doUchos Fox

Wings pale brown; medium sized bee
(8-10 mm); Mexico and southward ....
laevigata Smith, toUeca Cresson

5. Vertex between lateral ocellus and eye

impunctate; scutum, axilla, and scutel-
lum with large ferruginous areas; scu-
tellum with large, rounded posterior

projection medially azteca Cresson

Vertex variously punctured; scutum,
axillae usually black; scutellum without

large, rounded posterior projection

6

6. Scutellum evenly, closely, deeply punc-

tured on dorsal surface; carina sepa-
rating dorsal and posterior surfaces in-
complete laterally, only a small median
portion raised above level of dorsal

surface 7

Scutellum sparingly punctured or im-
punctate medially; carina separating
dorsal and posterior surfaces entire,
median half raised above level of dorsal
surface 8

7. Thoracic and metasomal fasciae dense,

conspicuous

slossoni arenicola Crawford

Thoracic and metasomal fasciae narrow,

inconspicuous

slossoni slossoni Viereck

8. Only the basal metasomal tergum ferru-

ginous across entire width

assumptionis Schrottky

Basal two or three metasomal terga ferru-
ginous across entire width

menthae Cockerell

9. Foveae (Figs. 9, 25B-H) on metasomal
tergum two (in C. mexicana foveae
may be very small on tergum two, but
they are in an otherwise impunctate
area) 10

Foveae not present or present on meta-
somal tergum three only 12

10. Foveae present on metasomal tergum

three mexicana Cresson

Foveae absent on metasomal tergum
three 11

11. Graduli complete on metasomal terga

two and three Boreocoelioxys

Graduli incomplete medially on meta-
somal terga two and three

Schizocoelioxys

12. Foveae absent, foveal area may be closely

punctured on metasomal terga two and

three 13

Foveae present only on tergum three
17

13. Scutellum impunctate medially, posterior

margin with flat, rounded projection

azteca Cresson

Scutellum punctured medially except
sometimes for an impunctate longi-
tudinal line, posterior margin variable
14

14. Scutellum in dorsal view with a conical

projection posteriorly

gonaspis Cockerell

Scutellum in dorsal view without a pro-
jection 15

15. Metasomal tergum six with inner mar-

gins of dorsal spines forming a broad

V zapoteca Cresson

Metasomal tergum six with inner mar-
gins of dorsal spines forming a broad
U 16

16. Thoracic and metasomal fasciae distinct;

prothoracic tubercle produced into thin,
plate-like structure; gradulus of tergum
two curved toward apical margin medi-
ally Xerocoelioxys

Thoracic and sometimes metasomal fas-
ciae inconspicuous; carina of protho-
racic tubercle conspicuous but not ex-
panded into thin, plate-like structure;
gradulus of tergum two almost straight
Coelioxys

17. Posterior margin of scutellum in dorsal

view with a median projection half as
long as axilla 18

Taxonomy of Five Nearctic Subgenera of Coelioxys

657

Posterior margin of scutellum without

median projection 19

1 8. Eastern United States .. germana Cresson
Mexico and southward

totonaca Cresson

V). Scutellum with narrow, impunctate,

longitudinal median carina

mexicana Cresson

Scutellum without longitudinal median

carina 20

20. Posterior margin of scutellum medially

subangulate in dorsal view; basal meta-
somal terga with ferruginous areas;

Mexico and southward

chichimeca Cresson

Posterior margin of scutellum broadly
rounded in dorsal view; metasoma
black or very dark brown throughout;
northern Mexico and northward .... 21

21. Metasoma with postgradular areas

closely and evenly punctured (maxi-
mum of one to two puncture widths
between punctures) with small punc-
tures; west of Rocky Mountains

deani Cockerell, gilemis Cockerell

Metasoma with postgradular areas less
closely punctured medially (maximum
of four to five puncture widths between
punctures) with medium sized punc-
tures; east of Rocky Mountains

scitula Cresson, modesta Smith

Subgenus Synocoelioxys Mitchell

Synocoelioxys Mitchell, 1973. A subgeneric
revision of the genus Coelioxys of the
Western Hemisphere. Cont. Dept.
Entomol. North Carolina State Univ. p.
57. Type species: Coelioxys texana
Cresson.

Table 1. Known host-parasite relationships for North American Megacliile and Coelioxys. References
to sources for species not covered by this review are given in footnotes. B = Boreocoelioxys, Co ■=.
Coelioxys, Cy ^ Cyrtocoelioxys, M ^ Melanocoelioxys, Sc ^ Schizocoelioxys, Sy =: Synocoelioxys,
X = Xerocoelioxys.

MEGACHILE and CHALICODOMA COELIOXYS

Subgenus

species

species

Chelostomoides

campanulae

Chelostomoides

subexilis

Delomegachile

frigida

Delomegachile

melanophoea

Delomegachile

melanophoea wootoni

Eutricharaea

concinna

Eutricharaea

rotundata

Litomegachile

brevis

Litomegachile

gentilis

Litomegachile

mendica

Litomegachile

onobrychidis

Litomegachile

texana

Megachile

centuncularis

Megachile

inermis

Megachile

montivaga

Megachile

relativa

Megachiloides

umatillensis

Melanosarus

xj'locopoides

Phenosarus

fortis

Pseudocentron

sidalceae

Sayapis

policaris

Sayapis

pugnata pugnata

Xanthosarus

latimanus

Xanthosarus

perihirta

modesta (Cy)^

gilensis (Cy)3

funeraria (Sc), moesta (B), porterae (Co)

rufitarsis (B), sodalis (Co)

rufitarsis (B), sodalis (Co)

moesta (B)

funeraria (Sc), gilensis (Cy)*, moesta (B),

novomexicana (B), octodentata (B), sodalis (Co)
novomexicana (B), octodentata (B), sayi (B)

novomexicana (B)

octodentata (B), sayi (B)

octodentata (B)

octodentata (B), moesta (B), rufitarsis (B),

sodalis (Co)
modesta (Cy)5, moesta (B), octodentata (B)
funeraria (Sc)
rufitarsis (B)
funeraria (Sc), modesta (Cy)6, moesta (B),

porterae (Co)

mesae (X)
dolichos (M)l
rufitarsis (B)
novomexicana (B)

texana (Sy)

alternata (Sy)

funeraria (Sc), rufitarsis (B)

grindeliae (X), octodentata (B), rufitarsis (B)

3. Krombein (1967).

4. From specimen labels in Washington, Canada.

5. Graenicher (1927).

6. Fye (1965).

658

The University of Kansas Science Bulletin

Figure 10. Dor.sal view of female Coelioxys texana, left half showing punctations, right half with
melanization and pubescence added.

Taxonomy of Fivl Nearctic Subgenera of Coelioxys

659

Figure 11. Frontal view of females of Synocodtoxys.
A, Head of C. texana. B-C, Left half of clypeus
and mandible. B, C. alternata; C, C. htiuteri.

Other bees in the subgenus Sytwcoeboxys
are C. alternata, C. apacheorum, C. hiinteri,
and C. erysimi. This group of bees, although
possessing many derived characters, possesses
several characters which appear to be ances-

tral: the seventh sternum of males is com-
plete, the sixth sternum of females has a
simple margin and is not unduly lengthened,
and the sixth tergum of males appears to be
most ancestral (i.e. resembles most closely the
condition found in some species of Megachile
and Liothyrapis) of any of the New World
Coelioxys.

Certain host bees are known for two
species in Synocoelioxys: Coelioxys texana
and C. alternata. Both utilize Megachile
hosts of the subgenus Sayapis (Table 1).

Most of the characters in the list below
are found in other subgenera of New World
Coelioxys and even in some of the Old
World Liothyrapis, but never in this combi-
nation. The simple seventh sternum and the
crenulate margin of the sixth tergum of
males are found in no other New World
Coelioxys. Italicized characters most clearly
differentiate Synocoelioxys from other sub-
genera in North America.

A. Ocellar area raised, impunctate. Me-
dian ocellus margined anteriorly by small
group of anteriorly directed setae (Figs. 10,
llA, 12A).

Figure 12. Dorsal views of males of Synocoelioxys. A-C, C. texana. A, Head; B, Mctasoma; C, Sixth
metasomal tergum. D-G, Sixth metasomal terga. D, C. alternata; E-F, C. apacheorum ; G, C. hunteri.

660

The University of Kansas Science Bulletin

B. Preoccipital carina complete medially

(Fig. 3).

C. Mandible robust to slender (Fig.
IIA-C).

D. Prothoracic tubercle with carina pro-
duced into thin plate-like structure.

E. Scutum moderately punctured (Fig.
10), fasciae indistinct to distinct.

F. Scutellum not flattened and usually
without a projection although a small carina
may be present (Fig. 14E).

G. Axilla well produced except in Coeli-
oxys alternata (Figs. 10, 14E).

H. Mesepisternum moderately punctured,
with minute punctures as well as larger
punctures (Fig. 8).

I. Gradular grooves incomplete medially,
becoming filled with squamose setae laterally
(Figs. 10, 12B).

}. Female: front coxal spine variable but
well produced in some species.

K. Female: lateral margin of sternum six

Figure 13. Hidden sclerites of males of Synocoelioxys. A-F, C. texatia. A, Seventh tergum; B, Margin
of fourth sternum, fifth sternum; C, Sixth strenum; D, Seventh sternum; E, Eighth sternum; F,
Genital armature (arrow indicates rugulose area). G, Sixth sternum, C. alttniata; H, Sixth sternum,
C. apaclieoriim; I, Genital armature, C. alternata. ~- •■ ■

Taxonomy of Five Nearctic Subgenera of Coelioxys

661

entire, acute or rounded apically (Figs. 10,
14A-C).

L. Male: hypostomal area of the gena
not modified into excavation (Fig. 3).

M. Male: joveal area of tergum two
sometimes very closely punctate (Fig. 12B).

N. Male: fifth tergum with lateral spine
close to apical margin.

O. Male: carina of tergum six forming
a crenulate plate (Fig. 12B-G).

P. Male: apex of tergum seven with a
conspicuous spine (Fig. 13A).

Q. Male: apex of sternum four with a
pair of teeth (Fig. 13B).

R. Male: sternum seven undivided , bi-
lobed (Fig. 13D).

Key to Females of Synocoelioxys

1. Sixth sternum acute apically (or very

narrowly rounded) (Fig. 14A, C) 2

Sixth sternum rounded apically (Figs. 10.
14B, D) 3

2. Axilla short (Fig. 14 E); mandible slen-

der (Fig. IIB) alternata

Axilla longer (as in Fig. 10); mandible
intermediate (Fig. UC) hunteri

3. Lateral ocellus closer to occipital margin

than to eye 4

Lateral ocellus equidistant from occipital
margin and eye texana

4. Sternum six broadly rounded (Fig. 14D)

erysimi

Sternum six more narrowly rounded (Fig.
14B) apacheorum

Figure H. A-D, Sixth metasomal tcrga and sterna
of females of Synocoelioxys. A, C. alternata; B,
C. apaclieonim; C, C. hunteri; D, C. erysimi.
E-F, Dorsal (E) and optical section (F) of scu-
teilum and axillae of C. alternata.

Key to Males of Synocoelioxys

1. Hypostomal area of the gena bare or with
small slender setae which do not obscure
integument 2

Hypostomal area of the gena obscured by
setae at least as wide if not as long as
setae on gena 3

2. Anterior portions of raised, impunctate

ocellar area (Figs. 1, 2) separated by
about one puncture width, usually a
longitudinal row of punctures in the
space; scutellum usually without a ca-
rina separating dorsal and posterior sur-
faces hunteri

Anterior portions of raised, impunctate
ocellar area separated by about two or
three puncture widths; scutellum usually
with a carina separating dorsal and pos-
terior surfaces (Fig. 14E) alternata

3. Ocellucular distance less than ocelloccip-

ital distance; scape and pedical usually

ferruginous texana

Ocellocular distance subequal to or greater
than ocelloccipital distance; scape and
pedicel usually piceous or black 4

4. Posterior surface of scutellum with no

appressed setae or a few slender ap-

pressed setae restricted to middle

erysimi

Posterior surface of scutellum with squa-
mose appressed setae across the full
width apacheorum

Coelioxys texana Cresson
Figures 10, llA, 12A-C, 13A-F, 15

Coelioxys texana Cresson, 1872, Trans.
Amer. Entomol. See. 4:272 (9, $
Texas, type 9 , in Philadelphia Acad-
emy of Sciences) ; Schwarz, 1896, Proc.
Entomol. Soc. Washington 4:24
(Sleeping); Robertson, 1926, Psyche 33:
177 (Phenology); Krombein, 1967,
Trap-nesting Wasps and Bees: Life
Histories, Nests and Associates, Wash-
ington, D.C. 570 p.

Coelioxys texaniis; Banks, 1902, J. New
York Entomol. Soc. 10:212 (Sleeping).

Coelioxys texana sonorensis Cockerell,
1914, Entomologist 47:116 ( <5 , Sonora,
Mexico, type in the National Museum
of History) NEW SYNONYMY.

Coelioxys texana texana: Mitchell, 1962,
North Carolina Agr. Exp. Sta. Tech.
Bull. 152:193, 194, 228 (Redescription,
key).

662

The University of Kansas Science Bulletin

FEMALE. (1) Length 9-14 mm; (2)
integument between punctures granular
on vertex, weakly so on thorax, dull on
metanotum and propodeum, shiny else-
where; black; scape, pedicel, apical por-
tion of labrum, pronotal lobe, tegula, front
coxal spine, trochanters, femora, tibiae,
ventral or medioventral portion of hind
coxa, lateroposterior margins of terga 1-4,
sometimes 5, and posterior portions of
sterna 1-5 ferruginous to reddish brown;
(3) pubescence white, slightly brown be-
low antenna, golden on tarsi, brown on
apices of tergum 6 and sternum 6; (4)
ocular hairs short (about 0.05 mm); (5)
clypeus visible medially, margin irregular
(Fig. 14A); (6) clypeoantennal distance
subequal to lateral margin of clypeus; (7)
paraocular area hidden laterally by setae
about 0.2 mm long; (8) rest of face punc-
tured up to ocellar area; (9) clypeoanten-
nal and interantennal distances equal,
greater than antennocular distance (Fig.
IIA); (10) ocellar area with impunctate
areas broad (Figs. 10, UA); (11) ocel-
loccipital distance subequal to ocellocular
distance, greater than or subequal to inter-
ocellar distance (Fig. 10); (12) vertex
moderately punctate (Figs. 10, 11 A);
(13) gena narrower than eye, setae not
obscuring integument, ventral angle about
90°; (14) hypostomal area of the gena
with setae shorter than on disc; (15) man-
dible short, broad (Fig. IIA); (16) scutum
with conspicuous fasciae (Fig. 10); (17)
lateral surface of mesepisternum with
punctures well separated, many more mi-
nute punctures than punctures, pubescence
short; (18) scutellum with posterior mar-
gin of dorsal surface subangulate in dorsal
view (Fig. 10); (19) axilla well produced
(Fig. 10) ; (20) metanotal setae appressed
medially, erect laterally; (21) front coxal
spine subtriangular, flattened, directed ven-
trally or slightly forward, inconspicuous;
(22) terga 1-5 with conspicuous fasciae, ter-
gum 6 rounded (Fig. 10) ; (23) sterna 1-5

fasciate apically, fasciae less conspicuous
posteriorly, punctures on sterna 1-4 with
short setae, punctures exceedingly small
and close medially on sternum 5 and sub-
apically on sternum 6, ventral apical mar-
gin of sternum 6 bare, shining, broadly
rounded (Fig. 10).

MALE. (24) Length 10-12.3 mm; (25)
integument as in female (2 above); (26)
pubescence white, golden on tarsi; (27)
ocular hairs of medium length (about
0.075 mm. Fig. 12A) ; (28) clypeus hidden
by setae in some specimens, coarsely ru-
gose; clypeal margin as in female (5
above); (29) clypeoantennal distance as in
female (6 above); (30) paraocular area
hidden by setae about 0.3 mm long; (31)
rest of face hidden up to ocellar area; (32)
antennocular distance less than clypeo-
antennal distance which is less than inter-
antennal distance; (33) ocellar area as in
female (10 above); (34) interocellar and
ocellocular distances equal, less than ocel-
loccipital distance (Fig, 12A) ; (35) vertex
as in female but longer (12 above. Fig.
12A); (36) gena as in female (13 above);
(37) hypostomal area of the gena with
setae as long as on discs; (38) mandible as
in female (15 above); (39) scutum as in
female (16 above) but with slender erect
setae on disc; (40-43) mesepisternum, scu-
tellum, axilla, metanotum as in female
(17-20 above); (44) front coxal spine about
twice as long as broad; (45) terga 1-6
fasciate, carina of tergum 6 crenulate, or
with sharp denticles, deeply emarginate,
often asymmetric (Fig. 12B-C), tergum 7
with apical spine (Fig. 13A); (46) sterna
1-4 evenly punctate with squamose setae
filling most punctures, sterna 1-3 apically
with irregular fasciae about 0.35 mm
wide; (47) sternum 4 submarginally cari-
nate, apex bidentate (Fig. 13B); margin
of sternum 5 slightly emarginate (Fig.
13B) ; sternum 6 with unbranched setae
medioapically (Fig. 13C) ; sternum 7 sim-
ple (Fig. 13D); sternum 8 with small,

Taxonomy of Five Nearctic Subgenera of Coelioxys

663

unbranched setae medially (Fig. 13E) ;
genital armature with area at base of penis
valves (arrow on Fig. 13F) sometimes
rugulose.

HOST RECORDS. Krombein (1967)
reported rearing Coelioxys texana from
trap nests provisioned by Megachile poli-
caris.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys texana is most abun-
dant in the southwestern United States
(Fig. 15). Exceptionally long series have
been taken at the Southwestern Research
Station five miles west o£ Portal, Arizona.

This bee has been collected from
March 27 to November 2 in Texas. One
specimen of Coelioxys texana was taken
November 17 at Ciudad Victoria, Tamu-
alipas, Mexico.

HABITAT. Coelioxys texana has

been taken at altitudes ranging from close
to sea level (Indian Wells, California) to
9,000 feet. Habitats range from humid
areas such as mangrove and floodplain
forests of Florida to xeric habitats in Ari-
zona and southern California where major
vegetational types include Larrea, Opuntia,
Cercidiiim, Bouteloita, Hilaria, Qiiaxus,
fiinipenis, Flourensia, and Franseria.
Coelioxys texana is apparently not com-
mon at higher altitudes although speci-
mens from two localities indicate it may
range up into coniferous forests (Shannon
Peak, Pinalena Mountains, Arizona, 9,000
feet and Madera Canyon, Santa Rita
Mountains, Pima County, Arizona, 4,800
feet).

GEOGRAPHIC VARIATION. Spec-
imens of Coelioxys texana from Florida
have wings which are noticeably darker
and a deeper yellow than wings of C.

Figure 15. Distribution of Coelioxys texana.

664

The University of Kansas Science Bulletin

texana from other areas. All (seven)
males from Florida also have the margin
of the carina of the sixth tergum produced
into sharp denticles (Fig. 12 C) ; elsewhere
the margin of the sixth tergum is crenu-
late (Fig. 12 B). A single female from
the Lower Matecumbe Key with dark,
yellowish wings (but typical in other char-
acters) has been associated with these
males.

COMPARATIVE COMMENTS. In

both sexes of Coehoxys texana the anten-
nal scape and pedicel are usually ferrugi-
nous, a condition found in no other North
American Coelioxys spp. Of the females
in this group, C. texana has the lateral
ocelli furthest removed from the occipital
margin (subequally distant from the oc-
cipital margin and eye).

Coelioxys alter nata Say
Figures IIB; 12D; 13G, I; 14A, E-F; 16

Coelioxys alter nata Say, 1837, Boston J.
Natur. Hist. 1 : 401 ( ^ , $ , Indiana,
type probably destroyed) ; Robertson,
1929, Flowers and Insects, Lancaster,
Pennsylvania Sci. Press (Flower rec-
ords); Graenicher, 1935, Ann. Entomol.
Soc. Amer. 28:304; Medler, 1964, Can.
Entomol. 96:918, 920, 921; Medler and
Lussenhop, 1968, Univ. Wisconsin Re-
search Bull. 274:51, 52 (Biology, key).

Coelioxys texana; Graenicher, 1911, Bull.
Publ. Mus. Milwaukee 1:243 (Wiscon-
sin, misidentification).

Coelioxys texana vegana Cockerell, 1912,
Can. Entomol. 44:166, 169 (New Mex-
ico, $ , key, type in the American
Museum of Natural History) NEW
SYNONYMY.

Coelioxys coc/^erelli Crawford, 1915, In-
secutor Inscitiae Menstruus 3:108 (9,
Colorado, type in the National Mu-
seum of Natural History) NEW
SYNONYMY.

Coelioxys wisconsinensis Cockerell, 1925,
Pan-Pacific Entomol. 1:145 ($, S,

Wisconsin, type in the University of
Colorado Museum) NEW SYN-
ONYMY.

Coelioxys alternata alternata; Mitchell,
1962, North Carolina Agr. Exp. Sta.
Tech. Bull. 152:193-195 (Redescription,
key).

Coelioxys alternata wisconsinensis; Mitch-
ell, 1962, North Carolina Agr. Exp. Sta.
Tech. Bull. 152:193, 194, 196; Medler,
1964, Can. Entomol. 96:920, 921 (Host
record).

FEMALE. Agrees with description
and figures of Coelioxys texana except as
follows: (1) Length 10-16 mm; (2) in-
tegument black or becoming dark reddish-
brown to ferruginous on femora, tibiae,
and metasoma; (3) pubescence white ex-
cept golden on tarsi; (4) ocular hairs 0.08
mm long; (5) clypeus with setal fringe
scant; (6) clypeoantennal distance shorter
than lateral margin of clypeus; (7) para-
ocular area rugose, almost hidden by 0.3-
0.4 mm setae; (8) rest of face with median
impunctate ridge; (9) antennocular dis-
tance subequal to clypeoantennal distance
and to interantennal distance; (10) ocellar
area with impunctate ridges not as con-
spicuous as those in Fig. 10; (11) inter-
ocellar distance greater than ocelloccipital
distance and less than ocellocular distance;
(12) vertex with longer, more copious setae
than shown in Fig. 10; (13) gena with
ventral angle acute, approaching 60°; (14)
hypostomal area of the gena appearing
bare; (15) mandible slender (Fig. IIB);

(16) scutum with peripheral fasciae less
conspicuous than shown in Fig. 10; thin
setae on disc longer than shown in Fig. 10;

(17) mesepisternum with lateral surface
closely punctate, pubescence long and
slender, lateral surface with as many or
more punctures than minute punctures;

(18) scutellum not angulate medially as
seen from above, with carina separating
surfaces (Fig. 14E, F); (19) axilla short
(Fig. 14E); (21) front coxal spine re-

Taxonomy of Five Nearctic Subgenera of Coelioxys

665

ducccl sometimes slightly curved posteri-
orly; (22) terga 1-6 in general more slen-
der than those of Fig. 10; (23) sterna 1-5
with apical fasciae of sparse hairs, punc-
tures laterally with short, slender setae,
punctures rather uniform throughout;
sternum 6 acute apically (Fig. HA),
ventrad with punctures contiguous basally,
in parallel rows.

MALE. Agrees with description and
figures of Coelioxys texana except as fol-
lows: (25) integument as in female (2
above); (27) ocular hairs of medium
length (about 0.1 mm); (2<S) clypeus hid-
den by setae in some specimens, coarsely
punctured, shining when revealed; (29)
clypeoantennal distance as in female (3
above); (30) paraocular area as in female
(7 above); (31) rest of face with weak
impunctate line to subocellar area; (32)
clypeoantennal distance subequal to inter-
antennal distance, greater than antennocu-
lar; (33) ocellar area as in female (10
above); (34) ocelloccipital distance sub-
equal to ocellocular distance and greater
than interocellar distance; (35) vertex with
more slender upright setae than shown in
Fig. 12A; (36-42) gena, hypostomal area
of the gena, mandible, scutum, mesepi-
sternum, scutellum, axilla as in female
(13-19 above); (43) metanotum with al-
most no appressed setae; (44) front coxal
spine pointing ventrally, about as long as
broad; (45) tergum 1 with median fascia
not as complete as shown in Fig. 12 B,
tergum 6 with carina denticulate (Fig.
12D), tergum 7 with apical projection
slio-htlv recurved, shorter than shown in
Fig. 13A; (46) sterna 1-4 with apical
fasciae interrupted medially, sternum 6
with a few slender scattered setae medio-
apically (Fig. 13G), genital armature
with ventral lobe of gonocoxite acute api-
cally (Fig. 131).

HOST RECORDS. Medler (1964) re-
ported rearing Coelioxys alternata from

trap nests provisioned by Megachile (Sa-
yapis) pugnata piignata.

DISTRIBUTION AND SEASON IN
FLIGHT. Although rare throughout its
range, Coelioxys alternata occurs as far
north as Edmonton, Alberta (Fig. 16). In
the southernmost extensions of its range,
C. alternata occurs only at higher eleva-
tions. One specimen from the American
Museum of Natural History labeled "Tex.
Collection Belfrage" is probably mis-
labeled. Somewhat surprising in view of
its coast to coast distribution is a lack of
specimens from California. The host bee,
Megachile pugnata piignata occurs as far
west in California as the eastern sides of
the Sacramento and San Joaquin Valleys.

Coelioxys alternata has been collected
from June 24 to October 13.

HABITAT. Mesic conditions prevail
throughout most of the locaHties from
which Coelioxys alternata has been col-
lected. Dryer vegetational types include
areas classified as bluestem prairie (Andro-
pogon), oak savana (Qiierciis, Andro-
pogon ), and oak-juniper woodland (Quer-
cus, Juniperus). Further north, C. alternata
appears to be associated with various conif-
erous and hardwood forests.

GEOGRAPHIC VARIATION. Vari-
ation in Coelioxys alternata does not seem
to be related to geography. The color of
the legs and tegulae grades from black
(previously considered to be C. alternata
wisconsinensis) to ferruginous. Medler
(1964) reared both black and ferruginous
forms from the same domicile. Actually
most specimens exhibit an intermediate
condition, and all degrees of melanism are
found throughout the range. The co-
existence of forms with all degrees of
melanism throughout the range, morpho-
logical agreement among forms, and iden-
tical host data were major factors con-
tributing to the synonymyzing of C. alter-
nata wisconsinensis and C. alternata
alternata.

666

The University of Kansas Science Bulletin

Figure 16. Distribution of CocUoxys altcniata.

COMPARATIVE COMMENTS.

Males of Coelioxys alternata are known
from other males in Synocoelioxys (ex-
cept C. humeri) by the bare appearance of
the hypostomal area of the gena. Unlike
all other species of Synocoelioxys, the
axillae in both sexes are usually very short
(Fig. HE). Females of C alternata pos-
sess sixth strena more acute apically (com-
pare Figs. 14A and 14C) and mandibles
more slender than those of C. hunteri
(compare Figs. IIB and IIC).

Coelioxys apacheontm Cockerell
Figures 12E-F, 13H, 14B, 17

Coelioxys apacheontm Cockerell, 1900,
Can. Entomol. 32:297, 299-307 (9,
New Mexico, key, type in the National
Museum of Natural History) ; Linsley
in Muesebeck et al., 1951. Hymenoptera

of America North of Mexico; Synoptic
Catalog USDA Agric. Mono. 2:1183
(^fragariae) .

Coelioxys fragariae Cockerell, 1912, Can.
Entomol. 44:167-169 ( <5 , California,
key, type in the American Museum of
Natural History); Cockerell, 1921,
Amer. Mus. Novitates 21 :3, 8-10 (Colo-
rado, redescription).

Coelioxys qiiercina Cockerell, 1912, Can.
Entomol. 44:167-169 ( $ , Arizona, key,
type in the American Museum of Nat-
ural History) NEW SYNONYMY.

FEMALE. Agrees with description
and figures of Coelioxys texana except as
follows: (1) Length 9-13 mm; (2) integu-
ment grainy on vertex, dull on thorax,
shiny on abdomen; black; scape, front
coxal spine, and coxae piceous; pronotal
lobe and tegula dark brown to brown;

Taxonomy of Fivi: Nearctic Subgenera of Coelioxys

667

femora, tibiae and basitarsi dark brown to
dark brown with ferruginous spots or seg-
ments to ferruginous throughout; kitero-
posterior margins of terga 1-4 and some-
times 5, and posterior margins of sterna
1-5 black to dark reddish-brown; (3) pu-
bescence brown on apex of sternum 6; (4)
ocukir hairs 0.05-0.1 mm long; (5) margin
of clypeus with 6 teeth about 0.05 mm
long; clypeus sometimes with irregular
impunctate line; (6) clypeoantennal dis-
tance shorter than lateral margin clypeus;
(7) paraocular area hidden laterally by
0.2-0.3 mm setae; (8) rest of face with an
indefinite impunctate line up to ocellar
area; (9) clypeoantennal distance greater
than antennocular distance and less than
or subeqtial to interantennal distance;
(10) ocellar area with arms of impunctate
area medially separated by punctures in a
band 2 or 3 punctures wide; impunctate
area sometimes not as broad as in Figures
10, 11 A; (11) ocelloccipital distance less
than or subequal to interocellar distance
which is less than ocellocular distance;
(12) vertex with slender erect setae longer
than shown in Figure 10, especially in in-
terocellar area; (13) gena with ventral
angle slightly acute; (14) hypostomal area
of gena with setae as long as or longer
than on disc; (15) mandible with outer
tooth longer than shown in Figure 11 A,
more like Figure IIC; (16) scutum with
erect setae longer and more copious than
shown in Figure 10; (17) mesepisternum
with lateral face closely punctate, more
punctures than minute punctures, pubes-
cence slender and long (about 0.35 mm);
(18) scutellum with posterior margin of
dorsal surface almost straight to broadly
curved in dorsal view; (21) front coxal
spine flattened and rounded apically,
slightly recurved posteriorly, inconspicu-
ous; (22) tergum 6 rounded apically (Fig.
14B) ; (23) sternum 6 moderately to nar-
rowly rounded apically (Fig. 14B).

MALE. Agrees with description and

figures of Coelioxys texana except as fol-
lows: (24) Length 8.5-10.5 mm; (25) in-
tegument as in female (2 above) ; (27)
ocular hairs 0.05-0.1 mm long; (28) clypeal
margin and median line as in female (5
above); (29) clypeoantennal distance as in
female (6 above); (32) clypeoantennal
distance subequal to or greater than an-
tennocular distance, less than interanten-
nal distance; (33) ocellar area as in female
(10 above); (34) ocellocular distance sub-
equal to ocelloccipital distance, greater
than interocellar distance; (35) vertex with
more erect, slender setae than shown in
Figure 12A; {%) gena as in female (13
above); (37) hypostomal area of gena as
in female (14 above); (39) scutum as in
female (16 above); (40) lateral surface of
mesepisternum closely but distinctly punc-
tured, pubescence as in female (17 above);
(41) scutellum as in female (18 above);
(44) front coxal spine 1.5 times longer
than wide; (45) terga 3-5 sometimes lack-
ing apical fasciae medially or entirely, ca-
rina of tergum 6 with irregular denticu-
lations (Fig. 12E, F); (46) sterna 1-4
evenly punctate with squamose setae fill-
ing most or only a few punctures, apically
with irregular fasciae 0.25-0.3 mm (maxi-
mum) wide; (47) sternum 6 with lateral
setae resembling C. alter nata (Fig. 13H).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys apacheonim ranges
from southwestern Canada to California
and east to South Dakota, Colorado, and
to the edge of the Edwards Plateau in
Texas (Fig. 17).

Extremes of the flight period are April
10 (Abilene, Texas) and September 23
(Sequoia National Park, California) with
the majority of specimens taken from June
to August.

HABITAT. In its northern range,
Coelioxys apacheonim is associated with
areas classified as a variety of western
coniferous forests, and one specimen was
taken at the edge of alpine meadow west

668

The University of Kansas Science Bulletin

Figure 17. Distributions of Coelioxys apacheonini (A) and C. crysimi. (o).

of Boulder, Colorado. Further south the
bee spreads into areas classified as western
shrub and grassland and forest-grassland
in Texas.

GEOGRAPHIC VARIATION. As

with Coelioxys altemata, variation in C.
apacheorum does not seem to be related
to geography. Color of the legs varies

Taxonomy of Five Nearctic Subgenera of Coelioxys

669

from black to ferruginous with an inter-
mediate condition in which the tibia is
dark medially with ferruginous proximal
antl distal areas. Bees exhibiting all de-
crees of melanism occur throutihout the

range.

The decision to synonymize Coelioxys
qiiercina and C. apacheonim is based
u|i()n morphological agreement of the
darker males (earlier associated with the
females of C. apacheonim^ with the
liu:hter colored males which resemble the
type of C. qiiercina.

COMPARATIVE COMMENTS.
Specimens of Coelioxys apacheonim are
usually smaller than the other species of
Synocoelioxys. Dark males of C. apach-
eonim most closely resemble those of C.
erysimi but have a fascia of squamose
setae on the posterior surface of the
scutellum. Dark females of C. apache-
onim resemble females of C. erysimi and
dark females of C. alternata but differ by
the narrowly to moderately rounded sixth
sterna (compare Fig. 14B and D).

Coelioxys erysimi Cockerell
Figures 14D, 17

Coelioxys erysimi Cockerell, 1912, Can.
Entomol. 44:166 {$ Colorado, type in
the American Museum of Natural His-
tory).

FEMALE. Agrees with the figures
and description of Coelioxys texana except
as follows: (1) Length 12-14 mm; (2) in-
tegument black, sometimes piceous on
tibiae; (3) pubescence white, sometimes
brownish below antennae, pale brown on
apices of tergum 6 and sternum 6; (4)
ocular hairs of medium length (about 0.08
mm); (5) clypeus with fringe of setae
shorter than shown in Figure 11; (6)
clypeoantennal distance shorter than lat-
eral margin of clypeus; (7) lateral quarter
of paraocular area hidden by setae, erect
setae about 0.3 mm long; (8) rest of face
with an indistinct impunctate median line;

(9) interantennal distance slightly greater
than clypeoantennal distance which is
greater than antennocular distance; (11)
interocellar distance slightly greater than
ocelloccipital distance and less than ocello-
cular distance; (14) hypostomal area of
gena with setae as long or longer than on
disc; (15) mandible with apical tooth
longer than shown in Figure llA, more
like Figure IIC; (16) fasciae of scutum
not as distinct as shown in Figure 10; (17)
mesepisternum with lateral surface closely
punctate, many more punctures than mi-
nute punctures, pubescence long (about
0.35 mm) and slender; (18) scutellum
with posterior margin of dorsal surface
broadly rounded; (19) axilla shorter than
shown in Figure 10, longer than shown in
Figure 14E; (20) metanotum with small
amount of setae appressed medially; (21)
front coxal spine with anterior surface 1.5
times as long as wide, small but conspicu-
ous; (22) terga 1-5 with gradular fasciae
inconspicuous; (23) sterna 1-5 with irreg-
ular fasciae.

MALE. Agrees with the figures and
description of Coelioxys texana except as
follows: (24) Length 10-12.5 mm; (25)
integument as in female (2 above); (27)
ocular hairs as in female (4 above); (28)
clypeal margin as in female (5 above);
(29) clypeoantennal distance as in female
(6 above); (30) paraocular area hidden by
setae about 0.4 mm long; (34) interocellar
distance equal to or less than ocellocular
distance, less than ocelloccipital distance;

(35) mandible as in female (15 above);

(36) scutal fasciae indistinct or lacking;
(40-42) mesepisternum, scutellum, axilla as
in female (17-19 above); (43) metanotum
with little or no appressed setae medially;
(45) tergum 2 with median portion of
gradular groove bare, carina of tergum 6
denticulate, apical spine of tergum 7
shorter than shown in Figure 13A; (46)
setae filling punctures of sterna 1-4 not
squamose.

670

The University of Kansas Science Bulletin

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys erysimi is probably
restricted to the higher elevations o£ the
western United States and Canada (Fig.
17). Judging from the small number of
specimens examined, this bee appears to
be rare throughout its distribution.

This bee has been taken between June
21 and September 9.

HABITAT. Coelioxys erysimi has
been taken from areas classified as grand
fir-Douglas fir (Abies-Pseitdotsiiga) and
pine-Douglas fir (Pirius-Pseudotsitga) for-
ests, from juniper-pinyon and juniper-
steppe woodland {Jiuiipenis, Finns, Ar-
temisia, Agropyron), from Montane chap-
arral (ArctostaphyJos, Castanopsis, Cea-
nothiis), and from Great Basin sagebrush
and sagebrush steppe {Artemisia, Agropy-
ron ) vegetation types.

GEOGRAPHIC VARIATION. Vari-
tion within Coelioxys erysimi does not
seem to be related to geography.

COMPARATIVE COMMENTS.
Coelioxys erysimi is variable in size but
seems to be invariably dark (black or
piceous) throughout. Females most closely
resemble dark females of C. apacheorum
but the sixth sternum is as broad as that
of C. texana (compare Figs. 10 and 14D).
Males of C. erysimi are distinguished from
dark males of C. apacheorum by the lack
of a fascia across the posterior area of the
scutellum.

Coelioxys hnnteri Crawford
Figures IIC, 12G, 14C, 18

Coelioxys hnnteri Crawford, 1914, Ann.
Entomol. Soc. Amer. 7:151 ( 9 , Texas,
type in the National Museum of Nat-
ral History); Mitchell, 1962, North
Carolina Agr. Exp. Sta. Tech. Bull.
152:193, 209 (Key, redescription).

FEMALE. Agrees with figures and
description of Coelioxys texana except as
follows: (1) Length 10.5-13.5 mm; (2)
integument with scape, legs and metasoma

darker than ferruginous, scape and meta-
some sometimes black; (5) clypeus medi-
ally with slightly raised impunctate line,
apically with 0.1 mm fringe of slender
setae barely obscuring margin, medio-
apically with 5-10 setae 0.2-0.3 mm long;
(6) clypeoantennal distance shorter than
lateral margin of clypeus; (7) lateral third
of paraocular area hidden by setae, medial
area visible between erect setae; (8) rest of
face closely punctate except for slightly
raised median line; (11) interocellar and
ocelloccipital distances equal, greater than
ocellocular distance; (15) mandible more
slender than Figure llA (Fig. IIC);
(17) lateral surface of mesepisternum with
about as many or more punctures than
minute punctures; (21) front coxal spine
rounded in frontal view; (22) tergum 6
narrowly rounded apically (Fig. 14C) ;
(23) sternum 6 acute apically (Fig. 14C).

MALE. Agrees with figures and de-
scription of Coelioxys texana except as fol-
lows: (24) Length 10-12.5 mm; (25) in-
tegument as in female (2 above); (29)
clypeoantennal distance as in female (5
above); (30) paraocular area hidden ex-
cept for small inner area by setae 0.4 mm
long; (34) interocellar distance less than
ocellocular distance which is less than
ocelloccipital distance; (37) hypostomal
area of gena with setae much shorter than
genal setae; (38) mandible as in female
(15 above); (40) mesepisternum as in fe-
male (17 above); (45) carina of tergum 6
crenulate or more commonly denticulate,
asymmetrical (Fig. 12G) tergum 7 with
apical projection slightly more attenuate
than shown in Figure 13A.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys hnnteri has a range
similar to that of C. texana except C.
hnnteri has been taken further north (Fig.
18). As with C. texana, the longest series
has been taken in the vicinity of the
Southwestern Research Station five miles
west of Portal, Arizona.

Taxonomy of Five Nearctic Subgenera of Coelioxys

671

Figure 18. Distribution of Coelioxys hitnteri.

Coelioxys hunteri has been taken from
March 14 (Florida) to September LS
(Oregon).

HABITAT. Coelioxys hunteri extends
through a wide variety of vegetation types
in the United States. In the West, it is
associated with areas classified as fir-hem-
lock (AbieS'Tsiiga), pine-spruce (Pinus-
Picea), juniper-pinyon (Jiiniperus-Piniis)
forests as well as western shrub (Atriplex-
Sarcobatiis) and shrub-grassland (Boiite-
loiia, Hilaria, Larrea) vegetation types. In
the central United States, C. hitnteri has
been collected in shinnery (Qitercits-An-
dropogon), bluestem prairie ( Andropogon,
Panicitm, Sorghastnim), and oak-hickory
(Oiierciis-Carya) vegetation types. Further
east, it extends from the northern hard-
wood forests down to marl-everglades
{Mariscits and Per sea, Taxodiiim), sub-
tropical-pine (Piniss-Tetrazygia), and
mangrove ( Aviceunia-Rhizophora) vegeta-
tion types.

GEOGRAPHIC VARIATION. Size,
using head width of males as an index,
was found to vary geographically in Coeli-
oxys hunteri (Females were not used in
this study because sample sizes from the
eastern United States were too small to
yield meaningful results). Head widths
of nine males from Florida (3.64 ± 0.133
mm) were significantly larger than those
of 15 males from the rest of the eastern
United States (3.48 ± 0.130 mm) or 109
males from Southern California, Arizona,
New Mexico, western Texas and Mexico
(3.38 ±0.144).

COMPARATIVE COMMENTS.

Coelioxys hunteri, a medium to large bee,
most closely resembles C. texana but dif-
fers in the males by the bare appearance
of the hypostomal area and in the females
by the apically very narrowly rounded
sixth sternum. Also in the males of C.
hunteri, the lateral ocellus is subequally
distant from the eye and occipital margin,

672

The University of Kansas Science Bulletin

whereas in C. texana the lateral ocellus is
closer to the eve than to the occipital

margni.

Subgenus Schizocoelioxys Mitchell

Schizocoelioxys Mitchell, 1973. A subgeneric
revision of the genus Coelioxys of the
Western Hemisphere. Cont. Dept.
Entomol. North Carolina State Univ. p.
50. Type species: Coelioxys jmieraria
Smith.

Coelioxys fitneruria is the lone New
World representative of an Old World group
of bees (Mitchell 1973). The lateral widen-
ing of the apical fascia on the first metasomal
tergum as well as the medial obliteration of
the metasomal graduli is characteristic of
many Old World Coelioxys spp. Host rec-
ords of C. funeraria fall into four subgenera
of Megachile (Table 1). Two host subgenera,
Eutricharaea and Megachile, are Holarctic in
distribution (the first by recent introduc-
tion); and the other two host subgenera.
Delomegachile and Xauthosams, are closely
resembled by two Palearctic groups ( Mitchell,
personal communication). One of the five
known hosts of C. juneraria has had only C.
jmieraria reported as a parasite: M. inermis.
The other four bees are exploited by one to
three other species of Coelioxys. This bee
seems to be a more generalized parasite than
the endemic groups of North American
Coelioxys such as Boreocoelioxys and Syno-
coelioxys. The host range and resemblance
of C. juneraria to Palearctic species suggest
that this bee may be a fairly late arrival on
the North American scene.

Schizocoelioxys is distinguished from
other subgenera of Coelioxys by the follow-
ing list of characters. The absence of fasciae
on the scutum and the hump on the inner
surface of the mandible (Fig. 22A) are
found in no other species of North American
Coelioxys. Italicized characters most clearly
differentiate Schizocoelioxys from other sub-
genera in North America.

A. Ocellar area closely punctured.

B. Preoccipital carina incomplete.

C. Mandible with hump on inner surjace
(Fig.22A).

D. Prothoracic tubercle with carina mod-
erately produced.

E. Scutum moderately punctured, ladl-
ing jasciae.

F. Scutellum moderately punctured (Fig.
23A).

G. Axilla short (Fig. 23 A).

H. Mesepisternum moderately punctured
with a few minute punctures in female, con-
tiguously punctured in male.

I. Metasomal terga with graduli incom-
plete medially.

]. Female: jore coxal spine carinate,
blending with medioanterior surjace oj coxa.

K. Female: sixth metasomal sternum
elongate and notched (Fig. 24 A).

L. Male: hypostomal area of gena with
excavation bare, shining (Fig. 4).

M. Male: conspicuous jovea on meta-
somal tergum two.

N. Male: spine reduced but present on
metasomal tergum five.

O. Male: sixth metasomal tergum with
conspicuous spines (Fig. 25 A).

P. Male: apex of tergum six rounded
apically (Fig. 19 A).

Q. Male: sternum four emarginate api-
cally (Fig. 19B).

R. Male: sternum seven represented by
two small sclerites (Fig. 19D).

Figure 1'». IluUlcn .sclerites of male of Coelioxys
jiiiuiaria. A, Seventh tergum; B, Margin of
fourth sternum, fifth sternum; C, Sixth sternum;
D, Seventh sternum (represented by two scle-
rites); E, Eighth sternum; F, Genital armature.

Coelioxys juneraria Smith
Figures 19A-F, 20, 22A, 25A-B

Coelioxys juneraria Smith, 1854, Cata-
logue of British Hymenoptera in the
collection of the British Museum I:
272 ( (5 , Canada, type in the British
Museum); Mitchell, 1962, North Caro-
lina Agr. Exp. Sta. Tech. Bull. 152:
195, 203 (Key, redescription, = C. lat-
eralis Cresson, = C. liter osa Cresson);

Taxonomy of Five Nharctic Subgenera of Coelioxys

673

Hohbs, 1968, Can. Entomol. 100: 783
(Host).

Coelioxys lateralis Cresson, 1864, Proc.
Entomol. Soc. Philadelphia 2:405 (6,
Pennsylvania, type # 2498 in iVcademy
of Natural Sciences of Philadelphia).

Coelioxys lucrosa Cresson, 1878, Trans.
Amer. Entomol. Soc. 7:97 (9, New
York, Colorado, type # 2496 in Acad-
emy of Natural Sciences of Philadel-
phia); Graenicher, 1905, Bull. Natur.
Hist. Soc. Wisconsin 3:160 (Biology);
Graenicher, 1927, Entomol. News 38:
233 (Biology); Graenicher, 1935, Ann.
Entomol. Soc. Amer. 28:300, 304 (Biol-
ogy); Medler, 1958, Can. Entomol. 90:
326-327 (Biology); Medler and Koer-
ber, 1958, Ann. Entomol. Soc. Amer.
51:343 (Biology).

Coelioxys /iicl{si Cockerel!, 1934, Amer.
Mus. Novitates 732:3 (9, Colorado,
type in the American Museum of Nat-
ural History) NEW SYNONYMY.

FEMALE. (1) Length 1246 mm; (2)
integument finely rugulose between punc-
tures, dull on propodeum; black to very
dark brown; (3) pubescence white to light
golden brown, golden on tarsi; (4) ocular
hairs of medium length (about 0.05 mm);
(5) clypeus visible medially, slightly con-
vex, closely punctured with small (about
0.018 mm) punctures; margin gently out-
curved, with 5 small denticulations (Fig.
22A); (6) clypeoantennal distance less
than lateral margin of clypeus; (7) para-
ocular area visible between 0.35 mm long
setae, closely punctured; (8) rest of face
closely punctured up to ocellar area except
for longitudinal median area indented by
a longitudinal median line; (9) inter-
antennal distance greater than clypeoan-
tennal distance which is greater than an-
tennocular distance; (10) ocellar area slop-
ing, coarseh' punctured; (11) interocellar
distance greater than ocelloccipital dis-
tance, less than ocellocular distance; (12)
vertex with punctures slightly larger and

slightly more crowded than shown in Fig-
ure 10; (13) gena subequal in width to
eye, tapering dorsally and ventrally; (14)
hypostomal area of gena with setae
slightly longer than on gena; (15) mandi-
ble with a distinct hump on inner surface
(Fig. 22A); (16) scutum with incon-
spicuous, slender setae; median line ex-
tending slightly more than half the dis-
tance from anterior margin; disc less
closely punctured than periphery; (17)
lateral surface of mesepisternum punctured
like scutum in Figure 10, but with a few
minute punctures among punctures; (18)
scutellum with posterior margin of dorsal
surface broadly rounded in dorsal view
(Fig. 23A); (19) axilla reduced in length
(Fig. 23A); (20) metanotum with incon-
spicuous, long, slender setae; (21) front
coxal spine a thickened carina blending
with medioanterior surface of coxa; (22)
terga 1-5 with graduli incomplete medi-
ally, with apical fasciae wider laterally;
graduli bare; tergum 6 elongate with
raised, impunctate median line (Fig.
24A); (23) sterna 1-5 fasciate apically; 1-4
more closely punctured laterally; sternum
5 very closely and minutely punctured api-
cally; sternum 6 minutely and closely
punctured with elongate punctures, elon-
gate, notched subapically (Fig. 24A).

MALE. (24) Length 942 mm; (25)
integument as in female (2 above); (26)
pubescence as in female (3 above); (27)
ocular hairs of medium length (about 0.08
mm) ; (28) clypeus and clypeal margin
hidden by setae 0.5 mm long, conforma-
tion as in female (5 above) ; (29) clypeo-
antennal distance shorter than lateral mar-
gin of clypeus; (30) paraocular area al-
most hidden by setae about 0.45 mm long,
closely punctured; (31) rest of face visible
through setae 0.5 mm long, punctures con-
tiguous to ocellar area; (32) clypeoanten-
nal distance less than interantennal dis-
tance and greater than antennocular dis-
tance; (33) ocellar area contiguously

674

The University of Kansas Science Bulletin

punctured; (34) ocelloccipital distance less
than ocellocular distance and subequal to
or greater than interocellar distance; (35)
vertex coarsely but contiguously punc-
tured; (36) gena slightly narrower than
eye (Fig. 4); (37) hypostomal area bare,
separated from gena by carina (Fig. 4);

(38) mandible as in female (15 above);

(39) scutum as in female (16 above); (40)
mesepisternum contiguously punctured;
(41) scutellum as in female (18 above);
(43"*' axilla slightly more produced than
shown in Figure 23A; (43) metanotum as
in female (20 above); (44) front coxal
spine robust, about twice as long as broad;

(45) terga 1-6 fasciate lateroapically; gra-
duli incomplete medially, mosdy bare;
tergum 2 with conspicuous foveae (Fig.
25B) ; tergum 6 with conspicuous dorsal,
ventral, and lateral spines (Fig. 25A);
tergum 7 rounded apically (Fig. 19A);

(46) sterna 1-4 fasciate apically, evenly
punctured; (47) sternum 4 submarginally
carinate, weakly bidentate apically (Fig.
19B); sternum 5 incurved apically (Fig.
19B) ; sternum 6 with abundant, long setae
(Fig. 19C), sternum 7 represented by 2
small sclerites (Fig. 19D) ; sternum 8
broadly rounded apically (Fig. 19E) ; (47)
genital armature with long, copious setae
on gonocoxites (Fig. 19F).

HOST RECORDS. Coelioxys fnner-
aria has been reported from more hosts
than any other North American Coelioxys.
Graenicher (1927) reported C. jiineraria
to be a parasite of Megachile latimaniis.
Medler (1958) reported C. jttneraria from
the nest of M. inermis, and Medler and
Koerber (1958) reported C. jnneraria
from the nest of M. relativa. Mitchell
(1962) listed M. jrigida as a host of C.
jnneraria, and Hobbs (1968) found a fifth
host species, M. rotitndata.

DISTRIBUTION AND SEASON OF
FLIGHT. The distribution of Coelioxys
funeraria extends from above the arctic
circle to southern New Mexico (Fig. 20).

Such a broad distribution, however, is no
more remarkable than the distribution of
one of the host bees, Megachile jrigida,
which raniies from Alaska to Arizona.

Season of flight lasts at least from May
26 (Ottawa, Ontario) to September 18
(Banff, Alberta) with most specimens
taken from mid-June through August.

HABITAT. In the northern portion
of its range, Coelioxys jnneraria is associ-
ated with areas designated as various co-
niferous forests, tundra-coniferous forest,
northeastern oak-pine forests, northern
floodplain forests, Appalachian oak forest,
northern grassland, oak savana, and fescue-
wheatgrass vegetation types. Further south
C. jnneraria is associated only with vege-
tation types which occur at higher eleva-
tions such as alpine meadow barrens in
Colorado and southwestern spruce-fir for-
est in New Mexico. In California (Gil-
roy), C. jnneraria was taken near an area
classified as the ecotone between California
oakwoods and fescue-oatgrass vegetation
types.

GEOGRAPHIC VARIATION. Al-
though there is considerable variation in
size in Coelioxys jnneraria, larger and
smaller specimens occur throughout the
range. One explanation for such variation
may be that its size is an expression of
host size. The large bees may utilize the
nests of Megachile inermis (15-22 mm).
Medium sized C. jnneraria may utilize the
nests of M. jrigida (12-15 mm) and M.
latimanns (13-14 mm), and the small C.
jnneraria may utilize the nests of M. rela-
tiva (9-12 mm) and M. rotnndata (8-9
mm). The distributions of the host bees
overlap over thousands of square miles,
and perhaps this overlap accounts for the
considerable variation within and between
populations of C. jnneraria from any one
area.

COMPARATIVE COMMENTS.
Coelioxys jnneraria is most easily confused
with the usually smaller C. moesta, but it

Taxonomy of Five Nkarctic Subgenera of Coelioxys

675

160 150 140 130 120 110 100 90 80 70 60 50

Figure 20. Distribution of Coelioxys jtmeraria.

is distinguished from C. moesta males by
the foveae of the second metasomal ter-
gum which are large and almost open
along the anterior margin in C. juneraria
(Fig. 25B) and small and enclosed in C
moesta (Fig. 25E). The female of C.
juneraria possesses a convex clypeus with
a slightly outcurved margin (Fig. 22A)
whereas the female of C. moesta possesses
a flat clypeus with a triangularly produced
margin (as in Fig. 22D).

Subgenus Boreocoelioxys Mitchell

Boreocoelioxys Mitchell, 1973. A subgeneric
revision of the genus Coelioxys of the
Western Hemisphere. Cent. Dept.
Entomol. North Carolina State Univ. p.
37. Type species: Coelioxys rufiic/rsis
Smith.

Other bees in the subgenus Boreocoelioxys
are C. banl^si, C. insita, C. moesta, C. novo-
mexicana, C. oaxacana, C. octodentata , C.
porterae, C. pratti , and C. sayi. These bees
are usually black or dark brown except for
the legs which may vary from black to bright
ferruginous.

Host-parasite relationships are not clear-
cut for Boreocoelioxys. The large number of
hosts utilized probably reflects a generalized
response of female Coelioxys to nests of
Megachile occurring within the habitat. Spe-
cies from the group have been reared from
the following subgenera of Megachile: Delo-
megachile, Eiitricltaraea, Litoinegachile, Me-
gachile, Psendocentron, and Xanthosarus
(Table 1).

Most of the characters in the list below
are found in other subgenera of New World
Coelioxys and even in some of the Old
World Liothyrapis, but never in this combi-
nation. This group is the only North Amer-
ican group with complete graduli on meta-

676

The University of Kansas Science Bulletin

Figure 21. Dorsal view of female Cotlioxys iiifilarsis, left half showing punctation, right half with
melanization and pubescence added.

Taxonomy oi- Five Nearctic Subgenera of Codioxys

677

somal terga two and three and notched sixth
metasomal sterna in females or fo\eae on
metasomal tergum two in males. Italicized
characters most clearly differentiate Boiro-
coclioxys from other subgenera in North
America.

A. Ocellar area closely punctured (Fig.

B. Preoccipital carina incomplete medi-
ally.

C. Mandible with outer surface broadly
rounded to humped (Fig. 22B-G).

D. Prothoracic tubercle with carina usu-
ally produced into thin, plate-like structure.

E. Scutum closely punctured, sometimes
with conspicuous fasciae (Fig. 21).

F. Scutellum contiguously to moderately
punctured (Fig. 23B-D).

G. Axillae well produced (Fig. 21,
23B-D).

B C

Figure 22. Left fialf of clypeus and mandible of
females of Coelioxys. A, C. jiineraria; B, itifi-
tarsis; C, Lateral view of clypeus of C. haiil{si;
D, C. i>an/(si; E, C. novotnexicana; F, C. portcrae;
G, C. sayi.

H. Mesepisternum with lateral surface
closely punctured.

I. Gniduli complete on metasomal terga
tivo and three (Figs. 21, 25B).

J. Female: fore coxal spine small, in-
conspicuous.

K. Female: sixth metasomal sternum
elongate and notched (Figs. 21, 24B-L).

L. Male: hypostomal area of gena with
excavation bare, shining.

M. Male: conspicuous joveae on meta-
somal tergum two (Fig. 25B).

N. Male: fifth metasomal tergum with
lateral spine (Fig. 25B).

O. Male: sixth metasomal tergum with
conspicuous spines (Fig. 25B).

P. Male: apex of metasomal tergum
seven rounded (Fig. 26A).

Q. Male: apex of metasomal sternum
four conspicuously emarginate (Fig. 26B).

R. Male: sternum seven represented by
two small sclerites (Fig. 26D).

Figure 23. Scutellum and axillae of females of
Coelioxys. A, C. ftineraria; B, C. nioesta; C, C.
oaxacana.

Key to Females of Boreocoelioxys

1. Clypeal margin with distinct emargina-

tion evident in frontal view with full

compliment of setae 4

Clypeal margin otherwise 2

2. Clypeal margin outcurved 3

Clypeal margin straight 5

3. Clypeus concave in profile ban\si

Clypeus straight in profile moesta

4. Clypeal margin with dense fringe of

white setae almost perpendicular to

surface sayi

Clypeal margin with dense fringe of yel-
lowish setae almost parallel to surface
novomexicana

Figure 24. Sixth metasomal tergum and sternum of
females of Coelioxys. A, C. jitncraria; B, C.
hanksi; C, C. moesta; D-L C. octodentata show-
ing \ariation in lateral angles of tergum, apex of
sternum; J, C. porterae, K, C. oaxacana.

678

The University of Kansas Science Bulletin

■^ O GO

D O O o

o°'

.00

„ ooooooo9,'Vi
r^OooooO ocuqO o

,-,00 o

o o

o ° o 000 OOo,

O o

o°o°oO 0_

0 00„o =

' O O

lO^O"© 000* •='<'%, 00 O ,-^

Figure 25. Various parts of males of Coelioxys. A-B, C. jnncniria. A, Sixth metasomal tergum; B,
Fovea. C, Metasoma, C. riifitarsis (arrow indicates median fascia). D-H, Foveal areas. D, C. batilist;
E. C. mocsta; F, C. porterae: G, C. octodentata; H, C. gnndcUae.

5. Scutellum with median longitudinal

ridge oaxacana

Scutellum without median longitudinal
ridge 6

6. Tergum six in dorsal view angulate as in

Figure 21 7

Tergum six in dorsal view less angulate
than shown in Figure 24 J 8

7. Ocular hairs long (about 0.15 mm); ter-

gum six with tiny setae ventral to an-
gles which have a velvety appearance;
tergum one with median fascia (see
arrow on Fig. 25C) of long, slender,

erect setae rufitarsis

Ocular hairs short (about 0.05 mm);
tergum six with squamose white setae
ventral to angles; tergum one with
median fascia of prostrate setae, if slen-
der than short octodentata*

8. Conspicuous fasciae laterally and ante-

riorly around scutum; ocular hairs long

(about 0.15 mm) pratti

Scutum with no conspicuous fasciae;
ocular hairs variable 9

9. Sterna one to five ferruginous apically.

octodentata*

Sterna one to five very dark to black
apically 10

10. Legs black or very dark brown .. porterae
Legs brownish to ferruginous 11

11. Tergum six broadly rounded apically

(Fig. 24E-J) octodentata*

Tergum six narrow apically (as in Fig.
24B) insita

* Occurs more than once in key.

Key to Males of Boreocoelioxys

1. Fovea on metasomal tergum two wide,

deep, short, conspicuous; margins of
fovea impunctate (Fig. 25C); bee ro-
bust riifitarsis

Fovea on metasomal tergum two vari-
able, if wide, then anterior margin
punctate; bee variable in form 2

2. Fovea on metasomal tergum two shallow

(as in Fig. 25D) or with anterior mar-
gin lower than posterior margin (as in
Fig. 25F) and sometimes closely punc-
tate 3

Fovea on metasomal tergum two deep
(as in Fig. 25G) with margins the

Taxonomy of Fivu Nearctic Subgenera of Coelioxys

679

Figure 26. Hidden sclerites of male of Coelioxys nifitarsis. A, Seventh tergum; B, Margin of fourth
sternum, fifth sternum; C, Sixth sternum; D, Seventh sternum (represented by two sclerites); E.
Eighth sternum F, Genital armature.

same height, anterior margin may be
moderately punctate 4

3. Scutellum with dorsal and posterior sur-

faces separated medially by a carina,
punctures deep and contiguous; legs
usually with ferruginous areas; fovea
on metasomal tergum two complete or
with anterior margin weak (as in Fig.

25D banf{si, insita

Scutellum convex, punctures contiguous
but so broad that edges become minute
carinae above the flat surfaces of the
puncture floors; legs usually black or
blackish-brown; fovea on metasomal
tergum two with anterior margin
weaker than posterior margin (Fig.
25F) porterae

4. Fovea on metasomal tergum two incon-

spicuous, sometimes shallow (Fig.
25E); integument usually black; me-
dium to small, slender bee moesta

Fovea on metasomal two deep and con-
spicuous (as in Fig. 25G) with small,
contiguous punctures immediately pos-

terior to it; integument variable in
color; medium to large bee 5

5. Posterior margins of gradular grooves on

metasomal terga two and three almost

obliterated medially sayi

Posterior margins of gradular grooves on
metasomal terga two and three distinct
medially 6

6. Ocular hairs long (about 0.15 mm);

punctation of vertex with conspicuous
interspaces; scutum with conspicuous
anterolateral fasciae; posterior surface
of scutellum with conspicuous fascia

pratti

Ocular hairs short (about 0.06 mm);
punctation of vertex variable; scutum
laterally and scutellum posteriorly with-
out conspicuous fasciae

novomexicana, octodentata

Coelioxys nifitarsis Smith
Figures 21, 22B, 25C, 26A-F, 27, 28

Coelioxys nifitarsis Smith, 1854, Catalogue

680

The University of Kansas Science Bulletin

of Hymenoptera in the British Mu-
seum 2:271 ( $ , United States, type in
the British Museum); Graenicher, 1905,
Bull. Wisconsin Natur. Hist. Soc. 3:
162-163 (Host) ; Graenicher, 1906, Bull.
Wisconsin Natur. Hist. Soc. 4:138 (Bi-
ology); Robertson, 1926, Psyche 3i:
116 (Host); Hicks, 1926, Univ. Colo-
rado Stud. 15:230 (Host); Graenicher,
1927, Entomol. News 38:233 (Biol-
ogy) ; Robertson, 1929, Flowers and In-
sects, Lancaster Pennsylvania Sci. Press:
9-142 (Flower Records) ; Medler and
Lussenhop. 1968, Univ. Wisconsin Re-
search Bull. 274:57 (Key, diagnosis,
biology).

Coelioxys dubitata Smith, 1854, Cat. Hy-
menoptera Brit. Mus. 2:272 ( 9 , Flor-
ida, type in the British Museum) ;
Robertson, 1897, Trans. Acad. Sci. St.
Louis 7:345 (=: nifitarsls); Graenicher,
1935, Ann. Entomol. Soc. Amer. 28:
300, 304 (Distribution, host).

Coelioxys comstocI{ii Cresson, 1878. Trans.
Amer. Entomol. Soc. 7:96 (9, New
York, type # 2497 in the Academy of
Natural Sciences of Philadelphia)
NEW SYNONYMY.

Coelioxys coloradensis Cresson, 1878,
Trans. Amer. Entomol. Soc. 7:98 ( c5 ,
Colorado, type # 2500 in the Academy
of Natural Sciences of Philadelphia)
NEW SYNONYMY; Hicks, 1926,
Univ. Colorado Stud. 15:233 (Host);
Graenicher, 1935, Ann. Entomol. Soc.
Amer. 28:300 (Host).

Coelioxys riifitarsis rhois Cockerell, 1903,
Ann. Mag. Natur. Hist. 12:452 (9,
New Mexico, type in the American
Museum of Natural History).

Coelioxys riifitarsis daripes Cockerell,
1925, Pan-Pacific Entomol. 1:146, 150
( $ , California, type # 1633 in the Cal-
ifornia Academy of Sciences).

Coelioxys riifitarsis riifitarsis; Mitchell,
1962, North Carolina Agr. Exp. Sta.

Tech. Bull. 152:193-194, 220 (Key, re-
description).

FEMALE. (1) Length 11-13 mm; (2)
integument granular on vertex between
punctures and on thorax; black, dark
brown on ventral surface of tergum 1, legs
distal to coxae black to ferruginous; (3)
pubescence white, golden on tarsi; (4)
ocular hairs long (about 0.15 mm) ; (5)
clypeus uniformly covered with short, ap-
pressed setae, surface visible, margin ir-
regular (Fig. 22B); (6) clypeoantennal
distance shorter than lateral margin of
clypeus; (7) paraocular area hidden by
setae about 0.25 mm long; (8) rest of face
rugose, closely punctured up to ocellar
area except for small median impunctate
ridge between antennae; (9) clypeoanten-
nal distance less than interantennal and
greater than antennocular distances; (10)
ocellar area closely punctate; (11) inter-
ocellar distance greater than ocelloccipital
and less than ocellocular distances; (12)
vertex moderately punctate (Fig. 21);
(13) gena narrower than eye, integument
visible, ventral angle about 90°; (14) hy-
postomal area of gena with setae longer
than on disc; (15) mandibles short, broad
(Fig. 22B); (16) scutum with slender,
erect hair on disc (Fig. 21); (17) lateral
surface of mesepisternum closely punctate
with long (about 0.35 mm), slender setae
which usually form anterior and posterior
fasciae; (18) scutellum with posterior
margin of dorsal surface rounded (Fig.
21); (19) axilla well produced (Fig. 21);
(20) metanotal setae erect except for 1-3
or 4 prostrate in center; (21) front coxal
spine subtriangular, inconspicuous; (22)
terga 2, 3 with conspicuous gradular
grooves, terga 1-5 with conspicuous apical
fasciae, tergum 6 angled with very small
setae at angles which resemble velvet in
dorsal view (Fig. 21); (23) sterna 1-5
fasciate apically, fasciae of sterna 2-4 usu-
ally interrupted medially, sterna 1-4 and
basal area of sternum 5 deeply and evenly

Taxonomy oi' Five Nearctic Subgenera of Coelioxys

681

punciuretl, sternum 5 apically very closely
and shallowly punctured, sternum 6 with
indistinct elongate punctures medially,
broadly lanceolate apically (Fig. 21).

MALE. (24) Length 9.5-12 mm; (25)
integument as in female (2 above), venter
of abdomen black to dark brown; (26)
pubescence white, white to yellow on face,
white to pale brown on thorax, golden on
tarsi; (27) ocular hairs long (about 0.15
mm) ; (28) clypeus hidden by copious
long (about 0.45 mm) setae, surface rugose
and shiny; clypeal margin slightly emargi-
nated with 5 small denticles; (29) clypeo-
antennal distance as in female (6 above) ;
(30) paraocular area hidden by long setae
(about 0.6 mm); (31) rest of face rugose
with median, short (0.25 mm), impunc-
tate line usually obscured by copious setae
up to 0.45 mm long; (32) interantennal,
clypeoantennal, antennocular distances as
in female (9 above); (33) ocellar area as
in female (10 above); (34) interocellar
and ocelloccipital distances equal, less than
ocellocular distance; (35) vertex as in fe-
male (12 above); (36) gena as in female
(13 above) but ventral angle greater than
90°; (37) hypostomal area of gena modi-
fied into distinct excavation (as in Fig. 4),
anterior portion with setae as in female
(14 above), posterior portion with integu-
ment clearly visible; (38) mandible, scu-
tum, mesepisternum, scutellum, axilla,
metanotal setae as in female (15-20 above) ;
(44) front coxal spine long (about 0.25
mm) and broad (about 0.2 mm) with
setae on dorsal and ventral surfaces,
rounded apically, directed anteriorly; (45)
terga with apical fasciae diminishing pos-
teriorly, gradular grooves of terga 4-6 with
fasciae, fovea on tergum 2 less than 0.05
mm long and 0.3-0.5 mm wide (Fig. 25C),
tergum 6 with dorsal spines stouter than
ventral spines (Fig. 25C); tergum 7 un-
modified (Fig. 26A); (46) sterna 1-4
evenly punctate with small slender seta
in each puncture, fasciate apically, sternum

1 fasciate medially; (47) sternum 4 sub-
margically carinate, apex emarginate to
bidentate, margin of sternum 5 slightly
emarginate (Fig. 26B), sternum 6 with
copious setae lateroapically (Fig. 26C),
sternum 7 with two sclerotized areas (Fig.
26D), sternum 8 with base broad (Fig.
26E), genital armature with abundant
setae on gonocoxite apically (Fig. 26F).

HOST RECORDS. Graenicher (1905)
reared Coelioxys nifitarsis from nests of
Megachile latimanits and M. melanophaea.
He later (1935) reported C. nifitarsis to
be a parasite of M. melanophaea woototii.
Hicks (1926) reared C. rufilarsis (= C
coloradensis) from a nest of M. monti-
vaga. Medler and Lussenhop (1968) re-
ported that D. H. Pengelly found C. nifi-
tarsis to be a parasite of M. texana. Mitch-
ell (1973) indicated M. jortis is another
host of C. nifitarsis. A new host record
from a label of C. nifitarsis from Alberta,
Canada, is M. perihirta.

DISTRIBUTION AND SEASON OF
FLIGHT. This bee extends from mid-
dle Canada well into Mexico (Fig. 27).
Although Smith (1854) described the fe-
male of Coelioxys nifitarsis from Florida,
specimens from south of the 35° parallel
are rare.

Coelioxys nifitarsis is in flight at least
from April 22 (Texas) and May 21
(Idaho) to October 3 (central California).

HABITAT. Coelioxys nifitarsis ranges
through areas which have been classified
as a variety of forest, savana, and open
grassland vegetation types. In the West
this bee has been taken from areas classi-
fied as forests containing such dominant
trees as spruce (Picea), cedars (Thuja Li-
brocednis), hemlock (Tsiiga), fir (Abies),
and pine (Piniis); from areas classified as
scrub and savana such as chaparral, moun-
tain mahogany-oak scrub and oak savana
(Cercocarpiis, Querciis, Andropogon), ju-
niper-oak savana (Jiiniperus, Querciis,
Andropogon), sagebrush (Salvia, Eriogo-

682

The University of Kansas Science Bulletin

Figure 27. Distribution of CoeUoxys nifitarsis.

num, Artemisia), and creosote bush (Lar-
rea); from areas classified as various grass-
lands containing grasses such as fescue
(Festitca), oatgrass (Danthonia), tule
(Scirpus), cattail (Typha), grama grass
(Boittelotia), and bluestem (Andropogon);
and from an area classified as alpine mea-
dows and barren.

In the East CoeUoxys nifitarsis has
been taken from areas classified as a vari-
ety of northern coniferous, hardwood and
mixed forests. In Florida it has been col-
lected in southern mixed forest.

GEOGRAPHIC VARIATION. As
with CoeUoxys alternata, leg color of C.
nifitarsis varies throughout its distribution

and seemingly without respect to geogra-
phy. However, an exceptionally large se-
ries of C. nifitarsis from California showed
possible geographic variation (Fig. 28).
The proportion of bees with dark legs was
greater for bees from higher elevations
than for bees from the San Joaquin and
Sacramento Valleys. However, bees with
both dark and ferruginous legs occurred
at high and low elevations.

Color of the pubescence of males
seemed to vary geographically. Some
males from the western portion of the
distribution of CoeUoxys nifitarsis possess
yellowish facial pubescence and slightly
brownish thoracic pubescence. This con-

Taxonomy of Five Nearctic Subgenera of Coelioxys

683

120

40

• : Femor a> Tibia piceous

%) z Femo,r piceous, Tibia ferruginous

\J - Femor & Tibia ferruginous

120

Figure 28. Distribution of Coelioxys rtifitarsis in California showing variation in leg color.

dition was especially evident in bees from
Zacapu, Michoacan, Mexico. However,
bees in this condition were also taken in
southwestern Canada and Ithaca, New
York. Perhaps the yellow or brown con-
dition of the pubescence is due to the
teneral conditions of these specimens.
Teneral adults of Megachile brevis and M.
mendica possess bright yellow (almost
orange-yellow in M. mendica) pubescence

whereas that of specimens from the field
is white.

The name Coelioxys comstochii is rep-
resented only by the type specimen in the
Academy of Natural Sciences of Phila-
delphia and seems to be a teratological
specimen. The metasoma appears to be
discolored; the metasoma is asymmetri-
cally punctured (some punctures appear
to be missing on one side) ; and the apex

684

The University of Kansas Science Bulletin

of tergum six appears to be broken. Pollen
on the sixth metasomal sternum indicates
the bee may have been worn before col-
lection. In other respects, however, the
specimen agrees with the description
above.

COMPARATIVE COMMENTS.
Coelioxys nifitarsis is most easily confused
with the larger western specimens of C.
octodentata but differs in the females by
the basal fascia on metasomal tergum one
which is composed of long, slender, erect
setae in C. nifitarsis (Fig. 21) and at least
partially composed of squamose setae in
C. octodentata. Males of C. rttfitarsis pos-
sess foveae (Fig. 25C) which are wider,
deeper, and shorter than those of C. ban\si
and C. insita (Fig. 25D) and are unlike
the more oval foveae of all other North
American Coelioxys which possess foveae
on the second tergum of the metasoma.

Coelioxys ban\si Crawford
Figures 22C-D, 24B, 25D, 29

Coelioxys ban\si Crawford, 1914, Ann.
Entomol. Soc. Amer. 7:155 (2, Vir-
ginia, type # 18221 in the National
Museum of Natural History); Medler
and Koerber, 1958, Ann. Entomol. Soc.
Amer. 51:343 (Biology); Mitchell, 1962,
North Carolina Agr. Exp. Sta. Tech.
Bull. 152:213 (= C moesta, misidenti-
fication).

Coelioxys anguUjera Cockerell, 1916, Po-
mona J. Entomol. Zool. 8:61 (9, <5 ,
California, type in the American Mu-
seum of Natural History) NEW
SYNONYMY.

FEMALE: Agrees with description
and figures of Coelioxys nifitarsis except
as follows: (1) Length 10-13 mm; (2) in-
tegument black, becoming brown to fer-
ruginous on tegula, legs distal to base of
femora, and metasomal venter; (4) ocular
hairs short (about 0.05 mm); (5) clypeal
margin outcurved and produced anteri-

orly, clypeal profile concave (Fig. 22C-D) ;
(7) paraocular area almost hidden by ap-
pressed setae about 0.15 mm long; (10)
ocellar area in some specimens with irreg-
ular, impunctate areas, usually narrower
than puncture width anterior to median
ocellus; (11) interocellar distance equal to
ocelloccipital distance, less than ocellocular
distance; (13) gena equal in width to light
portion of eye; (14) hypostomal area of
gena with setae as long as or slightly
longer than setae on lower disc; (16)
scutum with slender, erect setae and with
anterolateral fascia sometimes less con-
spicuous than shown in Figure 21; (17)
mesepisternum in some specimens with an
occasional minute puncture among large
punctures, usually closely punctured; (22)
terga 2-5 slightly more closely punctured
on postgradular area than shown in Fig.
21, gradular groove of tergum 3 less con-
spicuous than shown in Fig. 21, tergum 6
not angled laterally but produced posteri-
orly (Fig. 24B); (23) sternum 6 elongate,
apex lanceolate (Fig. 24B).

MALE: Agrees with figures and de-
scription of C. nifitarsis except as follows:
(24) Length 9.5-11.5 mm; (25) integu-
ment faintly rugulose on metasoma be-
tween punctures; tegula black to brown;

(26) pubescence as in female (3 above) ;

(27) ocular hairs of medium length (about
0.09 mm); (31) rest of face coarsely, con-
tiguously punctured, with long (0.3 mm),
slender setae; (33) ocellar area contigu-
ously punctured; (34) ocelloccipital dis-
tance greater than interocellar distance,
less than ocellocular distance; (37) hypo-
stomal area of gena with pubescence on
anterior portion shorter than on gena;
(39) scutum as in female (16 above); (43)
metanotum with erect, slender setae; (45)
fovea on tergum 2 about 0.06 mm long
and 0.3 mm wide (Fig. 25D); tergum 6
with depression between dorsal spines
slightly deeper and wider than shown in
Figure 25C.

Taxonomy of Five Nearctic Subgenera of Coelioxys

685

Figure 29. Distributions of Coelioxys bank^si and C. insita.

HOST RECORDS. Medler and
Koerber (1958) reported rearing Coelioxys
ban\si from nests of Megachile relativa.

DISTRIBUTION AND SEASON OF
FLIGHT. This species has been collected
throughout the western half and north-
eastern quarter of the United States (Fig-
ure 29). It appears to be rare throughout
its range.

Coelioxys banl{si is in flight from at
least April 16 (central California) to Sep-
tember 20 (Toronto, Ontario).

HABITAT. Coelioxys ban\si has
been taken mainly from areas classified as
western coniferous and northern hard-
wood forests. The western forests are
dominated by such trees as cedar (Tluija),
hemlock (Tsiiga), and Douglas fir {Pseit-

dotsitga), grand fir (Abies), pine (Pinus),
and spruce (Picea). Two areas, one in
California, classified as coastal sagebrush
(Salvia, Eriogoniim ) and the other in
Utah classified as mountain mahogany-
oak shrub ( Cercocarpits, Qiterciis), were
the only western areas not dominated by
conifers. In the East this bee was taken
from an area classified as mixed meso-
phytic forest (Acer, Aescitlns, Fagits, Liri-
odendron, Quercus, Tilia) and an area
classified as northern hardwoods forest
(Acer, Betula, Fagtts, Tsiiga). Perhaps C.
ban{si is excluded from the western, cen-
tral, and eastern grasslands by C. insita.

GEOGRAPHIC VARIATION. The
variation described above does not seem to
be related to geography. Coelioxys ban/{si

686

The University of Kansas Science Bulletin

shows a problem that is recurrent in New
World Coelloxys; this species and C. insita
(treated next) differ largely in the con-
figuration of the anterior margin of the
clypeus of the female (males are indistin-
guishable). In C. ban\si the clypeal mar-
gin is produced medially (Fig. 22C-D)
whereas the clypeal margin is almost
straight to slightly outcurved in C. insita.
There seem to be no intermediates (15
specimens of C. han\si and 20 specimens
of C. insita). Whether these alternate
conditions are the result of polymorphism
within one species or indicate the presence
of two species cannot be determined from
dead specimens. Unless it is demonstrated
that there is only one species, it seems best
to recognize two.

Similar pairs of species separated by the
anterior margin of the clypeus of the fe-
male (males indistinguishable) are found
again in this group {Coelioxys octodentata
and C. novomexicana. further discussion
below) and in a Neotropical group (C.
zapoteca and C. clypeata).

COMPARATIVE COMMENTS. Fe-
males of Coelioxys banJ{si differ from all
other females in this group by the anterior
margin of the clypeus which is produced
and raised medially, making the profile
distinctly concave (Fig. 22C-D). Males of
C. banhji differ from other males in this
group (except those of C. insita which are
distinguishable by distribution only) by
the almost linear shape of the fovea on the
second metasomal tergum (Fig. 25D).
The anterior margin of the fovea is usu-
ally very closely punctured.

Coelioxys insita Cresson
Figure 29

Coelioxys insita Cresson, 1872, Trans.
Amer. Entomol. Soc. 4 :273 ( $ , Texas,
type # 2499 in the Academy of Natu-
ral Sciences of Philadelphia).

Coelioxys ntdis Cockerell, 1934, Amer.
Mus. Novitates 732 :3-4 ( 9 , Colorado,

type in the American Museum of Nat-
ural History) NEW SYNONYMY.

FEMALE. Agrees with description
and figures of Coelioxys ban\si except as
follows: (5) clypeal margin almost
straight to slightly undulating or slightly
outcurved; clypeal profile straight.

MALE. Agrees with description and
figures of Coelioxys ban\si.

^ DISTRIBUTION AND SEASON OF
FLIGHT. This bee ranges from Ne-
braska to northern Mexico and to Arizona
(Fig. 29).

Coelioxys insita is in flight at least from
April 11 (Texas) to September 15 (Ari-
zona) and October 15 (Mexico).

HABITAT. The distribution of Coeli-
oxys insita extends westward to an area
classified as pine-Douglas fir forest (Pinits,
Pseiidotsitga) in Colorado and eastward
to an area classified as oak-hickory forest
{Oiterciis. Carya) in Texas. This bee has
also been taken from areas classified as
eastern ponderosa pine forest, oak-juniper
woodland (Jiiniperits), mesquite-oak sa-
vana (Prosopis, Quercus, Andropogon),
mesquite-buffalo grass (Buchloe), creosote
bush-tarbush (Larrea, Floitrensia), and
grama-tobosa-shrubsteppe (Boitteloiia, Hi-
laria) vegetation types. The remaining
areas are classified as western grassland
and central and eastern grassland contain-
ing such dominant grasses as grama, to-
bosa, buffalo grass (Buchloe), bluestem
(Andropogon), switchgrass (Panicum),
and Indian grass (Sorghastnim).

Coelioxys moesta Cresson , .
Figures' 23B, 24C, 25E, 30

Coelioxys moesta Cresson, 1864, Proc.
Entomol. Soc. Philadelphia 2:403 (9,
Connecticut, type # 2491 in the Acad-
emy of Natural Sciences of Philadel-
phia); Graenicher, 1927, Entomol.
News 38:233, 274 (Biology); Medler
and Koerber, 1958, Ann. Entomol. Soc.
Amer. 51:337, 343 (Biology); Medler,

Taxonomy of Five Nearctic Subgenera of Coelioxys

687

1959, Can. Entomol. 91:114-115 (Biol-
ogy); Bohart, 1970, Utah State Univ.
41st Faculty Honor Lecture, p. 9 (Im-
mature).

Coelioxys tristis Provancher, 18<S2, Natur.
Can. 13:241 (Lapsus for C. woesta).

Coelioxys liitzi Cockerell, 1921, Amer.
Mus. Novitates 21:5-6 ($, $, Utah,
Colorado, type in the American Mu-
seum of Natural History) NEW SYN-
ONYMY.

FEMALE. Agrees with description
and figures of Coelioxys nifitarsis except
as follows: (1) Length 10-12 mm; (2) in-
tegument black, tegula, legs and apex of
metasoma black to dark brown; (3) pu-
bescence white, scattered golden erect setae
on clypeus, golden setae on tarsi; (4) ocu-
lar hairs short (about 0.7 mm) ; (5) clypeus
covered by short appressed setae and erect
golden setae which are more numerous
near apical margin, margin broadly out-
curved, clypeal profile straight; (7) para-
ocular area almost obscured by setae about
0.2 mm long; (11) interocellar distance
slightly greater than ocelloccipital distance
and less than ocellocular distance; (14)
hypostomal area with setae subequal in
length to those on disc; (16) scutum with
almost no squamose setae; (17) setae of
mesepisternum usually not forming dis-
crete fasciae; (19) axilla short (Fig. 23B);
(20) metanotal setae erect, (22) terga 2, 3
with less conspicuous gradular grooves
than shown in Figure 21, gradular groove
of tergum 3 almost obliterated medially,
postgradular areas more closely punctured
than shown in Figure 21, fasciae of terga
4, 5 almost obliterated medially, tergum 6
not angled, elongate (Fig. 24C); (23)
sternum 6 elongate, lanceolate apically
(Fig.24C).

MALE. Agrees with description and
figures of Coelioxys nifitarsis except as fol-
lows: (24) Length 7-11 mm; (25) integu-
ment as in female (2 above) ; (26) pubes-
cence white, golden on tarsi; (27) clypeal

margin irregular with 5 or 6 small denti-
cles; (30) paraocular area with integu-
ment rugose, visible among long (0.35
mm), slender, erect setae; (31) rest of face
contiguously, deeply punctured to ocellar
area, visible among slender setae (about
0.5 mm long) ; (34) ocelloccipital distance
subequal to or greater than interocellar
distance, less than ocellocular distance;
(35) vertex more closely punctured than
shown in Figure 21; gena in lateral view
subequal in width to eye, ventral angle
greater than 90°; (37) hypostomal area of
gena with dorsal carina of excavation ex-
tending to the preoccipital carina, exca-
vation with setae of anterior fascia denser
ventrally and longer than setae on disc;
(39) scutum as in female (16 above); (41)
fasciae of mesepisternum indistinct; (42)
axilla as in female (19 above) ; (43) meta-
notum as in female (20 above) ; (45) terga
1-4 with apical fasciae incomplete medi-
ally, apical fascia missing on tergum 5;
gradular grooves on terga 4-6 with fasciae
weak; gradular groove on tergum 2 almost
obliterated medially; fovea on tergum 2
inconspicuous, shallow, about 0.05 mm
long, 0.1 mm wide (Fig. 25E) ; (47) setae
on gonocoxites not as copious as shown in
Figure 26F.

HOST RECORDS. Medler and Koer-
ber (1958) reported Coelioxys moesta to
be a parasite of Megachile relativa; Medler
(1959) recorded M. centiincitlaris as a
host; and Medler and Lussenhop (1968)
related that D. H. Pengelly reared C. moe-
sta from the nests of M. texana and M.
jrigida. Hobbs (1968) noted C. moesta as
a probable parasite of M. rotiindata. A
new host record is from an anonymous
label on a pin with a female of C. moesta
from Tucson, Arizona: M. concinna.

DISTRIBUTION AND SEASON OF
FLIGHT. The distribution of Coelioxys
moesta is practically the same as that of C.
fitneraria (compare Figs. 20 and 30).

Coelioxys moesta is in flight at least

688

The University of Kansas Science Bulletin

Figure 30. Distribution of Coclioxys mocsta.

from May 27 (Virginia) to September 29
(Ontario, Canada).

HABITAT. The habitat of CoeUoxys
moesta agrees closely with that of C. jii-
neraria with both bees taken at many of
the same localities.

GEOGRAPHIC VARIATION. The
variation in CoeUoxys moesta described
above does not seem to be related to
geography.

COMPARATIVE COMMENTS. Fe-
males of CoeUoxys iiioesta are probably
most easily confused with dark females of
C. ban\si but differ in that the profile of
the clypeus is straight in C. moesta. Males
of C. moesta are probably most easily con-

fused with dark males of C. ban}{si but
differ by the dorsal carina of the genal
excavation which extends to the preoccip-
ital carina in C. moesta. Both sexes of C.
moesta differ from all others in this group
by the prothoracic tubercles which are not
produced into a thin, almost translucent
carina. Axillae of both sexes are shorter
than those of other species in this group.
The prothoracic tubercles and axilla of C.
moesta more closely resemble those of C.
fitneraria than those of other bees in the
rufitarsis group. However, C. moesta pos-
sesses complete graduli on metasomal
terga two and three (incomplete in C.
juneraria) .

Taxonomy of Five Nearctic Subgenera of Coelioxys

689

Coelioxys novomexicana Cockercll
Figures 22E, 31

Coelioxys sa\i novomexicana Cockerell,
1909^ Entomol. News 20:9 (9, New
Mexico, type in University of Colorado
Museum).

Coelioxys novomexicana; Crawford, 1914,
Ann. Entomol. Soc. Amer. 7:149
(Key); Bechtel, 1958, Pan-Pacific
Entomol. 34:12-13 (Host).

Coelioxys texana; Bohart, 1970, Utah State
Univ. 41st Faculty Honor Lecture, p. 8
(Misidentification) .

FEMALE. Agrees with description
and figures of Coelioxys ntfitarsis except
as follows: (1) Length 8-13 mm; (2) in-
tegument black; scape, pedicle, and apical
portions of metasomal sterna black to dark
reddish-brown; apical portion of clypeus,
tegula, and legs distal to trochanters fer-
ruginous; (3) pubescence white; golden
on tarsi; clypeus and face with scattered
erect light brown setae; (4) ocular hairs
short (about 0.05 mm long) ; (5) clypeus
uniformly covered by short, appressed
setae and scattered erect setae, margin pro-
duced into 2 lobes (Fig. 22E) ; (7) para-
ocular area obscured by setae about 0.2
mm long plus scattered erect setae; (8)
rest of face rugose, closely punctured up
to ocellar area, with or without a small
median impunctate ridge; (10) ocellar
area raised, impunctate to closely punc-
tured; (11) interocellar distance subequal
to ocelloccipital, less than ocellocular dis-
tance; (12) vertex moderately punctate,
usually with small impunctate area latero-
posterior to ocelli; (13) gena practically
obscured by prostrate setae; (14) hypo-
stomal area of gena with setae ec]ual in
length to those on disc; (16) scutum with
long erect setae on disc less conspicuous
and anterior fascia more distinct than
shown in Figure 21; (17) lateral surface
of mesepisternum with setae shorter
(about 0.15 mm) and sometimes squamose

on disc; (18) scutellum with posterior
margin straight as in Figure 23C; (20)
metanotal setae erect or prostrate medi-
ally; (21) front coxal spine acute, directed
ventrally, inconspicuous; (22) tergum 6
angled as in Figure 24E-G; (23) sternum
6 ovate to broadly lanceolate as in Figure
24E-G.

MALE. Agrees with description and
figures of Coelioxys nifitarsis except as
follows: (25) Integument black; tegula,
venter of metasoma brown; leg distal to
coxae ferruginous; (26) pubescence white,
golden on tarsi; (27) ocular hairs short
(about 0.06 mm) ; (28) clypeus obscured
by short (about 0.33 mm) setae; (30) para-
ocular area obscured by short setae (about
0.4 mm); (31) rest of face rugose, ob-
scured by copious setae up to 0.2 mm long;
(32) interantennal distance subequal to
clypeoantennal distance, greater than an-
tennocular distance; (33) ocellar area as
in female (10 above); (34) ocelloccipital
and ocellocular distances subequal, greater
than interocellar distance; (35) vertex as
in female (12 above); (36) anterior por-
tion of gena practically obscured by pros-
trate setae; (37) hypostomal area of gena
with distinct excavation (as in Fig. 4),
anterior and ventral portion of excavation
with fascia shorter than setae on gena;
dorsal carina of the excavation reaches
preoccipital carina; (39) scutum as in fe-
male (16 above) ; (40) mesepisternum as
in female (17 above); (41) scutellum as
in female (18 above) ; (43) metanotum as
in female (20 above); (45) fovea on ter-
gum 2 about 0.03 mm long, about 0.13
mm wide (as in Fig. 25G) ; tergum 6 with
dorsal spines variable: sometimes reduced,
sometimes as in Fig. 25C, sometimes more
slender than shown in Fig. 25C; (47)
sternum 4 with apical emargination less
conspicuous than shown in Fig. 26B; api-
cal margin of sternum 5 straight, lateral se-
tae more plumose than shown in Fig. 26B.

HOST RECORDS. Bechtel (1958) re-

690

The University of Kansas Science Bulletin

Figure 31. Distributuin of females of Coelioxvs novomexicana.

ported males of Coelioxys novomexicana
(See comparative comments for C. novo-
mexicana.) emerged from a nest of Mega-
chile gentilis. Another possible host is
indicated by a bee collected by Dr. F.
Werner at Tucson, Arizona, close to a nest
of M. sidalceae. A third host bee is M.
brevis from the labels of female C. novo-
mexicana from three localities in Califor-
nia: Summerville, Sacramento, and Tur-
lock. A fourth host record is from the
labels of four small (8-10 mm) females
from Davis and College City, California,
where Dr. R. W. Thorp reared C. novo-
mexicana from nests of M. rotiindata.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys novomexicana is
most common in the southwestern United
States and northern Mexico (Fig. 31).

Season of flight is at least from April
12 (Texas) to October 1 (southern Cali-
fornia).

HABITAT. Coelioxys novomexicana
ranges from various areas classified as
western pine forests (Finns) into areas
classified as juniper-pine (]uniperits), oak-
juniper (Qiiercits), mountain mahogany-
oak ( Cercocarptis ) scrub and woodlands.
In California C novomexicana also has
been taken from areas classified as chap-
arral (Adenostoma, Archtostaphylos, Cea-
nothits), coastal sagebrush (Salvia, Erio-
goniim), California steppe (Stipa), and
tule marshes (Scirpiis, Typha). Areas
classified as creosote bush-bur sage (Lar-
rea, Fanseria) and creosote bush-tarbush
(Flourensia), Trans-Pecos shrub savana,
juniper-oak savana (with Andropogon)
are more southerly vegetation types. This
bee also occurs in areas classified as open
habitats such as alpine meadows and bar-
ren (Agrostis, Carex, Festiica. Poa), and
grama tobosa prairie and shrubsteppe
(Boiiteloua, Hilaria, Larrea). In the east-

Taxonomy of Five Nuarctic Subgenera of Coelioxys

691

ernmost portion of its range, C. novomex-
icana has been taken from an area classi-
fied as oak-hickory forest (Carya).

GEOGRAPHIC VARIATION. Size
(using head width as an index) of fe-
males of Coelioxys novomexicana from
central California was found to vary sig-
nificantly from size of females outside this
region. Head widths of 59 females from
central California (3.06 ± 0.153 mm)
were highly significantly less than the
head widths of 125 females from the rest
of the range (3.39 ± 0.186 mm). Perhaps,
as was postulated for C. funeraria (p. 102),
this difTerence in size is an expression of
host size. The only evidence for this ex-
planation is that three small females from
Summerville, Sacramento, and Turlock,
California were reared from nests of Meg-
acliile brevis whereas a large female was
taken close to a nest of M. sidalceae in
Tucson, Arizona. Since M. sidalceae (11-
14 mm) is a larger bee than M. brevis
(7-12 mm), the cells of M. sidalceae are
doubtless larger and could support a larger
parasite.

Females of Coelioxys novomexicana
are smaller in size in central California,
but in the same general area males have
been taken which had earlier been con-
sidered to be C. novomexicana but which
are larger than true C. novomexicana. It
seems improbable that these males belong
to the same species (assuming that C.
octodentata and C. novomexicana are not
the same species, see discussion under C.
octodentata) as the smaller females. A
more likely explanation is that these males
belong to C. octodentata.

COMPARATIVE COMMENTS. Fe-
males of this species are most likely to be
confused with Coelioxys octodentata al-
though the females of C. novomexicana
have the clypeal margin modified into two
conspicuous lobes (Fig. 22E). The only
other bee in this group possessing a bi-
lobed clypeal margin is C. sayi from which

C. novomexicana differs by the fringe of
setae on the clypeal margin which is di-
rected ventrally (directed posteriorly in C.
sayi). Males of C. novomexicana are most
likely to be confused with males of C.
octodentata from which they are separable
only by distribution (where the distribu-
tions of the females do not overlap).
Males of C. sayi are similar but in C. sayi
the posterior margin of the gradular
groove on the third metasomal tergum is
almost obliterated medially (distinct in C.
novomexicana). In the southern extreme
of their range, males of C. novomexicana
may be confused with males of C. pratti
from which they differ by their short ocu-
lar hairs (long in C. pratti).

Coelioxys oaxacana, new species
Figures 23C, 24K

FEMALE. Agrees with description
and figures of Coelioxys rufitarsis except
as follows: (1) Length 12.5 mm; (2) in-
tegument black; mandible rufous medi-
ally; legs, tegula ferruginous; (3) pubes-
cence white, ochreous on face, golden on
tarsi; (4) ocular hairs short (about 0.06
mm); (5) clypeus hidden by appressed
setae about 0.15 mm long, margin ob-
scured by dense fringe 0.45 mm wide
medially; margin widely emarginate; (6)
paraocular area hidden by setae about 0.2
mm long; rest of face visible between
short, appressed setae, strongly punctured,
broadly conical; (9) clypeoantennal dis-
tance subequal to interantennal distance,
greater than antennocular distance; (10)
ocellar area with impunctate median lon-
gitudinal line anteriorly; (12) vertex
sparsely punctured anteriorly, moderately
punctured posteriorly; (14) hypostomal
area of gena with setae shorter than those
on gena; (16) scutum with fasciae more
distinct than shown in Figure 21; (17)
lateral surface of mesepisternum moder-
ately punctured with many more very
small punctures than larger punctures;

692

The University of Kansas Science Bulletin

disc with very short, scattered, slendered
setae; distinct fasciae anteriorly and pos-
teriorly; (18) scutellum with median lon-
gitudinal impunctate line (Fig. 23C);
(19) axilla slender, acute (Fig. 23C) ; (22)
terga 2, 3 with gradular grooves complete
but almost obliterated medially; tergum 6
elongate (Fig. 24K).

COMPARATIVE COMMENTS.
Coelioxys oaxacajia is distinguished from
all other females in this group by the
broadly conical condition of the ventral
portion of the rest of the face; the di-
vergent, acute axillae (Fig. 23C); and the
elongate, conspicuously angled sixth meta-
somal tergum (Fig. 24K).

TYPE. Holotype female, ten miles
southeast Tapanatepec, Oaxaca, Mexico,
August 8, 1963 (F. D. Parker, L. A.
Stange) in the collection of the University
of California, Davis.

Coelioxys octodentata Say
Figures 24D-I, 25G, 32

Coelioxys octodentata Say, 1824, In Keat-
ing, Narrative of Long's 2nd Expedi-
tion etc. 2:353 (5, type probably de-
stroyed); Say, 1837, Boston J. Nat.
Hist. 1:400 (9); Robertson, 1897,
Trans. St. Louis Acad. Sci. 7:345 (^^
C. altilis, = C. hrevis) ; Fox, 1900,
Entomol. News 11:553 (Host); Hicks,
1926, Univ. Colorado Stud. 15:22 (Bi-
ology); Robertson, 1926, Psyche 33:116
(Host) ; Robertson, 1929, Flowers and
Insects, Lancaster Pennsylvania Sci.
Press 9-217 (Flower records); Grae-
nicher, 1935, Ann. Entomol. Soc. Amer.
28:300, 304 (Host, Wisconsin); Mich-
ener, 1953, Univ. Kansas Sci. Bull. 35:
1059-1060, 1737-1742 (Biology) ; Linsley,
1958, Hilgardia 27:582 (Ecology);
Medler, 1965, Proc. Entomol. Soc.
Washington 67:113-115 (Biology); Ste-
phen, Bohart and Torchio, 1969, Biol-
ogy and External Morphology of Bees:

29 (Immatures) ; Baker, 1971, J. Kansas
Entomol. Soc. 44:225-235 (Immatures).

Coelioxys brevis Cresson (nee Evers-
mann), 1864, Proc. Entomol. Soc. Phil-
adelphia 2:402-403, (9, Connecticut,
New Jersey, Pennsylvania, Maryland,
type # 2490 in the Academy of Natu-
ral Sciences of Philadelphia).

Coelioxys altilis Cresson, 1878, Trans.
Amer. Entomol. Soc. 7:219 (new name
for C. brevis Cresson, not Eversmann).

Coelioxys cressoni Dalla Torre, 1896, Cat-
alogus Hymenopterorum 10:485 (new
name for C. brevis Cresson, not Evers-
mann).

Coelioxys coquilletti Crawford, 1914, Ann.
Entomol. Soc. Amer. 7:157 (9, Cali-
fornia, type # 18224 in the National
Museum of Natural History) NEW
SYNONYMY.

Coelioxys megatricha Cockerell, 1916, J.
Entomol. Zool. 8:60 ((5, California,
type in the American Museum of Nat-
ural History) NEW SYNONYMY.

Coelioxys crassula Cockerell, 1919, Can.
Entomol. 51 :27 ( 9 , Colorado, type in
the American Museum of Natural His-
tory) NEW SYNONYMY.

Coelioxys mediata Cockerell, 1925, Pan-
Pacific Entomol. 1:146-147 (9, Cali-
fornia, type # 1634 in the California
Academy of Sciences) NEW SYN-
ONYMY.

Coelioxys atlantica Mitchell, 1962, North
Carolina Agr. Exp. Sta. Tech. Bull.
152:193, 195, 198 (9, 5 Canada to
Florida in the eastern United States,
type in the collection of T. B. Mitchell)
NEW SYNONYMY.

FEMALE. Agrees with descriptions
and figures of Coelioxys riifitarsis and C.
novomexicana except as follows: (1)
Length 7-12 mm; (2) integument black;
scape, pedicle, apical portions of meta-
somal sterna black to dark ferruginous;
tegula, tibiae, tarsi fulvous; femora black I
to fulvous; (5) clypeus slightly convex,

Taxonomy of Five Nearctic Subgenera of Coelioxys

693

clypeal margin straight or slightly emargi-
nate as in Figure 22B; (22) tergum 6 with
lateral angles variable (Fig. 24D-I); (23)
sternum 6 with apex variable, broadly
ovate to lanceolate (Fig. 24D-I).

MALE. Agrees with descriptions and
figures of Coelioxys nifitarsis and C. noi^o-
tyiexicana except as follows: (45) tergum
6 with dorsal spines usually slightly more
slender than shown in Figure 25G.

HOST RECORDS. Fox (1900) reared
Coelioxys octodentata from a nest of Meg-
achile mendica, a record repeated later by
Medler (1965). Hicks (1926) reared C.
octodentata from a nest of M. hrevis, re-
peated later by Michener (1953) and Baker
(1971). Medler (1958) reared this bee
from a nest of M. centiincidaris, and
Hobbs (1956) reared it from a nest of M.
perihirta. Medler and Lussenhop (1968)
reported that D. H, Pengelly reared C.
octodentata from a nest of M. texana. A
new host record is a female of C. octo-
dentata labeled: Cornish, Utah, IX-10-59,
reared from cell of Meg. onohrychidis.
Another new host record is from three
females from College City, California
reared from three nests of M. rotiindata
by Dr. R. W. Thorp (as would be ex-
pected, these bees are small, 7-8.5 mm).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys octodentata ranges
from southern Canada to the southern
United States (Fig. 32). In the south-
western portion of its range the distribu-
tion of C. octodentata overlaps with that
of C. novomexicana (compare Figs. 31
and 32).

This bee is in flight at least from May
8 (Kansas) to October 22 (central Cali-
fornia).

HABITAT. In the western United
States, Coelioxys octodentata has been col-
lected from a variety of areas classified as
needleleaf forests and broadleaf and nee-
dleleaf forest combinations. This bee has
also been taken from a number of western

shrub lands, western grasslands and west-
ern shrub and grassland combinations.
Elsewhere this bee has been taken from
central and eastern grasslands, grassland
and forest combinations, and eastern
broadleaf and broadleaf and needleleaf
forests.

However, more specific habitat obser-
vations have been reported for Coelioxys
octodentata (Hicks, 1926; Michener, 1953).
The host bee, M. breuis, nests in open
meadows and unshaded sites usually in,
on or near the ground. These are also
types of areas in which C. octodentata is
found flying near the ground and from
which nests containing C. octodentata
larvae have been recovered (Hicks, 1926;
Michener, 1953). Occasionally M. men-
dica nests in the ground at the edges of
open areas, a situation which C. octoden-
tata might exploit.

GEOGRAPHIC VARIATION. Six

characters of Coelioxys octodentata were
analyzed for geographic variation: size,
degree of angulation of the sixth tergum,
density of the medial fascia of the first
metasomal tergum, leg color, punctation
of the vertex, and degree of elongation of
the sixth metasomal sternum. Females
were used for this analysis as the males
in southern Oregon, California, northern
Arizona and New Mexico, Utah, and Col-
orado cannot be distinguished from those
of C. novomexicana. In Figure 32 each of
the six sides of a symbol summarizes data
on one character in the form of a histo-
gram for the group of bees from the area
indicated (a key to characters and scale
appears in the upper right portion of the
map).

Size varies markedly even in bees from
the same general area (in Fig, 32 size was
scored as small, 0-9.0 mm; medium, 9.1-
10.9 mm; large, 11-12 mm). Perhaps, as
was postulated for Coelioxys jitneraria and
C. novomexicana, size is an expression of
host size in C. octodentata.

694

The University of Kansas Science Bulletin

Figure 32. Distributions of fcmaks of Coclioxys octodentata showing variation of six characters given in (upper

right) key.

Degree of development of the lateral
angle of the apical tergum does not seem
to vary geographically except perhaps for

bees from the Southeast. Bees from the
East with well developed angles and api-
cal sterna were formerly considered to be

Taxonomy of Five Nearctic Subgenera of Coelioxys

695

80

Figure 32. (continued) Each sitle of the hexagon is the base line of a hi!>togram. The number inside each
hexagon is the number of specimens examined from the area indicated.

Coelioxys atlantica. Bees from the West somal tergum were considered to be C.

with well developed angles and apical coqiiilletti. Bees intermediate in all con-

sterna and medial fasciae of the first meta- ditions were considered to be C. octo-

696

The University of Kansas Science Bulletin

dentata throughout the distribution. Bees,
especially small specimens, with poorly
developed sixth tergal angles, apical sterna,
and medial fasciae were formerly consid-
ered to be C. salinaria or C. pratti (a mis-
identification; original description of C.
pratti emphasizes the long, brown ocular
hairs). Specimens with poorly developed
angles and long sterna and specimens with
extreme angles and short sixth sterna (Fig.
24D-I) represent mixed conditions among
the extreme examples. Most bees from the
Southeast have the lateral angles moder-
ately developed (Fig. 32).

The medial fascia of the first meta-
somal tergum in general is poorly de-
veloped in the Southeast (minimum in
Fig. 32) but well developed in large speci-
mens from the northwestern and western
United States (maximum in Fig. Z2). In
the West there seems to be a correlation
between size and density of this medial
fascia in that larger bees (formerly con-
sidered to be Coelioxys coqiiilletti) usually
have medial fasciae which are more dense
and contain more squamose setae than
those of smaller bees (formerly considered
to be C. crassitla and C. octodentata) .
However, if only the larger bees are con-
sidered, the medial fascia of the first meta-
somal tergum gradually becomes less dense
and the proportion of squamose setae
drops as a cline across the northern United
States from west to east.

Leg color, especially color of the fe-
mora, varies from black to fulvous with
the color of most specimens intermediate
(ferruginous). Bees of the eastern United
States have fulvous to ferruginous legs
whereas western specimens vary from ful-
vous to black with both large and small
bees displaying both extremes of leg color.

Punctation of the vertex varies from
closely punctate to a condition in which
impunctate areas occur lateroposteriorly to
the ocelli (in Fig. 32 this character was
scored as impunctate areas, minimum;

closely punctured, maximum). Most bees
were intermediate or with impunctate
areas, but a few (mainly from the North-
east) were closely punctured.

The apex of the female sixth sternum
seemed to be highly variable in contrast to
most species of North American Coelioxys
in which the apex of the sixth sternum
may be uniform throughout the whole
series. In C. octodentata, however, the
apex of the sixth sternum ranged from
broadly ovate (Fig. 241, minimum in Fig.
32) to lanceolate (Fig. 24D, maximum in
Fig. v.).

Females with maximum lanceolate
apices of the sixth metasomal sternum
were generally medium to large, although
some small bees had moderately lanceolate
apices (as in Fig. 24G-H).

COMPARATIVE COMMENTS.

The large specimens of this species are
most likely to be confused with Coelioxys
rufjtarsis although the ocular hairs are
short and the medial fascia of the first
metasomal tergum is composed of short,
appressed setae in C. octodentata (long
ocular hairs and long, erect setae on ter-
gum one in C. nifitarsis). In the West,
females of C. octodentata differ from those
of C. novomexicana by their almost
straight clypeal margin (bilobed in C.
novomexicana') although males of C. octo-
dentata are not separable from those of C.
novomexicana. Females of C. octodentata
differ from females of C. sayi in the same
way as from females of C. novomexicana
and males differ from C. sayi by the well
produced posterior margins of the gradu-
lar grooves of metasomal terga two and
three (posterior gradular margin almost
obliterated medially in males of C. sayi).

Coelioxys porterae Cockerell
Figures 22F, 24J, 25F, 33

Coelioxys [lucrosa var.?) porterae Cock-
erelK 1900, Can. Entomol. 32:297-301
( 9 , c^ , New Mexico, key, type #

Taxonomy of Five Nearctic Subgenera of Coelioxys

697

20230 in the National Museum of Nat-
ural History).
Coelioxys porterae; Cockerell, 1904, Ann.
Mag. Natur. Hist. (7) 13:34 (at flow-
ers of Frasera sp.).

Coelioxys diibitata melanopoda Viereck,
1916, Connecticut Geol. Hist. Surv.
Bull. 22:747 ( c^ , Connecticut, type #
66187 in the National Museum of Nat-
ural History); Mitchell, 1962, North
Carolina Agr. Exp. Sta. Tech. Bull.
152:220 (=C. rnfiiarsis, misidentifica-
tion).

Coelioxys hypodonta Cockerell, 1925, Pan-
Pacific Entomol. 1:150-151 {i, Cali-
fornia, type # 1637 in the California
Academy of Sciences) NEW SYN-
ONYMY.

FEMALE. Agrees with description
and figures of Coelioxys nifitarsis except
as follows: (1) Length 9-14.5 mm; (2)
integument granular on vertex between
punctures, rugulose on thorax, faintly
rugulose on metasoma; black; tegula, ven-
ter of metasoma black to piceous; (3)
pubescence white to faintly brownish on
clypeus; (4) ocular hairs of medium
length (about 0.1 mm) ; (5) clypeus also
with erect setae (about 0.16 mm long)
scattered on surface; (7) paraocular area
obscured laterally by setae up to 0.4 mm
long; (11) ocelloccipital distance slightly
greater than interocellar distance and
much less than ocellocular distance; (13)
gena with ventral angle less than 90°;
(14) hypostomal area of gena with setae
subequal to setae on disc; (15) mandible
with hump on middle of outer surface
(Fig. 22F); (18) scutellum with posterior
margin of dorsal surface slightly less
rounded than shown in Figure 21; (19)
axilla as or slightly less produced than
shown in Figure 21; (20) metanotal setae
erect; {22) terga 2 and 3 posterolaterally
less closely punctured than shown in Fig-
ure 21 ; tergum 6 less conspicuously angled

than shown in Figure 21 (Fig. 24J) ;
sternum 6 lanceolate apically (Fig. 24J).

MALE. Agress with description and
figures of Coelioxys nifitarsis except as
follows: (24) Length 9-12 mm; (25) in-
tegument black; tegula, legs distal to
coxae, venter of metasoma black to pic-
eous; (26) pubescence white, golden on
tarsi; (27) ocular hairs of medium length
(about 0.13 mm) ; (28) clypeus obscured
by copious setae about 0.5 mm long; (31)
rest of face closely punctured up to ocellar
area, almost obscured by setae up to 0.7
mm long; (37) hypostomal area of gena
with setae on anterior portion of excava-
tion shorter than on disc; (41-43) scutel-
lum, axilla, metanotum as in female (18-
20 above) ; (45) terga 1-5 with apical
fascia sparse but complete; gradular
grooves of terga 5 and 6 fasciate; fovea on
tergum 2 about 0.1 mm long, about 0.25
mm (Fig. 25E) ; dorsal spines of sternum
5 with denticles slightly more conspicuous
than shown in Fig. 25C.

HOST RECORDS. Mitchell (1962)
listed M. relativa as a host of C. porterae.
A bee from Pibrach, Alberta was reared
from the nest of M. frigida by Dr. G. A.
Hobbs. Hobbs also collected two females
at a log in which M. frigida was nesting
in Scandia, Alberta.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys porterae ranges
across southern Canada and down the
major mountain ranges of the United
States (Fig. 33).

This bee is in flight at least from June
1 (Massachusetts) to September 9 (Medi-
cine Hat, Alberta).

HABITAT. In the Southwest C. por-
terae occurs only at higher elevations in
areas classified as various western needle-
leaf forests. From central California north-
ward it occurs in areas classified as west-
ern needleleaf forests and western shrub
and grasslands. In the Northeast and
mountains of the Southeast, C. porterae

698

The University of Kansas Science Bulletin

Figure ii. Distribution of Coelioxys portcmc.

occurs in areas classified as eastern needle-
leaf and broadleaf forests and combina-
tions of these two vegetation types.

GEOGRAPHIC VARIATION. The
variation described above does not seem to
be related to geography.

COMPARATIVE COMMENTS.
Coelioxys porterae is most likely to be
confused with dark specimens of C. octo-
dentata from which it can be distinguished
by the medium length of the ocular hairs
(short in C. octodentata) in females and
the larger fovea of the second metasomal
tergum (Fig. 25F) in males (small in C.
octodentata, Fig. 25G). Another similar
bee is C. sodalis which has long ocular
hairs and no subapical notches on the sixth
sternum in the female and no foveae on
the second tergum in the male.

Coelioxys pratti Crawford
Figure 34

Coelioxys pratti Crawford, 1914, Ann.
Entomol. Soc. Amer. 7:159 ( 9 , Texas,
type # 18225 in the National Museum
of Natural History).

FEMALE. Agrees with description
and figures of Coelioxys nifitarsis except
as follows: (1) Length 9.5-13 mm; (2)
integument black; tegula, legs ferruginous
to fulvous; lateral margins of metasomal
terga 1 and 2, posterior areas of sterna 2-5
ferruginous; (3) pubescence white, brown-
ish tinge on face and thorax; ocular hairs
light brown; golden on tarsi; (4) ocular
hairs medium to long (about 1.3 mm);
(5) clypeus uniformly covered with short
appressed setae and erect brown setae
(about 0.25 mm long) ; (7) paraocular
area obscured laterally by setae about 0.3
mm long; (11) interocellar and ocellocip-
ital distances subequal, less than ocello-
cular distance; (12) vertex slightly more
closely punctured than shown in Figure
21; (13) gena with ventral angle slightly
less than 90°; (15) mandible with hump
on outer surface slightly less conspicuous
than shown in Figure 22F; (16) scutum
with anterolateral fascia more conspicuous
and erect setae on disc less conspicuous
than shown in Figure 21; (IS) scutellum
with posterior margin less rounded than

Taxonomy of Five Nkarctic Subgenera of Coelioxys

699

shown in Figure 21; (20) metanotum with
erect setae; (22) tergum 6 not as angled
laterally as shown in Figure 21, more like
Figure 24J.

MALE. Agrees with description and
figures o£ Coelioxys nifitarsis except as
follows: (24) Length 9-11 mm; (25) in-
tegument as ill female (2 above), terga 2-4
with lateral margins ferruginous; (26)
pubescence as in female (3 above) ; (27)
ocular hairs long (about 0.15 mm); (28)
clypeus obscured by copious long (about
0.5 mm) setae; (30) paraocular area ob-
scured laterally by long (up to 0.6 mm)
setae; (31) rest of face rugose with long
(up to 0.6 mm) setae sometimes obscuring
integument; (34) ocelloccipital distance
greater than interocellar distance, less than
ocellocular distance; (35) vertex as in fe-
male (12 above) ; (36) gena as in female
(13 above); (37) hypostomal area of gena

with anterior fascia of excavation shorter
than genal setae; (39) scutum as in female
(16 above); (41) scutellum as in female
(18 above); (43) metanotum as in female
(20 above) ; (45) fovea on tergum 2 about
0.05 mm long, about 0.1 mm wide as
shown in Figure 25G; (47) margin of
sternum 5 slightly emarginate; sternum 8
with setae longer, more abundant than
shown in Figure 26E; genital armature
with copious, long (half the length of
gonocoxite) setae on gonocoxite.

DISTRIBUTION AND SEASON OF
FLIGHT. Although the type was col-
lected in Texas, most specimens of Coeli-
oxys pmtti were taken in Mexico and
southward (Fig. 34).

Coelioxys pratti is in flight from at
least March 21 (Cuernavaca, Morelos) to
December 24 (Tamazunchale, San Luis
Potosi).

Figure 34. Distribution of Coelioxys pratti.

700

The University of Kansas Science Bulletin

HABITAT. The type locality (Kerr-
ville, Texas) is in an area classified as
juniper-oak savana (funipenis, Qiiercits,
Andropogon). The remaining localities
are at least 600 miles farther south and at
elevations above 1,000 feet.

COMPARATIVE COMMENTS.
Both sexes of Coelioxys pratti are most
similar to C. octodentata but can be dis-
tinguished by the long, brown ocular hairs
(short and yellowish to white in C. octo-
dentata) .

Coelioxys sayi Robertson
Figures 22G, 35

Coelioxys sayi Robertson, 1897, Trans.
Acad. Sci. St. Louis 7:346 (9,5, Flor-
ida, Illinois, type in the Illinois Nat-
ural History Survey); Robertson, 1926,
Psyche 33:116 (Host); Robertson, 1929,
Flowers and Insects, Lancaster Penn-
sylvania Science Press, p. 9-216 (Flower
records); Mitchell, 1962, North Caro-
hna Agr. Exp. Sta. Tech. Bull. 152:
192, 195, 222 (Key, redescription, = C.
mendacina) ; Medler, 1965, Proc.
Entomol. Soc. Washington 67:113-115
(Host); Krombein, 1967, Trap-nesting
Wasps and Bees, Smithsonian Press, P.
322 (Host); Baker, 1971, J. Kansas
Entomol. Soc. 44:225-235 (Immatures).

Coelioxys mendacina Cockerell, 1921,
Amer. Mus. Novitates 21 :3, 7 ( 9 , Vir-
ginia, type in the American Museum of
Natural History).

FEMALE. Agrees with description
and figures of Coelioxys nifitarsis except
as follows: (1) Length 9.5-12 mm; (2)
integument granular on vertex between
punctures, less so on thorax; (3) pubes-
cence white, slight golden-brown tinge on
clypeus, golden on tarsi; (4) ocular hairs
of medium length (about 0.08 mm) ; (5)
clypeus covered with short appressed setae
plus scattered erect setae; surface visible,
margin thickened and sUghtly bilobed

(Fig. 22G) ; (7) paraocular area almost
obscured by appressed and erect setae
about 0.2 mm long; (8) rest of face rugose
with long (0.45 mm) mediosagittal im-
punctate ridge extending from ocellar
area; (9) clypeoantennal distance subequal
to antennocular distance and less than
interocellar distance; (10) ocellar area
coarsely punctured; (14) hypostomal area
of gena with setae subequal to setae on
disc; (15) mandible with outer margin
straight (Fig. 22G) ; (16) scutum with
erect setae shorter and less conspicuous
than shown in Figure 21; (17) lateral sur-
face of mesepisternum moderately punc-
tured with short (0.2 mm) slender setae
on disc and distinct anterior and posterior
fasciae; (18) scutellum with posterior
margin less rounded than shown in Figure
21; (20) metanotal setae sometimes erect,
sometimes prostrate medially; (22) ter-
gum 3 with gradular groove medially,
fasciae on terga 1-5 not as conspicuous as
shown in Figure 21; tergum 6 with lateral
angles as shown in Figure 24H.

MALE. Agrees with description and
figures of Coelioxys nifitarsis except as
follows: (24) Length 9-10 mm; (26) pu-
bescence as in female (3 above) ; (27)
ocular hairs short (about 0.8 mm) ; (28)
clypeus obscured by long, copious setae
(about 0.3 mm) ; (30) paraocular area ob-
scured by copious setae (about 0.3 mm
long); (31) rest of face rugose, punctured
closely up to ocellar area, with copious
setae (about 0.3 mm long) ; (33) ocellar
area as in female (10 above); (34) ocel-
loccipital distance greater than interocellar
distance, less than ocellocular distance;
(35) vertex usually with impunctate areas
lateroposterior to ocelli; (36) gena with
ventral angle about 90°; (37) excavation
of hypostomal area of gena with setae of
anterior fascia shorter than setae on disc;
(39-43) scutum, mesepisternum, scutel-
lum, axilla, metanotal setae as in female
(16-20 above); (45) gradular grooves of

Taxonomy of Five Nearctic Subgenera of Coelioxys

701

100

Figure 35. Distribution of Coelioxys sayi.

terga 2, 3 as in female (22 above) ; fovea
on tergum 2 about 0.03 mm long; 0.06
mm wide (as in Fig. 25G) ; dorsal spines
of tergum 6 variable, sometimes more, as,
or less slender than shown in Figure 25C;
(47) genital armature with setae not as
abundant on gonocoxite apically as shown
in Figure 26F.

HOST RECORDS. Medler (1965) re-
ported Megachile mendica to be a host of
Coelioxys sayi. This bee has also been re-
covered from nests of M. hrevis.

DISTRIBUTION AND SEASON OF
FLIGHT. Except for scattered records
from the West, Coelioxys sayi is confined
to the eastern United States (Fig. 35).

This bee is in flight from at least March
3 in Florida, April 30 in Texas, and May
8 in Kansas to October 3 in New York
and October 23 in Texas.

HABITAT. In the eastern United
States, Coelioxys sayi has been collected
from areas classified as eastern broadleaf,

eastern broadleaf and needleleaf forests,
and eastern grassland and forest combina-
tions. In the West, C. sayi has been taken
from areas classified as western needleleaf
forests. However, as with C. octodentata,
more specific habitats have been reported
for its host Megachile mendica, as well as
for C. sayi (Medler, 1965). These bees
have been reared from trap nests placed
in open woodland, savana, and forest-edge
habitats. The nests usually occur well
above the ground (although A/, mendica
has been observed to nest in the ground).
C. sayi has been taken from M. brevis
nests at ground level, and has been ob-
served at a woodland edge flying slowly
along the ground as though searching for
host nests.

GEOGRAPHIC VARIATION. As in
the case of Coelioxys alternata, leg color
in C. sayi does not seem to be meaning-
fully correlated to geography. Bees with
all degrees of leg color occur at many lo-

702

The University of Kansas Science Bulletin

calities throughout the range, and the
chance of finding any particular condition
at a given location appears to hinge on
the sample size rather than latitude and
longitude. As with C. texana, however,
wing color does vary geographically. Bees
from Florida to Texas have slightly darker,
very slightly more yellowish wings than
those from elsewhere in the distribution.
COMPARATIVE COMMENTS. Fe-
males of Coelioxys sayi are most similar
to those of C. rnfitarsjs and C. octodentata
although the clypeus is bilobed in C. sayi
and straight in the other two species. In
the West, C. sayi might be confused with
dark females of C. novomexicana but dif-
fers conspicuously in the setal fringe of the
clypeus which is posteriorly directed (ven-
trally directed in C. novomexicana). Males
of C. sayi are most easily confused with
those of C. octodentata, differing mainly
in the reduced posterior margins of the
gradular grooves of terga two and three
(well formed in C. octodentata).

Subgenus Xerocoelioxys Mitchell

Xerocoelioxys Mitchell, 1973. A subgeneric
revision of the genus Coelioxys of the
Western Hemisphere. Cont. Dept.
Entomol. North Carolina State Univ. p.
44. Type species: Coelioxys edita Cres-
son.

Xerocoelioxys is intermediate between
Boreocoelioxys and the subgenus Coelioxys.
The distinct foveae of males of Boreocoeli-
oxys are represented in some members of
Xerocoelioxys by densely punctate fovea!
areas. The notched sixth sternum of females
of Boreocoelioxys is also found in one species
of Xerocoelioxys although the configuration
is not exactly reproduced (compare Figs. 21
and 38B). The subgenus Coelioxys is sepa-
rable from Xerocoelioxys by the reduced fas-
ciae of the mesosoma and metasoma and by
the reduced carina of the prothoracic tuber-
cles.

Bees in Xerocoelioxys fall clearly into two
subgroups: (1) Coelioxys edita, C. boharti, C.
mesae, C. galactiae, C. grindeliae, C. nodis, C.
piercei, and C. soror; and (2) C. aperta and
C. bisoncornua. Reasons for this subdivision

are given under the comparative comments
for C. aperta.

Host records are known for two species,
Coelioxys mesae and C. grindeliae (Table
1), and the host bees fall into two subgenera
of Megachile.

Most of the characters in the list below
are found in other subgenera of New World
Coelioxys and even in some of the Old
World Liothyrapis, but never in this combi-
nation. Italicized characters most clearly dif-
ferentiate Xerocoelioxys from other sub-
genera in North America.

A. Ocellar area variously punctured.

B. Preoccipital carina complete or incom-
plete medially.

C. Mandible robust, variously modified
(Fig.36A-C).

Figure 36. Left half of clypeus and mandible of

females of Xerocoelioxys. A, C. edita; B, C. aperta;

C, C. bisoncornua.

D. Prothoracic tubercle ivith carina pro-
duced into thin, plate-lil{e structure.

E. Scutum moderately to closely punc-
tured, ti'ith conspicuous fasciae.

F. Scutellum closely punctured (Fig.
37A-E).

Figure 37. Scutellum and axillae of females of
Xerocoelioxys. A, C. edita; B, C. aperta; C, C.
grindeliae; D, C. soledadensis; E, C. bisoncornua.

G. Axillae well produced (Fig. 37A-E).

H. Mesepisternum with lateral surface
moderately to closely punctured, with distinct
fasciae.

I. Gruduli complete on metasomal terga
ttvo and three.

J. Female: fore coxal spine variable.

Taxonomy of Five Nearctic Subgenera oe Coelioxys

703

K. Female: sixth metasomal sternum
with apical margin straight or constricted
(notched in Coelioxys hisoncornua, Fig. 38H)
as in Figtires 42 A and 3S/1.

L. Male: hypostomal area of gena with
excavation (Fig. 6).

M. Male: foveal area of metasomal ter-
gum two densely punctured in some species.

N. Male: fifth metasomal tergum with
lateral spine.

Figure 38. A-F, Sixth metasomal tergum and
sternum of females of Xeiocoelioxys. A, C. edita;

B, C. hisoncornua; C, C. galactiae; D, C. sokda-
densis; E, C. grindeliae; F, C. piercei. G-J, Sixth
metasomal tergum of males of XerococUoxys. G,

C. edita; H, C. aperta; I, C. soledadensis; J, C.
nod is.

O. Male: sixth metasomal tergum with
dorsal spines variable, reduced in some spe-
cies (Fig. 38G-J).

P. Male: seventh metasomal tergum var-
iable apically.

Q. Male: apex of metasomal sternum
four conspicuously emarginate (Fig. 39R).

R. Male: sternum seven represented by
two small sclerites (Fig. 39B).

Key to Females of Xerocoelioxys

1. Fore coxal spine reduced to obtuse angle

2
Fore coxal spine conspicuous, directed an-
teriorly or ventrally 3

2. Mandible with distinct flange on outer

surface (Fig. 36C); ocular hairs very

short (about 0.03 mm) hisoncornua

Mandible without distinct flange on outer
surface (Fig. 36B); ocular hairs short
(about 0.06 mm) aperta

3. Sixth metasomal tergum with distinct

subapical protrusions (Fig. 38F)

piercei

Sixth metasomal tergum without protru-
sions 4

4. Sixth metasomal sternum elongate, con-

stricted subapically (Fig. 38E); legs
black to piccous; gena usually hidden by
dense setae; scutcllum sometimes with

raised longitudinal median line

grindeliae

Sixth metasomal sternum variable, if

elongate then legs ferruginous; gena

visible through subappressed setae; scu-

tellum without longitudinal medial line

5

5. Sixth metasomal sternum not or little

constricted subapically (Figs. 38D, 38E)

6

Sixth metasomal sternum more conspicu-
ously constricted subapically (Figs. 38A,
38C 7

6. Lateral surface of mesepisternum with

posterior margin with conspicuous fascia
of appressed, squamose setae; scuto-scu-

tellar suture with dense fascia soror

Lateral surface of mesepisternum with
posterior margin with fascia sparse, in-
distinct or formed from erect, slender
setae; scuto-scutellar suture with fascia
indistinct mesae

7. Mesepisternum with lateral surface con-

tiguously punctured; axilla usually with
small carinae defining dorsal surface
(Fig. 37A); scutum closely punctured
edita

Mesepisternum with lateral surface mod-
erately punctured; axilla not carinate;
scutum moderately to sparsely punctured
galactiae

Key to Males of Xerocoelioxys

1. Sixth metasomal tergum with dorsal

spines reduced, broad as in Figures 38H,

381 2

Sixth metasomal tergum with dorsal
spines at least as conspicuous as shown
in Figure 38G 4

2. Hypostomal area of the gena with dis-

tinct excavation (as in Figs. 5-6); fore
coxal spine conspicuous, without brown

setae soror

Hypostomal area of the gena without dis-
tinct excavation; fore coxal spine re-
duced, with fascia of brown setae 3

704

The University of Kansas Science Bulletin

Figure 39. Hidden sclerites of males of Xtrocoelioxys. A-D, C. edit a. A, Sixth sternum; B, Seventh
sternum (represented by two sclerites); C, Eighth sternum; D, Genital armature. E-G, C. aperta.
E, Sixth sternum; F, Eighth sternum; G, Genital armature. H-I, C. bisoncorniia. H, Eighth sternum;
I, Genital armature. J, C. boharii, genital armature.

Clypeal margin tridentate medially (Fig.
36B); legs, venter o£ metasoma usually
black or piceous aperta

Clypeal margin bidentate medially (Fig.
36C); legs, venter of metasoma usually
with ferruginous areas bisoncornua

Lateral surface of mesepisternum almost
covered by appressed, squamose (almost
scale-like) setae; axilla without carina on

dorsal surface; ocelloccipital distance at
least subequal to ocellocular distance ....
boharti

Lateral surface of mesepisternum without
appressed squamose setae, or if with
squamose setae then axilla with small
carina (Fig. 37A) or ocelloccipital dis-
tance less than ocellocular distance 5

Gena with dense fascia hiding surface;

Taxonomy of Five Nearctic Subgenera of Coelioxys

705

Figure 39. (continued) K-M, C. mcsae. K, Sixth sternum; L, Eighth sternum; M, Genital armature.
N, C. galactiae, genital armature. O-Q, C. grindeliae. O, apical margin of fourth sternum, fifth
sternum; P, Sixth sternum; Q, Genital armature. R-T, C. nodis. R, margin of fourth sternum; S,
Eighth sternum; T, Genital armature. U, C. soledadensis, eighth sternum.

foveal area of metasomal tergum two
with raised, impunctate area (Fig. 25H)
grindeliae

Gena usually with surface visible; foveal
area of metasomal tergum two without
raised impunctate area 6

Lateral surface of mesepisternum with
squamosa setae; scutum with anterior
fascia distinct, of squamose setae .. edita

Lateral surface of mesepisternum with
plumose but slender setae; scutum with
anterior fascia variable, of plumose but
slender setae 7

Scutum moderately to sparsely punctured;
ocular hairs short (about 0.06 mm) .... 8

Scutum closely punctured; ocular hairs

medium length (about 0.09 mm)

mesae

8. Usually 10 mm or longer, midwestern

distribution nodis

Usually 10 mm or less; eastern distribu-
tion galactiae

Coelioxys edita Cresson
Figures 36A, 37A, 38A, G, 39 A-D, 40

Coelioxys edita Cresson, 1872, Trans.
Amer. Entomol. Sec. 4:272 ( S , Texas,
type # 2502 in the Academy of Natu-

706

The University of Kansas Science Bulletin

ral Sciences of Philadelphia) ; Craw-
ford, 1914, Ann. Entomol. Soc. Amer.
7:153 {} = deplanata); Hill, 1936,
Entomol. News M -2^1 ( = deplanatd).

Coelioxys deplanata Cresson, 1878, Trans.
Amer. Entomol. Soc. 7:96 ( 9 , Kansas,
Colorado, type # 2487 in the Academy
of Natural Sciences of Philadelphia).

FEMALE. (1) Length 9.5-14 mm; (2)
integument granular on vertex, very finely
rugulose on thorax and metasoma, black;
antenna black to piceous; coxae, trochan-
ters, venter of metasoma piceous to ferru-
ginous; legs piceous to ferruginous; basal
metasomal terga black to ferruginous; (3)
pubescence white, yellowish on clypeus,
golden on tarsi; (4) ocular hairs short
(about 0.03 mm); (5) clypeus with sur-
face rounded horizontally, rugose, contig-
uously punctured, covered with very short
setae; margin with two strong denticles
(Fig. 36A) obscured by dense setal fringe;
(6) clypeoantennal distance less than lat-
eral margin of clypeus; (7) paraocular
area hidden by appressed setae about 0.1
mm long; (8) rest of face coarsely, closely
punctured, upper median area bare; (9)
clypeoantennal distance greater than an-
tennocular distance and less than inter-
antennal distance; (10) ocellar area with
impunctate areas slightly less conspicuous
than shown in Figures 10, 11 A; (11) in-
terocellar distance greater than ocelloccip-
ital distance and less than ocellocular dis-
tance; (12) vertex less closely punctured
anteriorly and more closely punctured pos-
teriorly than shown in Figure 10; (13)
gena narrower than eye, strongly con-
stricted ventrally, surface visible; ventral
angle greater than 90°; (14) hypostomal
area of gena with setae shorter than those
on disc; (15) mandible with outer surface
closely punctured (Fig. 36A) ; (16) scu-
tum moderately punctured medially,
closely punctured laterally; distinct ante-
rior fascia interrupted medially; slender,
suberect setae in every puncture on disc;

(17) lateral surface of mesepisternum con-
tiguously punctured with large punctures,
scattered squamose setae on disc; conspic-
uous fasciae anteriorly and posteriorly;

(18) scutellum with posterior margin
slightly emarginate in some specimens
(Fig. 37A); (19) axilla with small carina
defining dorsal surface (Fig. 37A) ; (20)
metanotal setae prostrate medially; (21)
fore coxal spine about 0.16 mm long,
rounded apically, directed ventrally; (22)
terga 2, 3 with conspicuous gradular
grooves, terga 1-5 with conspicuous apical
fasciae, tergum 1 with sparse basal fascia;
punctures moderately (tergum 1) to
closely punctured (tergum 6), tergum 2
with foveal area closely punctured; ter-
gum 6 with subapical carina (Fig. 38A) ;
(23) sterna 1-5 fasciate apically, fascia of
sternum 1 interrupted submedially, fascia
of sternum 5 incomplete medially; punc-
tures on sterna 1-4, basally on 5, 6 moder-
ately spaced; sterna 5, 6 apically very
closely punctured; sternum 6 with margin
constricted subapically (Fig. 38A).

MALE. (24) Length 8.5-10.5 mm; (25)
integument as in female (2 above); (26)
pubescence white, golden on tarsi; (27)
ocular hairs short (about 0.05 mm) ; (28)
clypeus hidden by copious setae about 0.3
mm long; surface rugulose and shiny;
margin as in Figure 36A; (29) clypeo-
antennal distance as in female (6 above) ;
(30) paraocular area hidden by setae about
0.3 mm long; (31) rest of face coarsely
punctured, hidden by setae about 0.3 mm
long; (32) interantennal distance subequal
to clypeoantennal distance, greater than
antennocular distance; (33) ocellar area as
in female (10 above); (34) interocellar
and ocelloccipital distances subequal, less
than ocellocular distance; (35) vertex
more closely punctured than shown in
Figure 10; (36) gena narrowed ventrally;
surface visible; ventral angle greater than
90°; (37) hypostomal area of gena modi-
fied into distinct excavation (Fig. 6); an-

Taxonomy of Five Nearctic Subgenera of Coelioxys

707

terior, posterior portions with setae shorter
than on disc of gena; (38-43) mandible,
scutLiiii, mesepisternum, scutellum, axilla,
metanotal setae as in female (15-20
above) ; (44) front coxal spine conspicuous
(0.3 mm long, 0.15 mm wide) with setae
on dorsal and ventral surfaces, rounded
apically, directed anteriorly; (45) terga 1-5
with distinct apical fasciae diminishing
posteriorly, gradular grooves of terga 5, 6
with conspicuous fasciae; foveal area of
tergum 2 closely punctured but not
sunken; tergum 6 with dorsal spines stout
(Fig. 38G) ; tergum 7 unmodified (as in
Fig. 26A) ; (46) sterna 1-4 with punctures
and fasciae as in female (23 above) ; (47)
sternum 4 submarginally carinate (as in

Fig. 26C), fascia incomplete, apex entire;
sternum 5 as in Figure 19C except mar-
gin with small, median emargination;
sternum 6 with conspicuous shoulders
(Fig. 39A); sternum 7 represented by two
sclerites (Fig. 39B); sternum 8 with base
broad (Fig. 39C); genital armature with
gonobase incomplete, setae on gonocoxite
sparse (Fig. 39D).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys edita ranges from
southern Canada into northern Mexico
(Fig. 40). Few specimens have been col-
lected east of the Mississippi River.

This bee is in flight at least from April
10 (Texas) to October 11 (Baja Cali-
fornia).

Figure 40. Distribution of Coelioxys edita.

708

The University of Kansas Science Bulletin

HABITAT. Coelioxys edita has been
collected from areas classified as a wide
variety of vegetation types. Its distribution
does not seem to be related to vegetation
type.

COMPARATIVE COMMENTS.
Coelioxys edita most closely resembles C.
bisoncorniia from which it can be distin-
guished by the slightly carinate axilla and
slightly emarginate scutellum (non-cari-
nate axilla and slightly produced scutel-
lum in C. bisoncorniia, compare Figs, 37A
and 37E).

Coelioxys aperta Cresson
Figures 36B', 37B, 38H, 39E-G, 41

Coelioxys aperta Cresson, 1878, Trans.
Amer. Entomol. Soc. 7:95 (9, Colo-
rado, type # 2488 in the Academy of
Natural Sciences of Philadelphia);
Cockerell, 1921, Amer. Mus. Novitates
21:4-10 {$ described, key).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 11-15.5 mm; (2) in-
tegument black; antenna, tegula, piceous;
legs piceous to dark ferruginous; (4) ocu-
lar hairs short (about 0.05 mm) ; (5) clyp-
eus slightly convex to rounded horizon-
tally, rugose, shining; surface visible except
for longitudinal median line covered with
short, appressed setae, lateral and dorsal
margins bare; margin with three strong
denticles (Fig. 36B); setal fringe scant;
(7) paraocular area visible between setae
about 0.3 mm long; (8) rest of face closely,
coarsely punctured except for irregular
mediosagittal line; (10) ocellar area closely
punctured; (11) interocellar and ocel-
loccipital distances subequal, less than
ocellocular distance; (12) vertex closely
punctured; (14) hypostomal area of gena
with setae longer than those on disc; (15)
mandible with outer surface rounder than
shown in Figure 36A (Fig. 36B); (16)
scutum closely punctured, longitudinal
median line complete; anterior fascia rep-

resented by median patch, weaker lateral
line; (17) mesepisternum with anterior
and posterior fasciae sparse; (18) scutel-
lum with posterior margin produced
slightly (Fig. 37B) ; (19) axilla without
carina (Fig. 37B); (21) front coxal spine
reduced; (22) tergum with foveal area
unmodified; (23) fascia of sternum 1 rep-
resented by median patch, fascia of ster-
num 5 complete.

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 9.5-12.5 mm; (25) in-
tegument as in female (2 above) ; (27)
ocular hairs of medium length (about 0.1
mm) ; (28) clypeus obscured by long setae
(about 0.45 mm) ; (30) paraocular area
obscured by setae about 0.45 mm long;
(32) clypeoantennal distance less than in-
terantennal distance, greater than anten-
nocular distance; (33) ocellar area as in
female (10 above); (34) ocellocular dis-
tance greater than interocellar distance,
less than ocelloccipital distance; {^6) gena
with ventral angle less than 90°; (37)
hypostomal area of gena not excavated,
but as in female (14 above); (38-43) man-
dible, scutum, mesepisternum, scutellum,
axilla, metanotal setae as in female (15-20
above) ; (44) front coxal spine reduced,
with dense, brown anteroventrally directed
spot of setae; (45) terga 3-5 with apical
fasciae incomplete medially, median fas-
ciae becoming more distinct posteriorly;
tergum 2 with foveal area unmodified;
tergum 6 with dorsal spines reduced (Fig.
38H); tergum 7 with apical spine about
half as long as shown in Figure 13A; (46)
sterna 1-4 with punctures and fasciae as in
female (23 above) except each puncture
with seta becoming more conspicuous pos-
teriorly; gradular grooves with fasciae;
(47) sternum 4 subapically carinate, fascia
complete; emargination of sternum 5
broad; shoulders of sternum 6 bifid, setae
inflated medioposteriorly (Fig. 39E) ;
sternum 8 with lateral angles extended,

Taxonomy of Fivk Nearctic Subgenera of Coelioxys

709

Figure 41. Distributions of Coelioxys apeita, C. bisoncornua, C. boharti, and C. galactiae. Coelioxys
aperta is also known from Colorado (state record only).

base narrow (Fig. 39F) ; genital armature
with slender branched setae on gonocox-
ites (Fig. 39G).

DISTRIBUTION AND SEASON OF
FLIGHT. This bee ranges from Colo-
rado to Guadalajara, Mexico (Fig. 41).

Coelioxys aperta is in flight at least
from July 19 (New Mexico) to October 5
(Arizona) and to October 16 (Guadala-
jara, Mexico).

HABITAT. Areas classified as west-
ern coniferous forests (Abies, Picea, Finns,
Psetidotsitga) appear to be the principal
habitat of Coelioxys aperta although it has
been collected in oak-juniper woodland
(Qtierciis, Jiiniperiis) and perhaps has
been collected in grama-tobosa shrubsteppe
(Bouteloiia, Hilaria, Larrea).

GEOGRAPHIC VARIATION. Al-

though the series of this species is small,
specimens from the southern part of the
range are larger with the clypeal marginal
denticles longer and the metasomal fascial
more distinct.

COMPARATIVE COMMENTS.
Coelioxys aperta resembles the larger,
darker specimens of C. edita but differs in
both sexes by the conspicuous median
tooth of the clypeal margin (Fig. 36B),
the small median projection of the scutel-
lum (Fig. 37B), and the lack of carinae
on the axillae (Fig. 37B). The reduced
fore coxal spines of both sexes, the spine
on tergum seven of males, the bifid shoul-
ders of sternum six in males, the wide
eighth sternum, and the complete gono-
base of the genital armature are different
from other bees in Xerocoelioxys except
for C. bisoncornua. The lack of an exca-

710

The University of Kansas Science Bulletin

vation of the hypostomal area of the gena
suggests males of Synocoelioxys (Fig. 3).

Coelioxys bisoncorniia Hill
Figures 36C, 37E, 38B, 39H-I, 41

Coelioxys bisoncorniia Hill, 1936, Entomol.
News 47:205 ( 9 , ^ , Nebraska, Kan-
sas, type in the collection of The Uni-
versity of Nebraska, Lincoln).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 12-15 mm; (2) integu-
ment black, antenna black to ferruginous;
mandibles rufous; legs, venter of meta-
somal rufous to ferruginous; tegula brown;
(4) ocular hairs very short (about 0.16
mm); (5) clypeus slightly convex; bare
or visible through short (about 0.15 mm)
appressed setae; margin with sparse fringe,
two strong denticles (Fig. 36C) ; (7) para-
ocular area with ventral half bare or with
scattered appressed setae; (8) rest of face
bare except for area about 0.3 mm wide
around base of antenna; (10) ocellar area
closely punctate; (11) ocellocular distance
subequal to ocelloccipital distance, greater
than interocellar distance; (12) vertex
very closely punctured laterally, contigu-
ously punctured with slighdy smaller
punctures medially; (13) gena constricted
ventrally; (15) mandible with flange on
outer surface (Fig. 36C) ; (16) scutum
with indistinct anterior fascia complete;
slender, erect setae on disc sparse, short;
(17) lateral surface of mesepisternum con-
tiguously punctured with medium sized
punctures, fascia indistinct; (18) scutel-
lum with posterior margin rounded, pro-
duced medially (Fig. 37E); (19) axilla
without carinae (Fig. 37E) ; (20) meta-
notum with dorsal setae subprostrate;
(21) front coxal spine reduced; (22) ter-
gum 3 with gradular groove shallow or
incomplete medially; terga 1, 5 closely
punctured, tergum 6 very closely punc-
tured, terga 2-4 moderately punctured;
foveal area of tergum 2 unmodified; ter-

gum 6 broadly rounded (Fig. 38B); (23)
sterna 1-5 sparsely fasciate apically, fascia
of sternum 1 a median patch; sterna 1-5
moderately punctured, sternum 6 with
contiguous, elongate, irregular punctures,
subapicallv notched in some specimens
(Fig. 38B).

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 11-14.5 mm; (25) in-
tegument as in female (2 above); (28)
clypeus obscured by long (about 0.45 mm)
copious setae; (30) paraocular area ob-
scured by setae about 0.45 mm long; (31)
rest of face obscured by setae about 0.45
mm long; (32) clypeoantennal distance
greater than antennocular distance, less
than interantennal distance; (33) ocellar
area as in female (10 above) ; (34) ocel-
loccipital distance less than ocelloccipital
distance, greater than interocellar distance;
(37) hypostomal area of gena not exca-
vated, but as in female (14 above); (38)
mandible with outer margin slightly
rounder than shown in Figure 36B; (39-
43) scutum, mesepisternum, scutellum,
axilla, metanotal setae as in female (16-
20 above); (44) front coxal spine as de-
scribed for male of C. aperta; (45) foveal
area of tergum 2 unmodified; tergum 6 as
in Figure 38H; tergum 7 as described for
C. aperta; (46) sterna 1-4 with conspicu-
ous apical fasciae, 2-4 with conspicuous
basal fasciae in gradular grooves; (47)
margin of sternum 4 broadly emarginate;
marijin of sternum 5 entire, sternum 6 as
in Figure 39E but with slightly more co-
pious setae; sternum 8 broadly rounded
basallv (Fig. 39H); genital armature large
(Fig.M).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys bisoncorniia ranges
from Minnesota through Nebraska and
Kansas (Fig. 41).

This bee is in flight at least from July
28 to September 13.

HABITAT. Coelioxys bisoncornua has

Taxonomy of Five Nearctic Subgenera of Coelioxys

711

been taken in areas classified mostly as
central and eastern grasslands, grassland
and forest combinations but also in eastern
ponderosa pine forest, northern flood plain
forest and oak-hickory forest.

Flower records indicate an open, dis-
turbed habitat as all but one are sunflower,
a flower of open areas rather than wood-
lands. Furthermore, Mr. K. W. Richards
captured a female "searching" along a dirt
hank near Lawrence, Kansas.

COMPARATIVE COMMENTS.
This bee is most likely to be confused with
large specimens of Coelioxys cdita because
C. edita is far more common than the
more similar C. aperta. Females of C.
hisoncorniia have a conspicuous flange on
the outer surface of the mandible (Fig.
36C) which is lacking in C. edita and C.
aperta. Males of C. bisoncorniia lack small
carinae on the axillae evident in C. edita
(compare Figs. 37A and 37E) and have
more ferruginous areas on the legs and
venter of the metasoma than the darker
C. aperta.

Coelioxys boharti Mitchell
Figures 39J, 41

Coelioxys boharti Mitchell, 1962, North
Carolina Agr. Exp. Sta. Tech. Bull.
152:194, 206-207 (6, Florida, type in
the collection of the National Museum
of Natural History).

! MALE. Agrees with description and

; figures of Coelioxys edita except as fol-
* lows: (24) Length 12-13.5 mm; (25) in-
tegument black; antenna, tegula, legs pic-
eous; venter of metasoma dark ferrugi-
nous; (26) pubescence white, slightly yel-
, low on face, golden on tarsi; (31) rest of
face hidden ventrally by appressed setae
about 0.15 mm long, dorsally by sub-
appressed setae about 0.45 mm long; (32)
clypeoantennal distance greater than an-
tennocular distance, less than interanten-
nal distance; (33) ocellar area moderately
punctured; (34) ocellocular distance less

than ocelloccipital distance, greater than
intcrocellar distance; (37) excavation of
hypostomal area of gena almost bare; (38)
mandible with outer surface rounded as
in Figure 67E; (39) scutum with anterior
fascia sparse medially; (40) mesepister-
num with lateral surface closely punctured
with large punctures; punctures filled
with appressed, sc]uamose setae which may
hide the surface; (42) axilla not carinate;
(45) foveal area of tergum 2 closely punc-
tured with small punctures, very slightly
sunken; tergum 6 with dorsal spines as in
Figure 38G; (47) apical margin of ster-
num 4, sternum 5 as in Figure 26B; ster-
num 8 as in Figure 26 E; genital armature
with gonobase entire, setae on gonocoxite
long (Fig. 39J).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys boharti has been col-
lected in Texas (College Station) and
Florida (Alachua County, Cocoa, and La-
belle, Fig. 41).

This bee is in flight at least from June
21 to August 17.

HABITAT. This bee, known only
from the male, has been taken from areas
classified as oak-hickory (Quercus-Carya),
southern mixed forest (Fagits, Liqiiidam-
bar, Magnolia, Finns, Qtiercus), and pal-
metto prairie (Serenoa-Aristida).

COMPARATIVE COMMENTS.
Coelioxys boharti resembles males of C.
galactiae but is larger. The lateral surface
of the mesepisternum is almost hidden by
appressed squamose setae in C boharti
but is almost bare with scattered, erect,
plumose (but slender) setae in C. galactiae.

Coelioxys galactiae Mitchell
Figures 38C, 39N, 41

Coelioxys galactiae Mitchell, 1962, North
Carolina Agr. Exp. Sta. Tech. Bull.
152:193-195, 204-206 (9,6 North Car-
olina, Florida, Illinois, type is collec-
tion of T. B. Mitchell).

712

The University of Kansas Science Bulletin

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 9-11 mm; (2) in-
tegument granular on vertex, thorax;
finely rugulose on metasoma; black; an-
tenna, tegula piceous; mandible, venter
and dorsum (basally) of metasoma black
to rufous; legs rufous to ferruginous; (3)
pubescence white, yellowish on face,
golden on tarsi; (4) ocular hairs short
(about 0.06 mm); (5) clypeus almost flat,
margin with 4-6 denticles obscured by
apical fringe; (10) ocellar area moderately
punctate anteriorly; (11) interocellar dis-
tance slightly greater than ocelloccipital
distance, less than ocellocular distance;

(12) vertex punctured as in Figure 10;

(13) gena constricted ventrally; surface
almost obscured; (15) mandible as in Fig-
ure 22E; (16) scutum punctured as ver-
tex in Figure 10; anterior fascia indistinct;

(17) lateral margin of mesepisternum
moderately punctured, fasciae distinct;

(18) scutellum sparsely punctured; (19)
axilla without conspicuous carinae; (22)
tergum 1 with basal fascia inconspicuous;
postgradular areas less closely punctate
than pregradular areas; tergum 6 nar-
rowly rounded apically (Fig. 38C) ; (23)
fascia of sternum 5 sparse but complete
medially; apical portion of sternum 6
longer than shown in Figure 3SA (Fig.
38C).

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 9-10 mm; (25) integu-
ment as in female (2 above) ; (26) pubes-
cence white, yellow on face, golden on
tarsi; (32) clypeoantennal distance greater
than antennocular distance, less than in-
terantennal distance; {H) ocellar area as
in female (10 above); (34) ocelloccipital
distance greater than interocellar distance,
greater than ocellocular distance; (38)
mandible as in female (15 above); (39)
scutum moderately punctured; (40) mes-
episternum moderately punctured, almost

obscured by setae about 0.45 mm long;
(41, 42) scutellum, axilla as in female (18-
19 above); (45) gradular groove of ter-
gum 4 with conspicuous fasciae laterally;
tergum 6 with apical spines slightly more
elongate than shown in Figure 39G; (47)
sternum 4 with subapical carina indented
medially, apical margin bidentate as in
Figure 390; sternum 6 as in Figure 39A;
sternum 8 with base about half as broad
as shown in Figure 39C; abundant setae
apically on gonocoxite (Fig. 39N).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys galactiae is confined
to the eastern United States (Fig. 41).

This bee is in flight at least from June
23 to August 6,

HABITAT. Coelioxys galactiae has
been collected from areas classified as a
mixture of bluestem prairie and oak-hick-
ory forest (Andropogon, Paniciim, Sor-
ghastnim, Qitercus, Carya), oak-hickory
forest, oak-hickory-pine forest (Piniis),
southern mixed forest (Fagits, Li qui dam -
bar, Magnolia, Pinus, Qitercus), and poco-
sin {Pinus, Ilex).

COMPARATIVE COMMENTS.

This bee is similar to Coelioxys edita but
differs by the more attenuate sixth ster-
num of females (compare Figs. 38C and
38 A) and more attenuate apical spines of
males (longer than shown in Fig. 38G).
Also the axillae of C. galactiae in both
sexes are not carinate and are more
sparsely punctured than those of C. edita
(Fig. 37A).

Coelioxys grindeliae Cockerell
Figures 25H, 37C, 38E, 390-Q; 44

Coelioxys grindeliae Cockerell, 1900, Can.
Entomol. 32:297-301 ( ci , 9 New Mex-
ico, key, type in the American Museum
of Natural History;

Coelioxys grindeliae denverensis Cockerell,
1912, Can. Entomol. 44:166-170 ( <5 ,
Colorado, key, type in the National
Museum of Natural History) ;

Taxonomy of Five Nearctic Subgenera gi-- Coelioxys

713

Coelioxys lamellicaiida Cockerell, 1921,
Amer. Mus. Novitates 21 :6-8 ( 6 , Col-
orado, type in the American Museum
of Natural History).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 9-11.5 mm; (2) in-
tegument black; antenna, legs, tegula,
venter of metasoma black to piceous; (3)
pubescence white, golden on tarsi; (5)
clypeus almost flat, very closely punctate,
almost hidden by dense, 1.5 mm long,
plumose setae and scattered erect setae;
margin with 5 small evenly spaced den-
ticles; (7) paraocular area hidden by
dense, plumose setae 2.0 mm long; (8)
rest of face closely punctured except for
longitudinal median line, almost hidden
by appressed, plumose setae; (10) ocellar
area moderately to closely punctured; (12)
vertex punctured as in Figure 21; (13)
gena constricted ventrally, surface visible
anteriorly; (15) mandible as in Figure
22E; (16) scutum closely punctured; an-
terior fascia sparse medially; (17) lateral
surface of mesepisternum nearly to com-
pletely contiguously punctured, disc bare
or with scattered, suberect setae; (18)
scutellum sometimes with longitudinal
median impunctate ridge (Fig. 37C) ; (19)
axilla long (Fig. 37C); (22) tergum 1
with basal fascia inconspicuous; tergum 2
with foveal area unmodified; tergum 6
more slender than shown in Figure 38A
(Fig. 38E); (23) sterna 1-5 moderately to
sparsely punctured, punctures very shal-
low medially, surface shining; sternum 5
apically contiguously punctured with
small punctures; sternum 6 elongate api-
cally (Fig. 38E).

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 9.5-11 mm; (25) in-
tegument as in female (2 above) ; (28)
clypeal margin with 3 distinct denticles;
{M) ocellar area as in female (10 above);
(36) ventral half of gena hidden by dense

setal fascia; (38, 39) mandible, scutum as
in female (15, 16 above); (40) mesepister-
num with lateral surface almost hidden
in some specimens; (41-43) scutellum,
axilla, metanotal setae as in female (18-20
above) ; (45) foveal areas of tergum 2
with small, raised, impunctate area (Fig.
25H) ; (46) sterna 1-4 with conspicuous
fasciae posteriorly; sternum 4 carinate sub-
apically, margin with 2 projections (Fig.
390) ; sternum 6 with mostly expanded,
rather than plumose setae, heavily pig-
mented laterally (Fig. 39P) ; sternum 8
with base broadly rounded; genital arma-
ture with gonobase complete (Fig. 39Q).
HOST RECORDS. Hicks (1926) re-
ported Coelioxys grindeliae to be a parasite
of Megachile perihirta.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys grindeliae is con-
fined to southern Canada and the western
United States (Fig. 44).

This bee is in flight at least from June
4 (northern California) to September 26
(Middle California).

HABITAT. Coelioxys grindeliae has
been taken from a diversity of western
forest, shrub and grassland vegetation
types, as well as central and eastern grass-
land vegetation types and northern flood-
plain forest.

COMPARATIVE COMMENTS.

Most specimens of Coelioxys grindeliae
have the gena obscured by a dense, white
fascia, a condition found in no other spe-
cies in this group. Males of C. grindeliae
are distinguished from other males by the
raised, impunctate area of the foveal area
(Fig. 25H). Females are distinguished by
the elongate sixth metasomal sternum
(Fig. 38E), dark legs, and sometimes by
the raised longitudinal median line of the
scutellum (Fig. 37C).

Coelioxys mesae Cockerell
Figures 39K-M, 42A, 43

Coelioxxs mesae Cockerell, 1921, Amer.

714

The University of Kansas Science Bulletin

Figure 42. Dorsal views of female CocUoxys. A, C. mesae; B, C. niitclieUi.

Taxonomy of Five Nearctic Subgenera of Coelioxys

715

Mus. Novitates 21 :6 ( 6 , Colorado,
type in the American Museum of Nat-
ural History).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (2) Integument finely granular,
dull on vertex, metasoma; black; antenna,
legs, venter of metasoma black to piceous;
(3) pubescence white, sometimes brown-
ish on mesosoma, golden on tarsi; (4)
ocular hairs medium length (O.OcS mm) ;
(5) clypeus convex, rugose, closely punc-
tured, shiny; disc bare, margin with 5
small denticles; apical fringe conspicuous
but not dense; (7) paraocular area hidden
by setae up to 0.3 mm long; (8) ventral
portion of face with sparse, appressed se-
tae about 0.1 mm long; dorsal portion
with mediosagittal portion bare or ob-
scured by longer setae; (10) ocellar area
closely punctured; (12) vertex coarsely,
evenly punctured (Fig. 42A); (13) gena
with ventral angle about 90°; (15) man-
dible with outer margin rounded as in
Figure 22E; (16) scutum closely, evenly
punctured; disc with short, inconspicuous
erect setae (Fig. 42A); (17) lateral surface
of mesepisternum largely bare, fasciae
sometimes indistinct; (18) scutellum with
posterior margin almost straight medially
(Fig. 42A); (19) axilla not carinate (Fig.
42A); (21) front coxal spine directed an-
teriorly; (22) metasomal terga closely
punctured; tergum 2 with foveal area un-
modified; tergum 6 subtriangular (Fig.
42A); (23) sternum 5 complete medially;
sternum 6 subtriangular apically (Fig.
42A).

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 9-11 mm; (25) in-
tegument as in female (2 above) ; (27)
ocular hairs medium (about 0.1 mm) ;
(28) clypeus hidden by copious setae about
0.36 mm long; margin as in female; (30)
paraocular area hidden by setae about 0.45
mm long; (31) rest of face closely punc-

tured; hidden by setae about 0.4 mm long;
(32) clypeoantennal distance less than in-
terantennal distance, greater than anten-
nocular distance; (33) ocellar area as in
female (10 above); (34) interocellar dis-
tance greater than ocelloccipital distance,
less than ocellocular distance; (35) vertex
slightly more irregularly punctured than
shown in Figure 42A; (36) gena slightly
narrowed ventrally; (38, 39) mandible,
scutum as in female (15, 16 above); (40)
mesepisternum with slender, plumose se-
tae on disc; (41-43) scutellum, axilla,
metanotal setae as in female (18-20
above) ; (45) tergum 6 with dorsal spines
more conspicuous than shown in Figure
38G; (47) sterna 4, 5 as in Figure 390
except for longitudinal ridges on rim of
sternum 4; sternum 6 as in Figure 39K;
sternum 8 with base rounded (Fig. 39L) ;
gonocoxite with abundant, plumose setae
(Fig. 39M).

HOST RECORDS. Dr. G. E. Bohart
has reared this bee from a nest of Mega-
chile itmatillensis at Cornish, Utah.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys mesae is confined to
the western half of the United States (Fig.
43).

This bee is in flight at least from May
15 (Texas) and June 10 (Oregon) to
August 29 (Utah).

HABITAT. Most localities from
which Coelioxys mesae has been collected
are classified as sagebrush steppe {Artemi-
sia-Agropyron). Two specimens were col-
lected from areas classified as grama-
tobosa prairie (Bouteloua, Hilaria) and
pine-Douglas fir forest (Pi mis, Pseii-
dotsitga).

COMPARATIVE COMMENTS.
Females of Coelioxys mesae differ from
those of C. edita by darker integument
and entire lateral margin of the sixth ster-
num (compare Figs. 42A and 38A) and
from those of C. mitchelli by the greater
amount of appressed, squamose setae on

716

The University of Kansas Science Bulletin

Figure 43. Distributions of Coclioxys mcsae, C. mitchcUi. and C. soledadensis.

the scutum, scutellum, and metanotum
(compare Figs. 42A and B). Males of C.
7nesae differ from those of C. edita by
longer ocular hairs, longer facial setae, and
darker integument. Males of C. mcsae
diflfer from those of C. mitchelU by the
entire, distinct basal fascia of tergum one
(incomplete in C. mitchelU).

Coclioxys nodis, new species
Figures 38J, 39R-T, 48

MALE. Agrees with description and
figures of Coclioxys cdita except as fol-
lows: (24) Length 10-11 mm; (25) in-
tegument black, antenna, dorsum of meta-
soma black to piceous; legs, tegula, venter
of metasoma piceous to ferruginous; (28)
margin of clypeus with six small denticles;
(30) paraocular area hidden by setae about
0.45 mm long; (31) rest of face hidden by
appressed setae about 0.25 mm long; (33)
ocellar area with small, irregular impunc-
tate areas anteriorly; (34) ocelloccipital
distance greater than interocellar distance,
less than ocellocular distance; (38) man-

dible with outer surface simple as in
Figure 22E; (39) scutum moderately
punctured, anterior fascia indistinct; (40)
lateral surface of mesepisternum closely
punctured, setae squamose to slender, fas-
cia indistinct; (41) scutellum slightly
rounded posteriorly, sparsely punctured;
(42) axilla without carinae; (45) tergum
6 with dorsal spines reduced, broad (Fig.
38J); (47) sternum 4 with submarginal
carina indented medially (Fig. 39R); ster-
num 8 with base truncate (Fig. 39S);
genital armature with gonobase lobed,
gonocoxite with long, abundant setae (Fig.
39T).

DISTRIBUTION AND SEASON OF
FLIGHT. Coclioxys nodis has been col-
lected in Illinois, Kansas, and Nebraska
(Fig. 48) from July 3-12.

HABITAT. Coclioxys nodis has been
collected in bluestem grama prairie (An-
dropogon, Boiitcloua).

COMPARATIVE COMMENTS.
Coclioxys nodis resembles larger males of
C. cdita but differs by lack of carinae on

Taxonomy of Five Nearctic Subgenera of Coelioxys

ni

Figure 44. Distribution of Coelioxys grindeliae.

the axillae, by the subapical carina of ster-
num four which is indented medially
(Fig. 39R), and by the dorsal spines of
tergum six which are broad (compare
Figs. 38J and 38G).

TYPES. Holotype male, Cams, Ne-
braska, July 3, 1902, Verbena (W. D.
Pierce), in the collection of the Univer-
sity of Nebraska at Lincoln. Paratypes
males; one, Sioux Co., Nebraska; one,
Trego Co., Kansas, 2450 feet, July 12, 1912
(F. X. Williams) ; one, 35530, Sands, Illi-
nois (Hart), 128. Paratypes in the collec-
tions of the University of Nebraska and
the University of Kansas,

The specific names refer to the lobes
on the gonobase which are more con-
spicuous in this species than any other in
this study.

Coelioxys piercei Crawford
Figure 38F

Coelioxys piercei Crawford, 1914, Ann.
Entomol. Soc. Amer. 7:152-153 ( 9 ,
Texas, type # 18219 in the National
Natural History Museum).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 9.5 mm; (2) in-

718

The University of Kansas Science Bulletin

tegument black; antenna, legs, tegula pic-
eous; (4) ocular hairs short (about 0.05
mm long); (5) clypeus with surface al-
most convex, clypeal margin with 3 small
denticles; (10) ocellar area closely punc-
tate; (12) vertex moderately punctured;

(15) mandible with outer margin rounded,
resembling that shown in Figure 22E;

(16) scutum moderately punctured, fascia
slightly more conspicuous laterally than
anteriorly; (17) lateral surface of mesepi-
sternum dorsally with slender, scattered
setae on disc; (18) scutellum with pos-
terior margin as in Figure 42B; punctures
as in Figure 37D; (19) axilla elongate as
in Figure 37B; (22) tergum 2 with foveal
area unmodified; tergum 6 with subapical
flattened projections (Fig. 38F) ; sterna 1-5
fasciate apically; sternum 6 with lateral
margin entire (Fig. 38F).

DISTRIBUTION AND SEASON OF
FLIGHT. Type specimen from CotuUa,
Texas, taken on April 17. No other speci-
mens are known.

HABITAT. The area surrounding
Cotulla has been classified as mesquite-
acacia savana (Prosopis, Acacia, Andro-
pogon, Setaria).

COMPARATIVE COMMENTS.

Coelioxys piercei resembles the females of
C. mesae and C. mitchelli but differs by
the flattened lateral projections of tergum
six (Fig. 38F).

Coelioxys soledadensis Cockerell
Figures 37D; 38D, I; 39U; 43

Coelioxys soledadensis Cockerell, 1909,
Entomol. News 20:9 {i , New Mex-
ico, type in the American Museum of
Natural History).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 8-12 mm; (2) in-
tegument shiny on vertex; black, antenna,
tegula black to piceous; legs, metasoma
basally and ventrally piceous to rufous;
(3) pubescence white, yellowish dorsally,

golden on tarsi; (5) clypeus slightly con-
vex, visible through minute setae, margin
with two denticles not as conspicuous as
shown in Figure 36 A; (7) paraocular area
hidden by appressed setae about 0.05 mm;
(8) lower portion of rest of face plus lon-
gitudinal median line up to ocellar area
visible, closely punctured; (9) clypeo-
antennal distance subequal to antennocu-
lar distance, less than interocellar distance;
(10) ocellar area moderately punctured;
(12) vertex moderately punctured; (14)
hypostomal area of gena with setae sub-
ec]ual in length to those on disc; (15)
mandible with outer surface rounded as
in Figure 22E; (16) scutum evenly, con-
tiguously punctured; anterior fascia dis-
tinct, narrowly interrupted medially; disc
with minute suberect setae in each punc-
ture; (17) lateral surface of mesepister-
num moderately punctured, disc bare,
fasciae distinct; (18) scutum with poste-
rior margin rounded (Fig. 37D); (19)
axilla not carinate (Fig. 37D) ; (22) terga
1, 2 very closely punctured; terga 3-5 mod-
erately punctured; tergum 6 very closely,
minutely punctured (Fig. 38D) ; (23)
sternum 6 slightly constricted subapically
(Fig. 38D).

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 8-10 mm; (25) in-
tegument black; antenna, legs, tegula,
venter of metasoma black to rufous; (27)
ocular hairs medium length (about 0.8
mm) ; (28) clypeal margin as in female
(5 above); (31) rest of face closely punc-
tured, rugulose; (32) clypeoantennal dis-
tance greater than antennocular distance,
less than interantennal distance; (33) ocel-
lar area as in female (10 above); (34) in-
terocellar distance greater than ocelloccip-
ital distance, less than ocellocular distance;
(35) vertex as in female (12 above); (37)
hypostomal area of gena covered with
small erect setae through which surface is
visible, anteroventral portion of excava-

Taxonomy of Five Nearctic Subgenera of Coelioxys

719

tion with setae longer than those on disc;
(3(S-43) mandible, scutum, mesepisternum,
scutellum, axilla, metanotal setae as in fe-
male (15-20 above); (45) terga 3-6 with
gradular grooves conspicuous, tergum 6
with dorsal spines reduced (Fig. 381);
(47) sterna 4, 5 as in Figure 390; ster-
num 6 much like Figure 19C, shoulders
slightly more distinct; sternum S with
base broad, apex narrowly rounded (Fig.
39U) ; gonocoxite with dense setae, gono-
base complete, much like Figure 39J.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys soledadensis has been
collected in southern Arizona and New
Mexico (Fig. 43).

This bee is in flight at least from July
5 to October 7.

FIABITAT. Coelioxys soledadensis has
been collected in areas classified as Ari-
zona pine forest ( Finns ), oak-juniper
woodland (Quercus, Jitnipents), creosote
bush-bur sage (Larrea, Franseria), and
grama-tobosa shrubsteppe (Bouteloita, Hi-
laria, Larrea).

COMPARATIVE COMMENTS.
Coelioxys soledadensis most closely resem-
bles C. edita from which it differs by the
shiny vertex, moderately punctured lateral
surface of the mesepisternum, slightly con-
stricted sixth sternum of the female (com-
pare Figs. 38A and 38D), and reduced
dorsal spines of the sixth tergum of the
male (compare Figs. 381 and 38G).

Subgenus Coelioxys Latreille

Coelioxys Latreille, 1809. Genera crusta-
ceorum et insectorum secundum ordinen
naturaleni in familias disposita ... 4:
166.

Type species: Coelioxys quadridentata
Linnaeus.

The subgenus Coelioxys contains C. hir-
sutissima, C. immaculata, C. initchelli, and
C. serricaudaUi. This subgenus differs from
Xerocoelioxys mainly in the reduced carina
of the prothoracic tubercle.

Three hosts, all in different subgenera of
Megachile, ha\e been reported for Coelioxys

sodalis (Table 1). Hosts have not been dis-
ccnered for the remaining species in this
group.

Most ol the characters listed below are
found in other groups ot: New World
Coelioxys ( particularly in Xerocoelioxys and
Boreocoelioxys) and even in some Old World
Liothyrapis, hut never in this combination.
This group is the only North American sub-
genus with reduced carinae of the prothoracic
tubercles, complete graduli on metasomal
terga two and three, sixth metasomal sterna
in females entire (as in Fig. 42B) or con-
stricted (as in Fig. 45E, not notched as in
Fig. 21), and foveal areas of metasomal terga
two in males closely punctured. Italicized
characters below most clearly differentiate
the subgenus Coelioxys from others in North
America.

A. Ocellar area moderately to closely
punctured.

B. Preoccipital carina incomplete medi-
ally.

C. Mandible with inner surface simple
(as in Figs. 45 A, B).

D. Prothoracic tubercle tvith carina dis-
tinct but not expanded into thin plate-li\e
structure.

E. Scutum moderately to closely punc-
tured, fasciae usually indistinct.

F. Scutellum usually rounded posteriorly
(Figs. 45C, D).

G. Axilla usually well produced (Figs.
45C,D).

H. Mesepisternum closely to contiguously
punctured, no minute punctures.

I. Gradular grooves complete on meta-
somal terga tii'o and three.

J. Female: fore coxal spine small, in-
conspicuous.

K. Female: sixth metasomal sternum
tvith margin entire (as in Fig. 42B) or con-
stricted subapically (as in Fig. 45E, F), not
notched as in Figure 21.

L. Male: hypostomal area of gena with
distinct excavation (as in Fig. 5).

M. Male: joveal area of metasomal ter-
gum ttvo closely punctured.

N. Male: fifth metasomal tergum with
lateral spines present but not conspicuous.

O. Male: sixth metasomal tergum with
dorsal spines long (as in Fig. 45G) or short
(as in Fig. 451).

P. Male: seventh metasomal tergum
broadly rounded apically (Fig. 46A).

Q. Male: fourth metasomal sternum sub-

720

The University of Kansas Science Bulletin

Figure 45. Parts of spp. of the subgenus Coelioxys. A-B, Left half of clypeus and mandible. A, C.
sodalis; B, C. serricatidata. C-D, Scutellum and axillae. C, C. sodalis; D, C. immacidata. E-F, Sixth
metasomal tergum and sternum of females. E, C. sodalis; F, C. serricaudata. G-I, Sixth metasomal
tcrgum of males. G, C. sodalis; H, C. serricaudata; I, C. im/nacidafa. ' -

apically carinate, apex bidentate (as in Fig.
26B).

R. Male: seventh sternum represented by
two small sclerites (as in Fig. 19D).

Key to Females of the Subgenus Coelioxys

1. Clypeus strongly convex, in lateral view

almost halt the width of eye; ventral

portion of rest of face convex

serricaudata

Clypeus less convex, in lateral view much
less than half the width of eye; ventral
portion of rest of face flat or slightly
rounded 2

2. Ocular hairs long (about 1.5 mm)

sodalis

Ocular hairs medium or short (about 1.0
to 0.5 mm) 3

3. Sixth metasomal sternum not or little con-

stricted subapically (Fig. 42 B)

tnitchelli

Sixth metasomal sternum conspicuously
constricted subapically 4

4. Clypeus slightly raised, shining, medio-

apically depressed or flat immaculata

Clypeus not raised, slightly rounded ver-
tically htrsHtissima

Key to Males of the Subgenus Coelioxys

1. Metasomal tergum six with dorsal spines
curved inward, upper surfaces sloping
ventrally from outer margin (Fig. 45
H) serricaudata

Metasomal tergum six with dorsal spines
directed posteriorly or slightly outward,
upper surfaces not sloping 2

2. Ocular hairs long (about 0.17 mm)

sodalis

Ocular hairs shorter (0.12-0.05 mm) .... 3

3. Excavation of hypostomal area appearing

bare 4

Excavation at least half covered by short
setae hirsutissima

4. Interocellar distance greater than ocelloc-

cipital distance; coast to coast distribu-
tion mitchelli

Interocellar distance subequal to or slightly
less than ocelloccipital distance; distri-
bution limited to southeastern United
States immaculata

Coelioxys sodalis Cresson
Figures 45 A, C, E, G; 46A-D; 47

Coelioxys qitadridentata; Cresson, 1864,
Proc. Entomol. Soc. Philadelphia 2:
409 (United States, Trenton Falls, mis-
identification); Hicks, 1926, Univ. Col-
orado Stud. 15:226 (Searching behavior,
misidentification) .

Coelioxys sodalis Cresson, 1878, Trans.
Amer. Entomol. Soc. 7:99 {$, New
York, Colorado, type in the Academy
of Natural Sciences of Philadelphia) ;
Mitchell, 1962, North Carolina Agr,

Taxonomy of Five Nearctic Subgenera of Coelioxys

721

Figure 46. Hidden sclcritcs of males of the subgenus Coelioxys. A-D, C. sodalis. A, Seventh tergum; B,

Sixth sternum; C, Eighth sternum; D, Genital armature. E, C. htrstttissima, genital armature. F-G,

C. imiyjacitlata. F. Sixth sternum; G, Eighth sternum. H-I, C. mitchelli. H, Eighth sternum; I,
Genital armature. J, C. serncandata, si.xth sternum.

E.xp. Sta. Tech. Bull. 152:193, 195, 226
(Key, redescription, = C. tristis, = C.
ribis).
Coelioxys ribis Cockerell, 1900, Can.
Entomol. 32:297, 301, 308 {i. New
Mexico, key, type in the National Mu-
seum of Natural History); Graenicher,

1935, Ann. Entomol. Soc. Amer. 28:
300, 304 (Biology); Mitchell, 1962,
North Carolina Agr. Exp. Sta. Tech.
Bull. 152:226 {= C. sodalis).
Coelioxys ribis var. /(incaidi Cockerell,
1904, Ann. Mag. Nat. Hist. (7) 13:33
(9 , S , Washington, type in the Amer-

722

The University of Kansas Science Bulletin

ican Museum of Natural History) ;
Mitchell, 1%2, North Carolina Agr.
Exp. Sta. Tech. Bull. 152:226 (= C.
sod alls).

FEMALE. (1) Length 11-13 mm; (2)
integument very finely rugulose, moder-
ately shiny between punctures; black, an-
tenna, legs, tegula, metasoma black to
piceous; (3) pubescence white, pale och-
raeous on face, mesosoma, golden on
tarsi; (4) ocular hairs long (about 0.15
mm); (5) clypeus convex, almost hidden
by appressed setae, longer, erect, scattered
setae; margin with 6 small denticles (Fig.
45A) ; (6) clypeoantennal distance shorter
than lateral margin of clypeus; (7) para-
ocular area almost hidden by setae up to
0.35 mm long; (S) rest of face rugose
except for a small impunctate median area,
closely punctured up to ocellar area; (9)
clypeoantennal distance subequal to an-
tennocular distance, less than interanten-
nal distance; (10) ocellar area closely
punctate; (11) interocellar distance greater
than ocelloccipital distance, less than ocel-
locular distance; (12) vertex closely punc-
tured; (13) gena narrower than eye, con-
stricted below; ventral angle greater than
90°; (14) hypostomal area of gena with
setae subequal to or longer than those on
disc; (15) mandible with short teeth (Fig.
45A); (16) scutum closely punctured;
erect, slender setae on disc about 0.33 mm
long; (17) lateral surface of mesepister-
num closely punctured, with erect, slender
setae about 0.35 mm long; (IS) scutellum
with posterior margin strongly rounded
(Fig. 45C) or slightly produced, carina
separating dorsal and posterior surfaces
indistinct; (19) axilla sometimes less pro-
duced than shown in Figure 45C; (20)
metanotal setae erect; (21) fore coxal
spine small (about 0.08 mm), triangular,
directed anteriorly; (22) terga 1-5 with
apical fasciae less conspicuous than shown
in Figure 21, punctured as in Figure 21
but with slightly smaller punctures; ter-

gum 6 elongate (Fig. 45E) ; (23) sterna 1
to basal portion of sternum 5 moderately
punctured, apical portion of sternum 5,
sternum 6 very closely punctured; sternum
6 constricted subapically (Fig. 45E).

MALE. (24) Length S-12 mm; (25)
integument as in female (2 above); (26)
pubescence white, golden on tarsi; (27)
ocular hairs long (about 0.15 mm); (28)
clypeus hidden by setae about 0.6 mm
long; surface rugose, shiny; margin as in
female (5 above) ; (29) clypeoantennal
distance as in female (6 above) ; (30)
paraocular area almost hidden by setae up
to 0.6 mm long; (31) rest of face closely
punctured, almost hidden by setae up to
0.6 mm long; (32) clypeoantennal, anten-
nocular, interantennal distances as in fe-
male (9 above) ; (33) ocellar area closely
punctate; (34) interocellar distance sub-
equal to ocelloccipital distance, less than
ocellocular distance; (35) vertex closely
punctured, sometimes with small, irreg-
ular, impunctate areas lateral to ocelli;
(36) gena as in female (13 above); (37)
hypostomal area of gena with distinct
excavation (Fig. 5) covered with setae
shorter than those on disc; (38-39) man-
dible, scutum as in female (15-16 above);
(40) lateral surface of mesepisternum with
setae about 0.45 mm long; (41-43) scutel-
lum, axilla, metanotal setae as in female
(18-20); (44) front coxal spine long
(about 0.33 mm), narrow (about 0.12
mm) with setae on dorsal and ventral
surfaces; rounded apically; directed an-
teriorly; (45) terga 1-5 with fasciae and
punctures as in female (22 above) ; tergum
2 with postgradular area very closely
punctured and rounded to densely punc-
tured and slightly but broadly sunken lat-
erally; tergum 6 with dorsal spines slender
(Fig. 45G) ; tergum 7 slightly protuberant
apically (Fig. 46A); (46) sterna 1-4 mod-
erately punctured, inconspicuous seta in
each puncture; (47) margin of sternum
4, sternum 5 as in Figure 26B except den-

Taxonomy of Five Nearctic Subgenera of Coelioxys

723

160 150 140 130 120 110 100 90 80 70 60 50

Figure 47. Distribution of Coelioxys sodalis.

tides of sternum 4 more rounded; sternum
6 with shoulders inconspicuous (Fig.
46B) ; sternum 7 represented by 2 small
sclerites; sternum 8 with narrow base
(Fig. 46C) ; gonocoxite with long setae,
gonobase complete (Fig. 46D).

HOST RECORDS. Graenicher (1927)
reared Coelioxys sodalis from nests of
Megachile melanophoea wootoni. Medler
(1968) reared C. sodalis from the cells of
M. texana. Hobbs (1968) reared C. sodalis
from a third host bee, M. rotiindata.

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys sodalis, like C. fii-
tieraria and C. tnoesta, ranges from Alaska
to high altitudes in Arizona and eastward
in the north to Nova Scotia (Fig. 47).

This bee is in flight at least from April
17 (Arizona) and June 5 (Yukon Terri-
tory) to September 8 (Alberta).

HABITAT. Coelioxys sodalis has been
collected mainly from areas classified as
coniferous forests, although it does range
down into sagebrush steppe (Artemisia,
Agropyron) in the northwestern United
States. In Canada it has been collected in
the ecotone between tundra and coniferous
forest, in the boreal forest, in the coastal
or moist coniferous forest, and in conif-
erous and deciduous forest combinations
in the east. In the Southwest, C. sodalis
is associated with various coniferous for-
ests dominated mainly by pine (Pinits).

GEOGRAPHIC VARIATION. Spec-
imens of Coelioxys sodalis from the south-
ern part of its range tend to have more
distinct setal fasciae. In addition, the fo-
veal area on the second metasomal tergum
of males tends to be more closely punc-
tured and very slightly and broadly

724

The University of Kansas Science Bulletin

sunken in males from the Southwest.
Western specimens closely resemble speci-
mens of the Old World Coelioxys qitadri-
dentata in that western specimens of C.
sodalis are more coarsely punctured.

COMPARATIVE COMMENTS.
Coelioxys sodalis is a fairly large, dark
bee which resembles C. jitneraria and
dark specimens of C. riifitarsis. Females
of C. sodalis differ from the latter two
species by the unnotched sixth metasomal
sternum. Males of C. sodalis differ from
C. jiineraria and C. riifitarsis by the lack
of a distinct fovea on the second meta-
somal tergum.

Coelioxys hirsittissima Cockerell
Figures 46E, 48

Coelioxys hirsiitissiiyia Cockerell, 1912,
Can.' Entomol. 44:168-169 ( $ , Califor-
nia, key, type in the American Mu-
seum of Natural History).

FEMALE. Agrees with description
and figures of Coelioxys sodalis except as
follows: (1) Length 9-12 mm; (2) in-
tegument black; antenna, venter of meta-
soma black to piceous; mandible piceous
to rufous; legs, tegula, subapical portion
of 6th tergum piceous to ferruginous; (3)
pubescence white, white to pale ochreous
on mesosoma, golden on tarsi; (4) ocular
hairs medium length (about 0.8 mm);
(5) clypeus visible to almost hidden by
short, appressed setae; slightly convex,
margin with 3-5 small denticles; (14)
hypostomal area of gena with setae shorter
than those on rest of gena; (16) scutum
closely punctured anteriorly and laterally,
rest moderately to closely punctured; erect
setae inconspicuous, about 0.2 mm long;
(20) some metanotal setae appressed medi-
ally; (22) terga 1-5 with apical fasciae as
or less conspicuous than shown in Figure
21; (23) sternum 6 with portion distal to
constriction 2/3-1/2 as long as shown in
Figure 45E,

MALE. Agrees with description and
figures of Coelioxys sodalis except as fol-
lows: (24) Length 8-10' mm; (25) in-
tegument as in female (2 above) ; (28)
clypeus hidden by setae about 0.45 mm
long, margin as in female (5 above) ; (30)
paraocular area hidden by setae up to 0.45
mm long; (31) rest of face almost hidden
by setae up to 0.45 mm long; (35) vertex
moderately punctured; (37) excavation of
hypostomal area of gena covered anteriorly
or completely by setae shorter than those
on gena; (39) scutum as in female (16
above); (40) lateral surface of mesepister-
num with setae about 0.35 mm long; (45)
terga 1-5 with fasciae as in female (22
above) ; (47) margin of sternum 4, ster-
num 5 as in Fig. 26B; sternum 6 as in
Fig. 46B, apical margin rounder (as in
Fig. 46F) ; gonocoxites blunter than shown
in Fig. 46D, gonobase simple (Fig. 46E).

DISTRIBUTION AND SEASON OF
FLIGHT. This bee ranges from northern
California southeastward into Texas (Fig.
48) and is in flight at least from March 15
(Texas) to July 16 (California).

HABITAT. Coelioxys hirsiitissima
ranges from areas classified as mixed conif-
erous forest (Abies, Pi mis. Pseiidotsiiga),
red fir forest (Abies), and oak-juniper
woodland (Ouerciis, Juniperiis) to areas
classified as chaparral ( Adenostoma, Arc-
tostaphyllos, Ceanothiis), and coastal sage-
brush {Salvia, Eriogoniim) and into areas
classified as California steppe (Stipa),
creosote bush-bur sage (Larrea, Franseria),
grama-tobosa prairie and shrubsteppe
(Boiiteloiia, Hilaria, Larrea), sagebush
steppe (Artemisia, Agropyron), and trans-
Pecos shrub savana {Floiirensia, Larrea,
also with Juniperiis).

COMPARATIVE COMMENTS.
Coelioxys hirsiitissima resembles C. sodalis
and C. immaciilata closely, but females
differ from C. sodalis by shorter ocular
hairs and brighter coloration of the legs
and from C. immacidata by the flat or

Taxonomy of Fivk Nuarctic Subgenera of Coelioxys

725

120

110

100

90

80

70

w

__[____^

1 \

1

, V

-^-^

/

W— ^J~

\

n

^

\i

x^

^A

fy

~r\

••

nodis : X
immaculata : •
hirsutissima ; ▲

120

110

100

80

Figure 48. Distributions of Coelioxys hirsiittssima (A), C. immaculata (o), and C. nodis (X).

slightly rounded clypeus (slightly raised
in C. immaculata). Males of C. hirsutis-
sima differ from males of C. sodalis by
shorter ocular hairs and brighter color-
ation of the legs and from males of im-
maculata by pubescence on the excavation
of the hypostomal area of the gena (al-
most bare in C. immaculata).

Coelioxys immaculata Cockerell
Figures 45D, I, 46F-G, 48

Coelioxys immaculata Cockerell, 1912,
Can. Entomol. 44:165 ( c5 , Indiana,
type in the American Museum of Nat-
ural History).

Coelioxys sculptijrons Crawford, 1914,
Ann. Entomol. Soc. Amer. 7:153 (9,
New York, type # 18220 in the Na-
tional Museum of Natural Flistory) ;
Mitchell, 1962, North Carolina Agr.
Exp. Sta. Tech. Bull. 152:210 ( = C.
immaculata).

FEMALE. Agrees with description

and figures of Coelioxys sodalis except as
follows: (1) Length 10-13 mm; (2) in-
tegument black; antenna black to rufous;
legs piceous to ferruginous apically; tegula
ferrruginous; (4) ocular hairs medium
(about 0.08 mm long) ; (5) clypeus as in
C. sodalis to bare, shining; raised, medio-
apically flat or slightly depressed; margin
with 5 small denticles; (9) clypeoantennal
distance slightly greater than antennocular
distance, less than interantennal distance;
(10) ocellar area moderately to closely
punctured; (11) interocellar slightly
greater than ocelloccipital distance, less
than ocellocular distance; (12) vertex mod-
erately punctured; (13) gena with ventral
angle less than 90°; (16) scutum closely
punctured anteriorly, posteriorly with 2
submedian slightly swollen, sparsely punc-
tured areas; (17) lateral surface of mese-
pisternum contiguously punctured; erect
slender setae about 0.4 mm long; (18)
scutellum with posterior margin broadly
rounded (Fig. 45D) ; {2i) sternum 6 with

726

The University of Kansas Science Bulletin

portion beyond constriction 2/3 to 1/2 as
long as shown in Figure 45E.

MALE. Agrees with description and
figures of Coelioxys sodalis except as fol-
lows: (24) Length 9-11 mm; (25) in-
tegument as in female (2 above), venter
of metasoma black to piceous; {26) pu-
bescence as in female (3 above) ; (27)
ocular hairs long (about 0.12 mm) ; (28)
clypeus hidden by setae about 0.45 mm
long; margin as in female (5 above); (30)
paraocular area hidden by setae about 0.45
mm long; (31) rest of face with setae
about 0.45 mm long; {^2) clypeoantennal,
antennocular, interantennal distances as
in female (9 above) ; {?>?>) ocellar area
shallowly punctured by large punctures;
{^7) hypostomal area of gena with pos-
terior portion bare; (39) scutum as in fe-
male (16 above), submedian swollen areas
moderately punctured; (41) scutellum
slightly more rounded posteriorly than
shown in Figure 45D; (43) axilla as in
Figure 45C; (45) tergum 6 with dorsal
spines blunt (Fig. 451); (47) margin of
sternum 4, sternum 5 as in Figure 26B;
sternum 6 with thin apical portion broad
(Fig. 46F); sternum 8 broad apically (Fig.
46G).

DISTRIBUTION AND SEASON OF
FLIGHT. This bee ranges throughout
the eastern United States (Fig. 48) and is
in flight at least from March 30 (Florida)
to July 14 (Massachusetts).

HABITAT. Coelioxys immaciilata has
been collected from areas classified as
bluestem prairie ( Andropogon, Paniciim,
Sorghastriim) and cross timbers (Qiiercus,
Andropogon) in the western portion of its
range. Further east it has been taken from
areas classified as various deciduous for-
ests such as Appalachian oak forest (Qiier-
cus), oak-hickory forest (Gary a), north-
eastern oak-pine forest (Piniis), oak-hick-
ory-pine forest, southern mixed forest
(Fagus, Liquidambar, Magnolia, Piniis,

Qiiercus), and southern floodplain forest
(Quercus, Nyssa, Taxodium).

COMPARATIVE COMMENTS.
Coelioxys immaciilata closely resembles C.
sodalis and C. hirsiitissitna but differs
from both by submedial areas of the scu-
tum which are usually conspicuously less
closely punctured than the surrounding
areas. In females the clypeus is slightly
raised and indented medioapically (flat or
slightly convex in C. sodalis and C. hirsii-
tissima).

Coelioxys mitchelli new species
Figures 42B, 43, 46H-I

Coelioxys piercei; Mitchell, 1962, North
Carolina Agr. Exp. Sta. Tech. Bull.
152:217 (misidentification).

FEMALE. Agrees with description
and figures of Coelioxys edita except as
follows: (1) Length 9-10 mm; (2) in-
tegument black; antenna, legs proximal
to tibia piceous; tegula, legs distal to fe-
mur, apical margins of sterna 5, 6 piceous
to fulvous; (3) pubescence white; white
to pale yellow on face, dorsal mesosoma;
golden on tarsi; (4) ocular hairs short
(about 0.06 mm long); (5) clypeus con-
vex, surface with scattered slender setae
0.05-0.1 mm long; margin with three dis-
tinct teeth almost hidden by fringe of
slender setae 0.3 mm wide; (7) paraocular
area hidden by setae about 0.3 mm long;
(8) rest of face visible to hidden; setal
length about 0.25 mm around antenna to
about 0.09 mm medially; (10) ocellar area
closely punctured; (12) vertex moderately
punctured (Fig. 42B); (15) mandible re-
sembles Figure 36B; (16) scutum closely
punctured medially, contiguously punc-
tured laterally (Fig. 42B); (17) lateral
surface of mesepisternum moderately to
closely punctured with large punctures;
fasciae variable; surface bare to almost
hidden by slender setae about 0.3 mm
long; (18) scutellum slightly curved pos-

Taxonomy of Five Nkarctic Subgenera of Coelioxys

111

teriorly (Fig. 42B) to broadly rounded;
(19) axilla not carinate (Fig. 42B); (21)
fore coxal spine about 0.09 mm long, di-
rected anteroventrally; (22) basal fascia of
teriium 1 as or more conspicuous than
shown in Figure 42B; tergum 2 with fo-
veal area not modified; (23) sterna 1-5
fasciate apically (medially on sternum 1
and fascia complete on sternum 5); ster-
num 1 medially closely punctured; ster-
num 6 with lateral margin entire (Fig.
42B).

MALE. Agrees with description and
figures of Coelioxys edita except as fol-
lows: (24) Length 9.5-10 mm; (25) in-
tegument as in female (2 above); (26)
pubescence as in female (3 above); (27)
ocular hairs medium length (0.12 mm);
(28) clypeal margin as in female (5
above); (30) paraocular area hidden by
setae about 0.45 mm long; (31) rest of
face with setae about 0.45 mm long; (32)
clypeoantennal distance less than inter-
antennal distance, greater than antennocu-
lar distance; {?)i) ocellar area as in female
(10 above); (34) interocellar distance
greater than ocelloccipital distance, less
than ocellocular distance; (35) vertex as
in female (12 above); (37) posterior por-
tion of excavation of hypostomal area of
gena almost bare; (38) mandibles as in
female (15 above); (39) scutum without
fasciae; (40-42) mesepisternum as in fe-
male (17-19 above) ; (43) metanotum with
few appressed setae medially; (45) gradu-
lar grooves of terga 4-6 with conspicuous
fasciae; tergum 6 resembles Figure 39K;
(46) fasciae complete on sterna 1-4; (47)
margin of sternum 4, sternum 5 as in Fig-
ure 19B; sternum 6 as in Figure 26C;
sternum 8 as in Figure 46FI; genital arma-
ture with small projection on gonobase
(arrow on Fig. 461).

DISTRIBUTION AND SEASON OF
FLIGHT. Coelioxys mitchelli ranges from
coast to coast (Fig. 43).

This bee is in flight from at least April

1 (Florida) and April 24 (southern Cali-
fornia) to June 22 (Nevada).

HABITAT. In the East, Coelioxys
mitchelli has been collected in areas classi-
fied as oak-hickory-pine (Qtiercus, Carya,
Finns) and southern mixed (Fagiis, Liq-
itidambar, Magnolia, Pinus, Quercus) for-
ests. In Kansas it was taken from blue-
stem-grama prairie (Andropogon, Boiite-
loita). Further west C. mitchelli was
taken from areas classified as sagebrush
steppe (Artemisia, Agropyron), creosote
bush-bur sage (Larrea, Franseria), creo-
sote bush, and saltbush-greasewood (Atri-
plex, Sarcobatus) vegetation types.

GEOGRAPHIC VARIATION. West-
ern specimens of Coelioxys 7nitchelli in
general have lighter colored tegulae and
legs as well as denser, more conspicuous
pubescence on the face, lateral surface of
the mesepisternum, and metasoma.

COMPARATIVE COMMENTS.
Coelioxys mitchelli closely resembles C.
mesae (for this reason the description was
compared with C. edita rather than C.
sodalis) but differs by its smaller size,
shinier integument, and lesser quantity of
appressed setae on the scutum, posterior
surface of scutellum, and on the meta-
notum. Western specimens of C. mitchelli
possess ferruginous or fulvous tegulae and
legs whereas those of C. mesae are black
or piceous.

TYPES. Holotype female, Raleigh,
North Carolina, May 23, 1959 (T. B.
Mitchell). Allotype male. Wake Co.,
N.C., May 22, 1960, on Oenothera (T. B.
Mitchell), both deposited in the North
Carolina State University Museum. Para-
tvpes females, Raleigh, N.C., May 1930,
Mav 19, 1932, Mav 9, 1935 (T. B. Mitch-
ell)', May 23, 1959, May 18, 1922, May
17, 1951, May 25, 1933, last three of
Oenothera, Wake Co., N.C., May 21, 1960,
May 26, 1960; males, 3 from Tallahassee,
Florida, April 1, 1944 (R. & G. Bohart),
Raleigh, N.C., May 23, 1922, on Oenothera

728

The University of Kansas Science Bulletin

(T. B. Mitchell). Paratypes are deposited
in the collections of the North Carolina
State University Museum, Utah State Uni-
versity, the National Museum of Natural
History, Snow Entomological Museum,
University of Kansas, and the Florida
State Collection of Arthropods, Division
of Plant Industry.

This species is named in honor of Pro-
fessor emeritus T. B. Mitchell of the
North Carolina State University.

Coelioxys serricaitdata, new species
Figures 45B, F, H, 46J, 49

FEMALE. Agrees with description
and figures of Coelioxys sodalis except as
follows: (1) Length 10-12 mm; (2) in-
tegument black; antenna, legs, tegula,
venter of metasoma black to piceous; (3)
pubescence white; white to golden on
head, mesosoma; golden on tarsi; (4) ocu-
lar hairs medium (0.08 mm long); (5)
clypeus with slender, inconspicuous setae;
strongly convex; closely punctured, shiny;
margin with 3-5 denticles (Fig. 45B); (8)
ventral portion of rest of face convex; (9)
clypeoantennal distance greater than an-
tennocular distance, shorter than inter-
antennal distance; (14) hypostomal area
of gena with setae subequal or shorter
than those on disc; (15) mandible with
hump on outer surface (Fig. 45B) ; (17)
lateral surface of the mesepisternum with
slender, prostrate setae up to 0.45 mm
long; (18) scutellum similar to Figure
45D; (19) axilla as in Figure 37D; (20)
metanotum with setae prostrate medially;
(22) terga 1-5 with apical fasciae similar
to or more conspicuous than those shown
in Figure 21; uniformly, closely punctured
with smaller punctures than shown in
Figure 21; tergum 6 very closely punc-
tured (Fig. 45F) ; (23) sterna 1 to basal
portion of 5 closely, uniformly punctured;
most of sternum 5, 6 very closely punc-
tured; sternum 6 with subapical margin

serrate (Fig. 45F, serrations visible in sil-
houette against light source).

MALE. Agrees with description and
figures of Coelioxys sodalis except as fol-
lows: (24) Length 9-11 mm; (25-26) in-
tegument, pubescence as in female (2-3
above) ; (28) margin of clypeus as in fe-
male (5 above); (32) clypeoantennal, an-
tennocular, interantennal distances as in
female (9 above); (35) vertex closely
punctured; (38) mandible as in female
(15 above); (41-42) scutellum, axilla as in
female (18-19 above); (45) terga 1-5 with
fasciae and punctures as in female (22
above) ; tergum 6 with dorsal spines
curved and sloped medially (Fig. 45H);
(47) margin of sternum 4 as in Fig. 390,
lacking longitudinal carinae; sternum 6
with shoulders distinct (Fig. 46J) ; gono-
coxites with setae 2/3 length shown in
Fig. 46E.

DISTRIBUTION AND SEASON OF
FLIGHT. This bee has been collected
from May 5 to July 6 in the Pacific coast
states (Fig. 49).

HABITAT. Coelioxys serricaitdata has
been taken from areas classified as mixed
conifer forest {Abies, Pinits, Pseitdotsiiga),
lodgepole pine-subalpine forest (Piniis,
Tsiiga). ponderosa pine shrub forest, and
California mixed evergreen forest (Qiter-
cits, Arbiitits. Pseiidotsiiga); areas classi-
fied as open grassland such as fescue-oat-
grass {Festitca. Danthonia), fescue-wheat-
grass ( Agropyron), California steppe
(Stipa), tule marshes (Scirpits, Typha),
and alpine meadows and barrens (Agro-
stis, Carex, Festitca, Poa).

COMPARATIVE COMMENTS.

Females of Coelioxys serricaitdata are
slender, dark bees which resemble C.
moesta, but differ from all other North
American Coelioxys by the strongly con-
vex clypeus and serrated margin of the
sixth metasomal sternum (Fig. 45F).
Males of C. serricaitdata are darker than
most Coelioxys males and differ from all

Taxonomy of Five Nearctic Subgenera of Coelioxys

729

Figure 49. Distributum of Coelioxys scnicuuthilu.

Other North American Coelioxys by the
dorsal spines of the sixth metasomal ter-
gum which are strongly curved and sloped
medially.

TYPES. Holotype female, Hastings
Nat. Hist. Reservation, Santa Lucia Mts.,
Jamesburg, Monterrey Co., California,
1900-2700 ft., June 12,' 1938 (C. D. Mich-
ener), deposited in Snow Entomological
Museum, the University of Kansas. Para-
types females, one. Prospect, Oregon, June
20, 1924 (C. L. Fox); one, Antelope Mt.
L. O., Grant Co., Oregon, El. 6500 ft.,
August 7, 1941 (M. & R. E. Rieder) ; one,
Corvallis, Oregon, July 1, 1910 (J. C. Brid-
well); one, Palouse, Washington, June 26,
1961 (R. W. Dawson); the remaining
paratypes are from California; one. Tan-
bark Flat, Los Angeles Co., June 20, 1956

(R. C. Bechtel) ; one, Twaine-Harte,
Tuolumne Co., 4000 ft., July 1937, Grin-
delia sp. (F. E. Blaisdell) ; one, Chile Bar,
Eldorado Co., July 5, 1948 (C. Chan);
Redwood City, San Mateo Co., June 10,
1961 (P. H. Arnaud, Jr.); one, 16 mi. E.
Glenville, Kern Co., June 25, 1961 (R. L.
Macdonald) ; one. Mormon Bar, 2 mi. S.,
Mariposa Co., June 16, 1959 (G. L Stage) ;
one, Miguel Meadows, Yosemite National
Park, Elev. 5200, July 6, 1940 (E. G.
Linsley); one. Pine Flat, Tulare Co., June
14, 1961 (G. L Stage, R. R. Snelling) ;
one, Ryan Creek, Mendocino Co., June 27,
1954, Godetia amoena (P. D. Hurd) ; one,
Nipinnawasee, Madera Co., July 4, 1960
(G. I. Stage, R. R. Snelling); one, Santa
Margarita, 6 mi. N. E., S. L. Obispo Co.,
June 22, 1958 (E. G. Linsley); two, Ar-

730

The University of Kansas Science Bulletin

royo Seco, Monterey Co., May 2, 1959
(Don Burdock), May 5, 1956 (Dave Kib-
ble); two, California Hot Springs, Tulare
Co., June 4, 1934 (E. C. VanDyke), (E. R.
Leach); five, Antioch, Contra Costa Co.,
two, May 6, 1939, one. May 16, 1936 (M.
Cazier), one, April 25, 1936 (M. Cazier),
one, June 23, 1957 (G. I. Stage). Paratypes
are in collections of the California Acad-
emy of Sciences; the University of Cali-
fornia at Berkeley and Davis; the Los
Angeles County Museum; the National
Natural History Museum; San Jose State
College; Utah State University; G. I.
Stage, the University of Connecticut; and
P. H. Timberlake, the University of Cali-
fornia, Riverside.

The specific name refers to the serrate
margin of the apical sternum of the
female.

LITERATURE CITED

Baker, J. R. 1971. Development and sexual di-
morphism of larvae of the bee genus Coeli-
oxys. J. Kansas Entomol. Soc. 44:225-235.

Baker, }. R. 1972. The biology and systematics of
certain nearctic groups of Coelioxys (Hyme-
noptera: Megachilidae). Thesis (deposited in
the Graduate School, the University of Kan-
sas). 270 p.

Bechtel, R. C. 1958. Notes on emergence and a
parasite of Megachile gentilis Cresson (Hyme-
noptera: Megachilidae). Pan-Pacific Entomol.
34:12-13.

COCKERELL, T. D. A. AND W. W. RoBBINS. 1910.

An introduction to the study of Rocky Moun-
tain bees. Univ. Colorado Stud. 7:179-195.

Dalla Torre, C. G. de, and H. Friese. 1894.
Synonymischer katalog der curopacischen
Schmarotzerbienen. Entomol. Nachrichten 20:
33-43.

Fox, W. J. 1900. Coelioxys i-dentata and its host.
Entomol. News 11:553.

Fye, R. E. 1965. Biology of Apoidea taken in trap
nests in northwestern Ontario (Hymenop-
tera). Can. Entomol. 97:863-877.

Graenicher, S. 1905. Some observations on the life
history and habits of parasitic bees. Bull.
Wisconsin Natur. Hist. Soc. 3:153-167.

Graenicher, S. 1927. On the biology of the para-
sitic bees of the genus Coelioxys (Hymen.,
Megachilidae). Entomol. News 38:231- 235,
273-276.

Graenicher, S. 1935. Bee-fauna and vegetation of
Wisconsin. Ann. Amer. Entomol. Soc. 28:
285-310.

Hicks, C. H. 1926. Nesting habits and parasites of
certain bees of Boulder County, Colorado.
Univ. Colorado Stud. 15:217-252.

HoBBS, G. A. 1956. Ecology of the leaf-cutter bee
Megachile perihirta Ckll. (Hymenoptera:
Megachilidae) in relation to production of
alfalfa seed. Can. Entomol. 88:625-631.

HoBBs, G. A. 1968. Controlling insect enemies of
the alfalfa leaf-cutter bee, Megachile rotiin-
data. Can. Entomol. 100:781-784.

Krombein, K. V. 1967. Trap-nesting wasps and
bees: Life histories, nests and associates.
Smithsonian Publication 4670. Smithsonian
Press, Washington. D.C. 570 p.

KiJcHLER, A. W. 1964. Potential natural vegetation
of the conterminous United States. (Map with
manual). Amer. Geogr. Soc. Spec. Pub. 36:
Legend, Map. 38 pp.

Latreille, p. a. 1809. Genera Crustaceorum In-
sectorum secundum ordinen naturalem in
familias disposita, inconibus exemplisque
plurimis explicata. 4. A. Koenig, Paris and
Argentorati. 399 p.

Medler, J. T. 1958. A note on Megachile iiiciniis
Prov. in trap nests in Wisconsin (Hymenop-
tera: Megachilidae). Can. Entomol. 90:325-
327.

Medler, J. T. 1959. A note on Megachile cen-
tiinadaris (Linn.) in Wisconsin (Hymenop-
tera: Megachilidae). Can. Entomol. 91:113-
115.

Medler, J. T. 1964. A note on Megachile (Sayapis)
ptignatc piignata Say in trap nests in Wiscon-
sin (hymenoptera: Megachilidae). Can.
Entomol. 96:^)18-921.

Medler, J. T. 1965. A note on Megachile mendica
Cresson in trap nests in Wisconsin (Hyme-
noptera: Megachilidae). Proc. Entomol. Soc.
Washington 67:113-116.

Medler, J. T., and T. W. Koerber. 1958. Biology
of Megachile relatica Cresson (Hymenoptera:
Megachilidae) in trap nests in Wisconsin.
Ann. Entomol. Soc. Amer. 51:337-344.

Medler, J. T., and J. F. Lussenhop. 1968. Leaf-
cutter bees of Wisconsin (Hymenoptera:
Megachilidae). Univ. Wisconsin Res. Bull.
274:1-80.

Mil HENER, C. D. 1953. The biology of a leaf-cutter
bee (Megachile hrcvis) and its associates.
Univ. Kansas Sci. Bull. 35:1659-1748.

Mitchell, T. B. 1962. Bees of the eastern United
States. II. North Carolina Agr. Exp. Sta.
Tech. Bull. 152: 557 pp.

Mitchell, T. B. 1973. A subgeneric revision of
the bees of the genus Coelioxys of the Western
Hemisphere. Contrib. Dep. Entomol. North
Carolina State Univ. 129 p.

Robertson, C. 1903. American Hymenoptera.
Synopsis of Megachilidae and Bombinae.
Trans. Amer. Entomol. Soc. 29:163-178.

Robertson, C. 1926. Phenology of inquiline and
nest-making bees. Psyche 33:116-120.

Robertson, C. 1929. Flowers and insects. Lan-
caster, Pennsylvania. Science Press: 217 pp.

Shuckard, W. E. 1866. British bees: an introduc-
tion to the study of the natural history and
economy of the bees indigenous to the British
Islands. L. Reeve and Company, London.
371 p.

Stephen, W. P., G. E. Bohart, and P. F. Torchio.
1969. The biology and external morphology
of bees with a synopsis of the genera of
northwestern America. Agr. Exp. Sta. Oregon
State Univ. Corvallis. 140 p.

I

Iv

i

g

ft:

.V

i

$:

I

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

r^

Y

1

ii:

PARASITID MITES ASSOCIATED WITH

BUMBLEBEES IN ALBERTA, CANADA

(ACARINA: PARASITIDAE;

HYMENOPTERA: APIDAE). I.

TAXONOMY

i

s

5:

By

S

ss

LYNN A. RICHARDS

$

S

ft: $
ft: $

S!.

^

s

s

s

ft:

5:

!g

Vol. 50, No. 13, pp. 731-773

February 4, 1976

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out at
the University of Kansas or by University faculty and students. Since its incep-
tion, volumes of the Bulletin have been variously issued as single bound volumes,
as two or three multi-paper parts or as series of individual papers. Issuance is at
irregular intervals, with each volume prior to volume 50 approximately 1000
pages in length. Effective with volume 50, page size has been enlarged, reducing
the length of each volume to about 750 pages.

The supply of all volumes of the Kansas University Quarterly is now ex-
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased at
$20.00 per volume. Where some of these volumes were issued in parts, individual
parts are priced at the rate of 2 cents per page. Current policy, initiated with
volume 46, is to issue individual papers as published. Such separata may be
purchased individually at the rate of 3 cents per page, with a minimum charge
of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at
the rate of $20.00 per volume. All communications regarding exchanges, sales
and subscriptions should be addressed to the Exchange Librarian, University
OF Kansas Libraries, Lawrence, Kansas 66045.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

The International Standard Serial Number of this publication is US ISSN
0022-8850.

«

Editor
Eugene C. Bovee

Editorial Board

William L. Bloom
Philip W. Hedrick

Rudolf Jander

Harvey Lillywhite

Charles D. Michener

Norman A. Slade

Henry D. Stone

George W. Byers, Chairman

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. 50, No. 13, pp. 731-773 February 4, 1976

Parasitid Mites Associated with Bumblebees in
Alberta, Canada (Acarina: Parasitidae; Hymenoptera:

Apidae). I. Taxonomy^

Lynn A. Richards

TABLE OF CONTENTS

Abstract 731

Introduction 731

Methods and Materials 732

Descriptions of Species 732

Panisitus hobbsi new species 733

Parasitus javus new species 735

Parasitus inc]Hilinobombus new species 736

Parasitus perthecatus new species 738

Parasitus fimetorum Berlese, 1903 741

Literature Cited 741

Figures 743-773

P. hobbsi n. sp.. Fig., i-7 _ 743-748

P. javus n. sp., Figs. 8-14 749-756

P. inquilinobombns n. sp., Figs. 14 21 I'^l-ld'i

P. perthecatus n. sp.. Figs. 22-31 766-773

ABSTRACT

Four new species of mites in the Genus Parasitus, found associated with bumblebees
of the Genus Bombus in Alberta, Canada, are described. They are: Parasitus hobbsi;
P. javus; P. inqudinobombus; P. perthecatus. For each of these new species the deuto-
nymph, female and male are described and for P. perthecatus the protonvmph and larva,
also. A fifth species P. jimetorum Berlese, 1903, fcjund in nests of bumblebees of the
Genus Bombus in Alberta, Canada, is also briefly discussed.

INTRODUCTION

The purpose of this study is to describe August 25, 1970, and April 25 through

new species of mites of the Genus Para- August 25, 1971, on the east side of Prairie

situs found with the Canadian bumblebees Bluf? Mountain, 49" 20' N, 114" 06' W,

of the Genus Bombus, species listed below. at about 5200 feet elevation. The biology

Field work was done in southwestern of these new species of Parasitus is to be

Alberta, Canada, from May 26 through reported in Richards and Richards (in

7 T" . press).

Contribution number 1529, Department of Entomol-
ogy, University of Kansas, Lawrence, Kansas, 66045. Previous work on ParasitUS associated

731

732

The University of Kansas Science Bulletin

with Bombiis has mainly been taxonomic:
Oudemans (1902, 1903, 'l904, 1905, 1906),
Berlese (1903, 1905), Tragardh (1904,
1910, 1912), Vitzthum (1930), Willmann
(1939), Schweizer (1949), Karg (1965),
Holzmann (1969), and Micherdzinski
(1969).

Bumblebees of the following species of
Bom bus and Psithynts were encountered
in the study area: B. appositus Cresson;
B. bifariits nearcticits Handlirsch; B. culi-
fornirus F. Smith; B. centralis Cresson;
B. fJcwifrons Cresson; B. frigid us F.
Smith; B. Iiuntii Greene; B. melanopygits
Nylander; B. mixtiis Cresson; B. nevaden-
sis Cockerell; B. occidentalis Greene; B.
riijocinctiis Cresson; B. ternariiis Say; B.
terricola Kirby; B. vagans F. Smith; P.
insidaris (F. Smith); P. siicl{leyi (Greene);
and P. jernaldae Franklin.

METHODS AND MATERIALS

Mites were mounted and cleared in
Hoyer's solution. Measurements in milli-
meters were taken from the specimens
used in figures, employing an ocular
micrometer. The measurements given are
maximal lengths and widths. No attempt
was made to measure more specimens be-
cause the measurements have only general
value. Differences in size are often caused
by handling (i.e., pressure applied to the
cover slip; amount of medium used; time
between collecting and preserving) and
by the condition of each mite at the time
it was collected (e.g., starved, engorged).
When males or females are referred to in
the text it can be assumed that they are
adults. Nomenclature for setae follows
Hirschmann (1957).

DESCRIPTIONS OF SPECIES

Mites in the family Parasitidae are
heavily sclerotized, with the dorsal shield
entire or divided into two plates. The
palpal tarsus has a three-pronged seta.

The female has a triangular genital shield
with its anterior apex flanked by large
metasternal plates. The male has a large
apophysis on leg II, and the spermato-
phoral process on the moveable digit of
the chelicera is fused to the distal area
of the digit.

Genus Parasitits Latreille, 1795

Both sexes have divided dorsal shields,
generally with 21 to 24 pairs of setae on
the propodosoma and 20 to 70 pairs on
the hysterosoma. Dorsal chaetotaxy of the
propodosomal shield in larvae resembles
that of Pergamasits Berlese, 1903, with
nine pairs of setae. Larval hysterosoma
have eight pairs of setae (Micherdzinski
1969). Protonymphs resemble those of
Pergamasits. Micherdzinski (1969) found
14 propodosomal setae and 15 hysterosomal
setae on protonymphs of Parasitits f^rae-
pelini Berlese. Sellnick (1938) and Karg
(1969) record more setae for Parasitits
immanits Berlese and P. remberti Oude-
mans than are noted in previous proto-
nymphal descriptions. Dorsal chaetotaxy
of deutonymphs is similar to that of
adults.

The epigynum of females in many
species is not clearly separated from the
ventro-anal shield. The vaginal sclerite is
either heavily sclerotized or present as a
weakly defined element. The male genital
opening is set far back from the sternal
shield. The gnathosoma of males is some-
times modified, with cleft corniculi and
nontypical rows of deuterosternal teeth.
The digitus mobilus of female chelicera
have one to three teeth, seldom four; of
the male, zero to four teeth, seldom three.
The anal shield of the larva has three
setae (i.e., an adanal pair and a long post-
anal hair). There are three pairs of sternal
setae, four pairs of setae between coxa III.
An anal shield is present; stigma and
paratreme, absent. The protonymph has
three pairs of sternal setae, two lyrifissures,

Parasitid Mites Associated with Bumblebees

733

one pair of genital setae, five pairs of
opistogastral setae, and three anal setae.
A stigma and short paratremes are pres-
ent, hut the anal shield and paratremal
shields are not always clearly defined.
Deutonymphs have pre-endopodal shields
and a distinct sternal shield with four
pairs of setae and three lyrifissures.

Le^s II and III are shorter than I and
IV; legs II are usually thicker than I, III,
or W. Tarsus I of most deutonymphs is
divided (distinguishing this genus from
other genera in the family, except Poechlo-
c hints) .

Parasitus hobbsi, new species

DEUTONYMPH (Fig. 1): Idiosomal
length 0.78 mm, width 0.62 mm. Propodo-
soma with 21 pairs of pilose setae; m 5
and i 2 longer; m 1, 2, and 3 shorter than
other setae on propodosomal shield.
Hysterosomal shield rounded posteriorly.
Sculpturing on dorsal shields netlike.

Pre-endopodal shield rectangular, 0.03
mm long, 0.01 mm wide. Front border
of sternal shield distinct, sclerotization
heavier behind first pair of pores. Sternal
setae smooth. Sculpturing of shield, net-
like polygons. Metapodal shields comma-
shaped, with truncate ends; 0.05 mm long.

Tectum (Fig. 2C) three pointed, mid-
dle point broad, blunt, all about same
length. Chelicera (Fig. 2B): Digitus mo-
biliis with three large teeth; digitus fixtus
with three large and two small teeth.

Length of leg I, 1.04 mm; II, 0.80 mm;
III, 0.80 mm; IV, 1.17 mm. Leg II (Fig.
2D) setae pilose; leg IV (Fig. 2E) setae
sparsely pilose or smooth, longer than
setae of leg II; tarsus IV pd 2 longest
setae on entire leg.

FEMALE (Fig. 3): Idiosomal length
1.16 mm, width 0.83 mm. Propodosomal
shield wider than long, with 21 pairs of
setae; all setae pilose except m 1, 2, and
3; m 5 longer than other setae; m 1, 2,
and 3 shorter than 4. Hysterosomal shield

and propodosomal shield separate. Hys-
terosomal shield shorter than propodo-
somal shield, semicircular. All hystero-
somal setae about the same length, pilose.
Sculpturing of dorsal shield, netlike.
Peripheral lines of propodosomal shield
form netlike pattern, becoming closer
together, following edge of plate; lines
indistinct nearer center.

First pair of sternal setae borne ofT
heavily sclerotized part of sternal shield.
All ventral setae pilose, except five pairs
of short setae and adanal setae. Ventro-
anal plate, V-shaped and fused with para-
tremal shield, has polygonal netlike sculp-
turing. Metasternal shield and anterior
halves of ventro-anal shield have netlike
sculpturing.

Gnathosoma (Fig. 4B) has 11 rows of
deuterosternal teeth. Tectum (Fig. 4A)
three-pointed, the middle point longest.
G 1 smooth, G 2, 3, and 4 pilose. Che-
licera (Fig. 4C): Digitus mobilus has
three teeth in upper half, two more apical
about the same size, third larger; digitus
fixtus has four teeth, the middle two
larger.

Length of leg I, 1.35 mm; II, 1.01 mm;
III, 1.04 mm; IV, 1.54 mm (Fig. 5B, C).
Setae pilose; leg II (Fig. 5A) setae needle-
like, except tarsus av 1 and pv 1, which
are short and thornlike; setae on tarsus
longer than other leg setae, especially pd
2, except al 1 and pi 1 which are short
and needlelike.

MALE (Fig. 6): Idiosomal length 1.11
mm, width 0.80 mm. Twenty-two pairs
of setae on propodosomal shield, all pilose
except m 1, 2, and 3; m 5 longer than
other setae; m 1, 2, and 3 shorter than 4.
Line dividing propodosomal shield from
hysterosomal shield extends to body mar-
gins. All hysterosomal setae pilose. Dor-
sal sculpturing like that of female.

Ventral setae pilose. Tritosternum,
base about '/b the total length, with lacinia
pile long. Sculpturing over entire sclero-

734

The University of Kansas Science Bulletin

tized ventral area has a polygonal netlike
pattern.

Gnathosoma (Fig. 7E) has 11 rows of
deuterosternal teeth. Tectum (Fig. 7D)
varies, has one to three points; only G4
pilose. Chelicera (Fig. 2A): Digitus mo-
bilits curved apically, with one tooth;
digitus jixtus curved apically, with two
teeth; cheliceral brush not stalked.

Length of leg I, 1.35 mm; II, 1.04 mm;
III, 0.98 mm; IV, 1.43 mm. Apophysis
on femur II well developed, axillary proc-
ess rounded. Genu and tibia II have
smaller, rounded apophyses. Leg setae
mostly smooth, some sparsely pilose. Leg
II (Fig. 7C) and IV (Fig. 7A, B), setae
needlelike; tarsus IV, setae longer, espe-
cially pd 2.

P. hobbsi resembles P. traegardhi
(Micherdzinski, 1969) (= bombophilus
Vitzthum, 1930), but differs in the follow-
ing ways: The female hysterosomal shield
is shorter than the propodosomal shield.
The dorsal hysterosomal setae are the
same length throughout: they are not
longer posteriorly as they are in P. trae-
gardhi. The male of ?. hobbsi is similar
to the male of P. traegardhi, but P. hobbsi
has pilose setae, whereas P. traegardhi has
smooth setae. The setae on and off the
hysterosomal shield of the deutonymph
are about the same length, with lateral
setae being somewhat shorter, whereas in
P. traegardhi the setae off the shield are
shorter.

Adult mites are found in Bombus
nests; deutonymphs in nests and on adult
bumblebees.

TYPES: Holotype of the deutonymph
collected 9 August 1970, at Prairie Bluff
Mountain, Alberta, Canada (49° 20' N,
114° 06' W), 5200 feet, from Bombus
bifarius nest; collector L. A. Richards.
Specimen located in Canadian National
Collection (CNC) no. 13187. Morphotype
of the male collected 9 August 1970, and

of the female collected 7 August 1971,
both from B. occidentalis nests; same
locality and collector as the holotype;
both specimens in CNC. Paratypes, all
from same locality as holotype. Eighteen
deutonymphs: Five collected 11 August
1971, from B. occidentalis nest; three 20
August 1971, from B. occidentalis nest;
two 14 August 1970, from B. calijornicus
nest; one 9 August 1971, from B. bifarius
nest; one 14 August 1971, from B. bifarius
nest; one 8 August 1970, from B. bifarius
nest; one 17 July 1970, from B. occidentalis
nest; one 18 July 1970, from B. huntii
queen; one 8 July 1971, from B. figidus
nest; one 9 June 1970, from B. huntii
queen; and one 19 August 1970, from
B. calif or nic us nest. Twelve females:
Three collected 11 August 1971, from B.
occidentalis nest; three 20 August 1971,
from B. occidentalis nest; two 11 August
1971, from Bombus nest; one 14 August
1971, from B. bifarius nest; one 22 August
1971, from Bombus nest; two 14 August

1970, from B. califomicus nest and Bom-
bus sp. nest. Fourteen males: Five col-
lected 20 August 1971, from B. occidentalis
nest; four 11 August 1971, from B. occi-
dentalis nest; one 9 August 1971, from
B. bifarius nest; one 9 August 1970, from
B. frigid us nest; one 22 August 1970,
from B. mixtus nest; one 2 August 1970,
from B. frigid us nest; and one 22 August

1971, from Bombus nest. Five deuto-
nymphs, three females and four males,
are deposited in the CNC. Four deuto-
nymphs, three females and three males,
are deposited in the British Museum
(Natural History) (BMNH). Four deu-
tonymphs, three females and three males,
are deposited in the United States Na-
tional Museum (USNM). Five deuto-
nymphs, three females and four males,
are deposited at the University of Kansas.

The species was named in honor of
Dr. Gordon A. Hobbs for his help and
encouragement with this project.

Parasitid Mites Associated with Bumblebees

735

Parasitiis juvits. new species

DEUTONYMPH (Fig. S): Idiosomal
length 0.77 mm, width 0.59 mm. Propo-
dosomal and hysterosomal shields have
scalelike surface; m 5 and i 2 longer than
other setae; m 5 and d 1 pilose. All hys-
terosomal setae smooth and needlelike.
Setae off the shield about the same length
as the short setae on the shield.

Pre-endopodal shields triangular. Front
borders of sternal shield indistinct, heavily
sclerotized part beginning at first pair of
pores. First pair of sternal setae pilose,
second pair with scarce pile or none, all
others smooth. Metapodal shields oval,
0.03 mm long.

Tectum (Fig. 9B) three-pointed, mid-
dle point longer and wider than others.
Chelicera (Fig. 9A) : Digitus mobilus has
three teeth, two in upper half of digit
about ^/i size of third tooth in middle of
digit; digitus fixtus has four teeth, all
similar in size.

Length of leg I, 0.83 mm; II, 0.68 mm;
III, 0.68 mm; IV, 0.98 mm. Leg II (Fig.
9C), setae needlelike, except av 1 and pv
1 of tarsi, which are thornlike. All setae
about same length, pilose. Leg IV (Fig.
9D), setae needlelike, pilose, al 3, pd 3,
and av 2 very long. All coxal setae pilose.

FEMALE (Fig. 10): Idiosomal length
1.29 mm, width 0.88 mm. Propodosomal
shield has 21 pairs of setae; m 5 and i 2
longer than other setae on shield; m 1,
2, and 3 shorter, than 4. Setae sparsely
pilose. Propodosomal shield wider than
long; separated from hysterosomal shield
by a slightly concave line. Propodosomal
shield wider than hysterosomal shield,
about the same length. Dorsal sculptur-
ing, netlike. Hysterosomal shield as wide
as long, triangular. Setae on hysterosoma
all about same length, most smooth; some
on membranous portion of dorsum, pilose.

Ventro-anal shield fused with para-
tremal shield, rounded posteriorly, pointed

posterior to anus, has nestlike sculpturing.
Ventral setae, moderately to sparsely
pilose.

Gnathosoma (Fig. 11 A) has 10 rows
of deutostcrnal teeth. Tectum (Fig. IIB)
has three points, middle one larger and
wider than lateral ones. G 1, 2 smooth,
CJ 3, 4 pilose. Chelicera (Fig. 12C): Each
digit with three teeth set wide apart.

Length of leg I, 0.98 mm; II, 0.83 mm;
III, 0.80 mm; IV, 1.12 mm. Setae thick;
pd 2 of genu II (Fig. 12B) longer than
other setae; av 1 and pv 1 of tarsi II, III,
IV thornlike, other tarsal setae longer than
setae on rest of leg. Tibia and trochanter
pv 1 thicker than other setae of these seg-
ments. Leg IV (Fig. 12A) setae needle-
like, ad 3 and pd 2 of tarsus IV very long.

MALE (Fig. 13): Idiosomal length
0.98 mm, width 0.68 mm. Propodosomal
shield has 22 pairs of setae, all smooth;
m 5 and i 2 lonirer than other setae.
Hysterosomal and propodosomal shield
divided by a concave line. Dorsal sctilp-
turing similar to that of female. Hystero-
soma polytrichous; setae smooth.

Tritosternum slender, base about same
length as lacinia. First pair of sternal
setae borne ofT heavily sclerotized portion
of sternal shield. First pair of sternal
setae pilose. Ventro-anal and sternal
shields have netlike sculpturing.

Gnathosoma (Fig. IIC) has nine rows
of deuterosternal teeth. Tectum (Fig.
IID) has one large, rounded point. G 1,
2, and 3 smooth, G 4 pilose. Chelicera
(Fig. 14C): Digitus mobilus Y^ the length
of the digitus fixtus, curved apically, with
one tooth; digitus fixtus slender, with
three widely spaced teeth, cheliceral brush
on a short stalk.

Length of leg I, 0.98 mm; II, 0.80 mm
(Fig. 14B); III, 0.74 mm; IV, 1.04 mm
(Fig. 14A). Femur II has well-developed
apophysis on ventral side. Genu and
tibia II also have small rounded apophyses,
tarsal setae longer except av 1 and pv 1

736

The University of Kansas Science Bulletin

which are thornHke, pd 2 of tarsus very
long.

P. jai'its appears to be closely related
to P. nitmismaticiis Vitzthum 1930, ac-
cording to the description by Micherdzin-
ski (1969). The description of the dorsum
of P. niimismaticiis by Vitzthum (1930)
differs from that by Micherdzinski (1969)
as having smooth, dorsal setae which are
shorter towards the posterior. The ventral
surface of the female of javiis is similar
to that of P. numismaticus except that
areas of small circles on sternal plate are
absent and the epigenial shield has a
broad elliptical apex (in P. numismaticus
it is narrow). The paragynial shield is
like that of P. numismaticus, but the setae
are smooth. In P. numismaticus only G
4 is pilose, but in P. javus both G 3 and
4 are pilose. P. javus females have 10
rows of deuterosternal teeth compared to
11 in females of P. numismaticus.

Adults of P. javus are found in the
nests of Bombus; deutonymphs occur both
in the nests and on adult bumblebees.
These mites were the most common Para-
situs in the studied area.

TYPES: Holotype of the deutonymph
collected 9 May 1971, at Prairie Bluff
Mountain, Alberta, Canada (49^^ 20' N,
114" 06' W), 5200 feet, from a Bombus
bijarius queen: collector L. A. Richards.
Located in CNC, no. 13188. Morphotypes
of male and female collected 30 May
1970, and 15 June 1970, respectively, from
a B. jrigidus nest at the same locality as
holotype, collector L. A. Richards; speci-
mens located in CNC. Paratypes, all col-
lected from same locality as the holotype.
Twenty-three deutonymphs: Three col-
lected on 18 August 1970, from B. bijarius
nest; three 5 June 1970, from P. insularis
queen; two 18 August 1970, from B.
bijarius nest; two 19 May 1971, from B.
occidentalis queen; one 3 August 1970,
from B. jrigidus nest; one 29 May 1971,
from B. rujocintus queen; one 28 May

1970, from B. jrigidus queen; one 28 May

1970, from B. occidentalis queen; one 4
June 1970, from B. melanopygus c]ueen;
one 3 August 1970, from B. occidentalis
nest; one 6 August 1970, from B. flavijrons
nest; one 28 July 1971, from B. rujocinctus
t]ueen; one 16 June 1971, from B. nevad-
ensis queen; one 3 June 1971, from B.
rujocinctus c^ueen; one 16 May 1971, from
P. suckleyi female; one 11 August 1970,
from B. flavijrons nest; and one 16 May

1971, from B. jrigidus queen. Twelve
males: Four collected 20 August 1971,
from B. occidentalis nest; two 14 August
1971, from B. bijarius nest; two 9 August
1970, from B. occidentalis nest; one 2
August 1970, from B. jrigidus nest; one
6 August 1970, from B. calijornicus nest;
one 14 August 1971, from B. vagans nest;
and one 28 July 1970, from B. mixtus nest.
Fourteen females: Three collected 11 Au-
gust 1971, from B. occidentalis nest; two
14 August 1971, from B. bijarius nest; one
9 August 1970, from B. occidentalis nest;
one 30 June 1971, from B. huntii nest; one
30 June 1970, from B. bijarius nest; one

8 August 1970, from B. bijarius nest; one
14 August 1971, from B. vagans nest; one

9 August 1971, from B. jrigidus nest; one
28 July 1971, from B. bijarius nest; one
30 July 1971, from B. huntii nest; and one
20 August 1971, from B. occidentalis nest.
Six deutonymphs, four females and three
males; located in CNC. Six deutonymphs,
three females and three males; in the
BMNH. Five deutonymphs, three females
and three males; in the USNM. Six deu-
tonymphs, four females, and three males;
at the University of Kansas.

The specific name, javus (i.e., honey-
combed) refers to the honeycomb pattern
on the sclerotized areas.

Parasitus inquilinobombus , new species

DEUTONYMPH (Fig. 15A, B): Idio-
somal length 0.62 mm, width 0.40 mm.
All propodosomal setae smooth; m 5

Parasitid Mites Associated with Rumblebees

737

longer than other setae; 1 4, 1 5, d 3, and
i 2 also long; m 1, 2, 3 and 4 short.
Hysterosomal shield semicircular, poly-
trichous, setae smooth. Dorsal sculptur-

ing, netlike.

Pre-endopodal shields 0.02 mm long,
teardrop-shaped. Sternal shield lacks dis-
tinct front border; has heavily sclerotized
area of netlike sculpturing which begins
at first pair of pores. All sternal setae
smooth. Metapodal shields sac-shaped,
have two anterior projections: one short
and round, the other long, tail-like, curv-
ing back, about '/3 the length of the shield
from its anterior extremity. Shield 0.03
mm long, 0.15 mm wide.

Tectum three-pointed, middle point
twice as long as wide. Chelicera: Digitus
mobihis has three teeth in apical half, all
about same size; digitus fixtus has four
teeth, three in apical half, fourth in distal
half, all about same size.

Length of leg I, 0.63 mm; II, 0.58 mm;
III, 0.52 mm; IV, 0.74 mm (Fig. 16A).
Most leg-setae slightly pilose, some smooth.
Genu II pd 2 longer than other setae.
Leg II (Fig. 16B) with most tarsal setae
thick, av 1 thornlike. No division of
tarsus I.

FEMALE (Fig. 17): Idiosomal length
0.94 mm, width 0.6(S mm. Propodosomal
shield has 24 pairs of setae. Setae smooth
and needlelike; m 5 longer than other
setae on propodosoma; m 1, 2, and 3
shorter than 4. Other setae of equal
length. Propodosomal shield wider than
long, larger than hysterosomal shield.
Dorsal sculpturing netlike; shield lines
forming pattern near the edges follow
the edges and become closer. Hystero-
somal shield wider than long, rounded
posteriorly.

Sternal shield has netlike, polygonal
sculpturing, beginning at first pair of
pores. First two pairs of sternal setae
pilose, all other ventral setae smooth.
Ventro-anal plate v-shaped, covered by a

netlikc pattern, fused with paratremal
shield.

Gnathosoma (Fig. 18D) with 11 rows
of deuterosternal teeth. Tectum (Fig.
18C) with three pf)ints, middle point
largest. G 1, 2, and 3 smooth, G 4 pilose.
Chelicera (Fig. 19C, D): Digitus mobilus
with three teeth, two evenly spaced in the
upper half and about half the size of the
third tooth in the lower half; digitus
fixtus, with four teeth, successively smaller
from the most proximal to the most distal
tooth.

Length of leg I, 0.77 mm; II, 0.62 mm;
III, 0.57 mm; IV, 0.89 mm. Tarsus II
(Fig. 19A) av 1, pv 1 heavy thornlike
setae, other setae also thick, most smooth.
Tarsus IV (Fig. 19B), setae longer than
other leg IV setae, pd 2 very long; av 1
and pv 1 of tibia and genu IV longer and
thicker than other setae of tibia and
irenu IV.

MALE (Fig. 20): Idiosomal length
0.83 mm, width 0.59 mm. Propodosomal
shield has 24 pairs of setae, all smooth;
m 5 and i 2 longest; m 1, 2, and 3 shorter
than 4. Line dividing propodosomal
shield and hysterosomal shield not ex-
tended to bodv margins. Hvsterosoma
polytrichous, setae all smooth and the
same length. Dorsal sculpturing like that
of female.

Tritosternum stout, base about ^4 the
total length. First pair of sternal setae
anterior to distinct part of sternum. Net-
like sculpturing of sternum extending to
coxa IV. All ventral setae smooth.

Gnathosoma (Fig. 18B) has 11 rows
of deuterosternal teeth. G 1, 2, and 3
smooth, G 4 pilose. Tectum (Fig. 18A)
has a single point, with row of teeth on
middle third of the edge of point. Che-
licera (Fig. 21A): Digitus mobilus curved
and pointed apically, with one tooth;
digitus fixtus slender, also curved apically
and with one tooth; cheliceral brush on
long stalk with few bristles,

738

The University of Kansas Science Bulletin

Length of leg I, 0.98 mm; II, 0.62 mm;
III, 0.55 mm; IV, 0.83 mm. Femur II has
well developed apophysis, rounded api-
cally, axillary process pointed apically.
Genu and tibia II have smaller apophyses.
Genual apophysis is round, tibial apophy-
sis longer and rounded apically (Fig. 2ID).
Coxal setae smooth or scarcely pilose. Leg
setae smooth and needlelike. Tarsus II,
III, IV, av 1, and pv 1 thornlike, mv, pi
3, al 3, pv 2, and al 2 thicker than other
tarsal setae. Most setae of tarsus IV (Fig.
21 B) thicker and longer than other leg
IV setae (Fig. 21C). Genu and tibia IV
with av 1, pv 1, thicker than other genal
and tibial setae.

This mite is distinguished by its small
size when compared to other Parasitns
collected from bumblebees at Prairie Bluff
Mountain. The deutonymphs are about
the same size as those of P. crinitiis Oude-
mans, 1903, and P. crinitosimilis Vitz-
thum, 1930. P. inquilinobombiis dififers
from P. crinitus by the lack of short setae
between the long ones on the hysterosomal
shield (Vitzthum 1930), and from P. cri-
nitosimilis by having shorter setae posteri-
orly. The deutonymph of P. inquilino-
bombiis differs from that of other Parasitns
species by the lack of the division of
tarsus I.

Adults of this mite are found in Bom-
bits nests; deutonymphs occur both in
nests and on adult bumblebees.

TYPES: Holotype of the deutonymph
collected 18 August 1970, at Prairie Bluff
Mountain, Alberta, Canada (49° 20' N,
114' 06' W), at 5200 feet, from B. bijarius
nest; collector L. A. Richards. Specimen
located in CNC, no. 13189. Morphotypes:
Male collected 11 July 1970, from B.
frigidiis nest; female collected 12 August
1970, from B. occidentalis nest; both from
same locality as holotype; collector L. A.
Richards, both specimens in CNC. Para-
types, all from same locality as holotype.
Sixteen deutonymphs: Thirteen collected

11 August 1971, from B. occidentalis nest;
one 18 August 1971, from B. bijarius nest;
one 9 August 1971, from B. bijarius nest;
and one 14 August 1971, from B. bijarius
nest. Fourteen males: Five 11 August
1971, from B. occidentalis nest; two 11
August 1971, from Bombus nest; one 11
July 1970, B. jrigidus nest; one 9 August
1971, from B. bijarius nest; one 22 July

1970, from B. occidentalis nest; one 3 Au-
gust 1970, from B. jrigidus nest; one 14
August 1970, from B. calijornicus nest;
one 31 July 1970, from B. jnixtus nest;
and one 9 August 1970, from B. jrigidus
nest. Fifteen females: Five 11 August

1971, from Bombus nest; four 8 August
1970, from B. bijarius nest; three 11 Au-
gust 1971, from B. occidentalis nest; one
2 August 1970, from B. appositus nest;
(me 9 August 1971, from B. bijarius nest;
and one 3 August 1970, from B. jrigidus
nest. Four deutonymphs, four females
and four males, in the CNC. Four deuto-
nymphs, three females and three males,
in the USNM. Four deutonymphs, four
females, three males, in BMNH. Four
deutonymphs, four females, and four
males at the University of Kansas.

The speces is named for its habitat,
the Bombus nest.

Parasitns perthecatus, new species

DEUTONYMPH (Fig. 22): Idiosomal
length 1.04 mm, width 0.74 mm. Pro-
podosomal shield has 22 pairs of setae,
all pilose and needlelike; m 5 longer, m
1, 2, 3, and 4 shorter than other propodo-
somal setae. Hysterosomal shield, trian-
gular. Setae off hysterosomal shield
shorter than those on shield, pilose. Dorsal
sculpturing netlike, similar to that of fe-
male. Pre-endopodal shields, triangular,
0.03 mm long and 0.05 mm wide. Sternal
shield has distinct front border. All ven-
tral setae pilose. Sternal shield has netlike
polygonal sculpturing. Metapodal shields.

Parasitid Mites Associated with Bumblebees

739

oval to pear-shaped, 0.33 mm long, 0.02
mm wide.

Tectum (Fig. 24 A) has three points,
middle one largest and rounded apically.
Chelicera (Fig. 24F): Both digits slender;
Digitus mobiliis has three teeth in upper
half; digitus fixtus has four teeth, apical
one smaller.

Length of leg I, 1.11 mm; II, 0.80 mm
(Fig. 23A): III 0.86 mm; IV, 1.26 mm
(Fig. 23B, C). Setae pilose. Tarsus IV
with ad 3, pd 2, very long.

FEMALE (Fig. 25): Idiosomal length
1.48 mm, width 1.04 mm. Twenty-two to
twenty-three pairs of setae on propodo-
somal shield; m 5 longer than other setae,
all dorsal setae pilose, 1 2, d 4, and i 2,
spatulate. Propodosomal shield smaller
than hysterosomal shield. Hysterosomal
shield covers entire hysterosoma. Hystero-
soma polytrichous.

First pair of sternal setae arise off
heavily sclerotized portion of sternal
shield. First two pairs of sternal setae
and genital setae, pilose. Anterior half
of ventro-anal shield has smooth setae,
other ventral setae pilose. Ventro-anal
shield V-shaped. Setae on posterior half
of ventral surface broader towards tip,
similar to 1 2, d 4, and i 2 of propodosomal
shield. Surface of sternal shield has net-
like sculpturing which continues through
epigynium, with smaller polygons anteri-
orly. Ventro-anal shield has netlike pat-
tern.

Gnathosoma (Fig. 26A) has 11 rows
of deuterosternal teeth. Tectum (Fig.
24B) has three points, the middle one
larger and rounded apically. Gl and 2
smooth, G 3 and 4 pilose. Chelicera (Fig.
24D) : Both digits slender, each has four
evenly-spaced large teeth.

Length of leg I, 1.23 mm; II, 0.98 mm
(Fig. 27A); III, 1.04 mm; IV, 1.54 mm
(Fig. 27B, C). Setae thick, pilose. Tarsus
II, III, IV with av 1 and pv 1 thornlike.
Coxal setae pilose.

MALE (Fig. 28): Idiosomal length
1.29 mm, width 0.98 mm. Propodosomal
shield has 22 setae; m 5 longer than other
setae; m 1, 2, and 3 shorter than m 4;
d 4 and i 2 knife-shaped. Propodosomal
shield and hysterosomal shield are di-
vided by a line extending to body margins.
All dorsal setae pilose. Hysterosoma poly-
trichous. Dorsal sculpturing like that of
female.

Tritosternum short, about same length
as coxa I, lacinia three times the length
of base. Sternal and genital setae conspic-
uously pilose, other setae slightly pilose.
Surface of sternal area has netlike pattern
which becomes less distinct towards anal
area.

Gnathosoma (Fig. 26B) has six rows of
deuterosternal teeth; tectum (Fig. 24C)
triangular, rounded apically. G 1, 2, and
3 smooth, G 4 pilose. G 2 shorter than
G 1, 3, and 4. Chelicera (Fig. 24E):
Digitus mobilus strongly curved apically,
with one tooth just above the middle;
digitus fixtus triangular, with dorsal point;
cheliceral brush a thick stalk 2/5 length
of digitus mobilus, with few bristles.

Length of leg I, 1.18 mm; II, 0.92 mm
(Fig. 29C); III, 1.01 mm; IV, 1.48 mm
(Fig. 29A, B). Femur II has well-devel-
oped apophysis; both axillary process and
apophysis rounded apically. Genu and
tibia II have smaller, rounded apophyses.
Leg-setae like those of female.

PROTONYMPH (Fig. 30): Idiosomal
length 1.18 mm, width 0.62 mm. Dorsal
shields not distinguishable. Propodosoma
and hysterosoma each with 14 pairs of
smooth setae. Tritosternum, base only
slightly shorter than lacinia. Three pairs
of sternal setae. No sternal shield visible.

Anal shield shaped like an upside-
down teardrop, with three setae, the ad-
anal pair and postanal or anal seta. Five
pairs of setae between sternal region and
anal shield, plus one seta on each side of
anal shield. Paratreme not visible.

740

The University of Kansas Science Bulletin

Length of leg I, 1.23 mm; II, 1.04 mm;
III, 1.17 mm; IV, 1.63 mm. Setae of
tarsus IV two to three times as long as
setae on other tarsi.

LARVA (Fig. 31): Idiosomal length
0.83 mm, width 0.55 mm. Nine pairs of
propodosomal setae, seven pairs of setae
on hysterosoma; m 5 longer than other
setae. Setae simple, smooth.

Ventral surface has three pairs of ster-
nal setae and four pairs of setae between
coxa III and anal shield. Adanal pair of
setae long, at least lYi times the length
of other ventral setae. Anal setae barely
distinguishable. Postanal setae very long,
four times length of adanals. Base and
lacinia of tritosternum about same length;
two smooth lateral branches accompany
lacinia. Gnathosoma has two pairs of
setae, 11 rows of deuterosternal teeth.

Length of leg I, 1.04 mm; II, 0.77 mm;
III, 0.86 mm.

This species is similar to P. jucoriun.
The shapes of 1 2, d 4, and i 2 of the female
are like those of the P. jiicontm female.
The chelicera is similar to that of the P.
jucorutn female. In P. jiicorum the fe-
male has a large hysterosomal shield not
completely covering the hysterosoma, as
it does in P. perthecatiis. On P. perthecatits
the sternal shield differs, the anterior por-
tion being less sclerotized than that of
P. fn coram.

The ventral setae of the male P. perthe-
catiis are pilose, whereas on P. jiicorum
males most ventral setae are smooth. The
cheliceral brush of males of P. fiiconin}
is not a stalk.

Ventral surfaces of deutonymphs of
P. jiicorum and P. perthecatiis are similar,
except for the absence of the vertical lines
in the sculpturing on the sternal shield of
P. perthecatiis.

P. perthecatiis are the largest Parasittis
found with the bumblebees in the study
area. Adults, protonymphs and larvae
were collected from nests, and deuto-

nymphs from nests and on adult bees.
These mites have also been collected at
Lake Hazen, EUesmere Island, Northwest
Territories in association with Bombiis
polaris Curtis and B. hyperboreits Schon-
herr by K. W. Richards in the summers
of 1967 and 1968.

TYPES: Holotype of the deutonymph
collected 12 June 1971, at Prairie Bluff
Mountain, Alberta, Canada (49° 20' N,
114° 06' W), at 5200 feet, from B. ap-
positiis nest; collector L. A. Richards.
Specimen located in CNC no. 13190.
Morphotypes: Male collected 29 June

1970, from B. bijariiis nest; female col-
lected 30 May 1970, from B. jrigidiis nest;
both from same locality and by the same
collector as the holotype. Larva reared 15
July 1971, from mating of male and fe-
male from Bom bus nest. Protonymph
reared 25 July 1971, from same source as
larva, both parents from same locality as
holotype. All these specimens located in
the CNC. Paratypes all from same lo-
cality as holotype. Nineteen deutonymphs:
Two collected 18 August 1970, from B.
bijariiis nest; two 6 July 1970, from B.
occidentalis queen; one 14 August 1971,
from B. bijariiis nest; one 24 May 1970,
from Bombiis sp. nest; one 29 May 1970,
from B. jrigidiis nest; one 14 June 1970,
from B. bijariiis queen; one 19 June 1971,
from B. calijorniciis queen; one 22 May

1971, from B. bijariiis queen; one 20 June
1971, from B. jrigidiis nest; one 6 June

1970, from B. flavijrons queen; one 3 June

1971, from B. appositiis queen; one 3 June
1971, from B. rujocinctus queen; one 19
May 1971, from B. occidentalis queen;
one 1 August 1970, from B. occidentalis
nest; one 16 June 1971, from B. nevaden-
sis queen; one 6 June 1970, from Hydro-
phylliim capitatum flower; and one 25
May 1971, from B. flavijrons queen. Seven
males: Two collected 9 August 1970,
from B. occidentalis nest; one 5 June 1970,
from B. bijariiis nest; one 8 August 1971,

Parasitid Mitls Associated with Bumblebees

741

from B. bijariiis nest; one 14 August 1970,
from B. calijornicus nest; one 9 August
1971, from B. bijariiis nest; and one 13
August 1970, from B. appositiis nest. Seven
females: Two 6 June 1970, from B. occi-
denlalis nest; two 2 August 1970, from
B. iippositits nest; one 9 July 1971, from
Bomhiis sp. nest; one 2 August 1970, from
B. appositiis nest; and one 14 August 1970,
from B. calijornicus nest. Five deuto-
nymphs; two females, and two males;
located in the CNC. Five deutonymphs;
two females, and two males; located in
the BMNFi. Four deutonymphs; one fe-
male, and one male; located in the USNM.
Five deutonymphs; two females and two
males; located at the University of Kansas.
The specific name is derived from
Latin, meaning well sheathed, referring to
the heavy dorsal and ventral shields cover-
ing most of the body of the adult mites.

Parasitiis fijnetoritm Berlese, 1903

Gannisits fimetoriim Berlese, 1903, Redia,

1 (1904) :238.
Parasitiis affinis Oudemans, 1904, Ent.

Ber. nederl. Ver., Gravenhage, 1:120.
Parasitiis fimetoriim Schweitzer, 1949, Res.

Rech. Scient. Pare Nation. Suisse (N.

F.) 2:8-9.
Parasitiis fimetoriim Costa, 1961, Ann.

Mag. Nat. Hist. (13)4:261.
Eiigamasiis fimetoriim Holzmann, 1969,

Acarologie 13:14.
Parasitiis fimetoriim Micherdzinski, 1969,

Panstwowe Wydawnictwo Naukowe.

p. 478.

The specimens from southern Alberta
differ slightly from those previously de-
scribed. Differences include the shape of
the tectum of the deutonymph, which is
triangular for the Alberta specimens;
other descriptions call it three-pointed.
On the gnathosoma of the male, accord-
ing to Costa (1961), eight rows of deutero-
sternal teeth are present; Micherdzinski
(1969) shows only a single basal row,

which is the same condition as in the
specimens from Alberta; Holzmann
(1969) shows two to three rows. Male,
female and deutonymphs of this species
have been described.

These mites are common in Europe
and have also been collected in Israel.
They are found in manure, but also oc-
cur under wood, stones, bark, and in soil
habitats. Generally the deutonymphal
stage is the most commonly encountered.
They have not previously been reported
from Bombits nests. Two specimens re-
ceived from Dr. R. Husband w^ere col-
lected from nests of B. jeruidiis (Fabric-
ius) and B. americanorum (Fabricius) in
Michigan. In Alberta this species was
found in Bombiis nests, but not on the
bees. Thus it is likely to be a soil in-
habitant with no specific relationship to
Bom bus such as seems to exist for the
preceding species.

LITERATURE CITED

Berlese, A. 1903. Acari nouvi. Manipulus I.
Redia 1:235-252.

. 1905. Monografia del gcnere Gamastis

Latr. Redia 3( 1906) :66-304. "

Costa, M. 1961. Mites recovered from the nests
of the Levant Vole (Microtits gitentheri) in
Israel. Ann. Mag. nat. Hist., London, ser.
13, 4(41):257-282.

HiRscHMANN, \V. 1957. Acorologie Gangsyste-
matik der Parasitiformes, Teil I, Rumpfbeha-
arung und Riickenflachen. Furth/Bayern.
20 p.

HoLZMAXN, C. 1969. Die Familie Parasitidae
Oudemans, 1901. .\carologie 13:3-54.

Karg, \V. 1965. Larvalsystematische und phylo-
genetische untcrsuchungen sowie revision des
systems der Gamasina Leach, 1815. Mitt,
zool. Mus. Berl. 41:193-340.

MicHERDZixsKi, W. 1969. Die Familie Parasitidae
Oudemans, 1901 (Acarina, Mesostigmata).
Panstowowc Wydawnictwo Naukowe, Kra-
kow. 690 p.

Oudemans, A. C. 1902. New list of Dutch Acari.
Second part, with remarks on known and
descriptions of a new subfamily, genera and
species. Tijdschr. Ent. 65:1-52.

■ — •. 1903. Acarologische anteekeningen M.

Ent. Ber., Amst. 50:83-88.

. 1904. .Acarologische anteekeningen IX.

Ent. Ber., Amst. 1:119-120.

742

The University of Kansas Science Bulletin

. 1905. Acariden von Borkum unci

Wangeroog (Parasitidae, Thrombiidae, Ori-
batidae), IX Series der Notes on Acari. Abh.
naturw. Ver. Bremen 18:77-98.

. 1906. Notes on Acari. XV series.

Tijdschr. Ent. 49:237-270.

Richards, L. A. and K. W. Richards. Parasitid
mites associated with bumblebees in Alberta
(Acarina: Parasitidae; Hymenoptera: Apidae)
II Biology. Kansas Sci. Bull. 50(14) (in
press).

ScHWEizER, J. 1949. Die Landmilben des Schvvei-
zerischen National-parkes. 1. Teil: Parasiti-
formes Rcuter 1909. Ergebn. wiss. Unters.
schweiz. Nat. Parks. 2:99.

SelltwCK, M. 1938. Milben ais Parasiten von

Insekten. Verb. 11 Intcrnat. Kongr. Entomol-
ogie, Berlin 2:1300-1307.
Tragardh, I. 1904. Monographie der arktischen
Acariilen. Fauna arct. 4:1-78.

. 1910. .Acariden aud dem Sarekgebirge.

Naturw. Unters. Sarkgebirg. 4:375-568.
. 1912. Danmark Ekspeditionen til Gr0n-

lands Nordostkyst 1906-1908. Saertryak af

"Meddelelser on Gr0nland 43 Kobenhaven

Bianico Lunos Bogtrykkeri 3(14):412-426.
ViTZTHUM, H. 1930. Die Bomhiis Parasiten. Zool

Jahrb. Syst., Okol., Geog. 60:1-45.
WiLLMANN, C. 1939. Terrestriche Acari der Nord

und Ostsee-Kiiste. Abh. naturw. Ver. Brem-

men 31:521-550.

Parasitid Mites Associated with Bumblebees

743

Fig. 1. Parasitiis hobbsi, deuton>mph. A. dorsal chaetotaxy, B. ventral side,

744

The University of Kansas Science Bulletin

Fig. 2. Parasitiis liobhsi A. male chelicera, B. deutonymph cheliccra, C. dcutonymph tectum, D. deuto-

nymph leg II, E. leg IV,

Parasitid Mites Associated with Bumblebees

745

200JJ

Fig. 3. Parasitus hobbsi, female. A. dorsal chaetotaxy, B. ventral side.

746

The University of Kansas Science Bulletin

lOOjL)

100 ju

yii

1 II J 1 1 1 1 1 1 1 .

< /

"I 1 1 1 J 1 1 1

'••'

••III 1 III! ,,/

Ht

"""Uii,

iiiii

'""""'/.,^

iiJii

"'""•*.«

„UHillU/.,^

B

Fk;. 4. Ptirasittis liohhsi, female. A. tectum, R. gnathosoma, C. chelicera.

PaKASITID MiTtS ASSOCIATI-.I) WITH BuMlJLEBliES

747

^^'^^

A

^^^T^

lOOju

Fig. 5. Parasitns liobhsi, female. A. leg 11, B. tibi.i IV and tarsus I\', C. coxa, genu, trochanter, femur.

v^^

Fin. 6. Parasi/ns liohbsi, male. A. ventral side, B. dorsal chactotaxv.

Parasitid Mites Associated with Bumblebees

749

ii.iii.^

Hum tiUt iiiiiiii'j'

s^l'

- ■ '<....•''

Fig. 7. Parasittis hobbsi, male. A. tarsus IV, B. coxa, trochanter, femur, genu, tibia IV, C. leg II,

D. tectum, E. gnathosoma.

750

The University of Kansas Science Bulletin

Fig. 8. Parasitits funis, dcutonv mph. A. ventral side, B. dorsal chactotaxy.

Parasitid Miti:s Associated with Bumblebees

751

lOQu

A

IOju

B

Fig. 9. Parasitiis jai'its, tlcutonymph. A. chcliccra, B. tectum, C. leg II, D. leg IV,

752

The University of Kansas Science Bulletin

ZOOoj

f\

B

Fig. 10. Parasitiis jai'tis, female. A. dorsal chaetotaxy, B, ventral side.

Parasitid MiTiis Associated with Bumblebees

753

Fig. 11. Parasittis favtis A. female gnathosoma, B. female tectum, C. male gnathosoma, D. male tectum.

lOOoj

Fig. 12. Pcvasi/m jams, female. A. leg IV, B. leg II, C. chtlicera.

Parasitid Mites Associati-d with Bumblebees

755

Fig. 13. Parasitiis faviis, male. A. dorsal chaetotaxy. B. ventral side.

756

The University of Kansas Science Bulletin

Fig. 14. Parasittis javus, male. A. leg IV, B. leg II, C. chelicera.

200^

Fig. 15. Parasitiis inquilinobombus, deutonymph. A. ventral side, B. dorsal chaetotaxy.

758

The University oe Kansas Science Bulletin

Fig. 16. Panis/fiis iiiqiiilinohonibits, tlcutonymph. A. leg IV, B. leg II.

Parasitid Mitks Associated with Bumblebees

759

Fig. 17. laiantiis inquilniohomhns, female. A. ventral side, R. dorsal chaetotaxy.

760

The University of Kansas Science Bulletin

lOOju

""*'I"H,JJ

Jll " I, ,.

^ • A *

•''U;m.MI^

B

Fig. 18. Parasittis inquilinobomhiis A. male tectum, B. male gnathosoma. C. female tectum, D. female

gnathosoma.

Parasitid Mites Associated with Bumblebees

761

c

lOOju

D

Fig. 19. Parasittis inqtiilinobomhtis, female. A. leg II, B. leg IV, C. chelicera left, D. chelicera right.

762

The University of Kansas Science Bulletin

Fig. 20. Parasitus inqinlinohombin, male. A. dorsal chactotaxy, B. ventral side.

Parasitid Mites Associated with Bumblebees

763

Fig. 21. Parasitus inqtiilinohombiis, male. A. chelicera, B. tarsus IV, C. coxa, trochanter, femur, genu,

tibia IV, D. kg II.

764

The University of Kansas Science Beilletin

Fig. 22. rarusilits pcithccatin. dcutiinympli. A. dorsal chactotaxy, B. ventral side.

Parasitid Mites Associated with Bumblebees

765

Fig. 23. Parasitiis perthecaltis, deuton\mph. \. leg II, B. tarsus IV, C. coxa, trochanter, femur, genu,

tibia IV.

766

The University of Kansas Science Bulletin

Fig. 24. Paiusittis perthecatits A. dcutonymph tectum, B. female tectum, C.

chclicera, E. male chelicera, F. dcutonymph chelicera.

male tectum, D. female

Parasitid Mites Associated with Bumblebees

767

Fk;. 25. Pcirasitiis pothccatits, female. .\. dorsal chaetotaxy, B. ventral side.

768

The University of Kansas Science Bulletin

-'«., ,A...<

A '""'" B

Fig. 26. Panisitiis perthccatiis \. female gnatho.soma, B. male gnathosoma.

Parasitid Miti:s Associated with Bumbllbees

769

100

'JU

Fig. 27. Paiasitits pcrtheaitiis, female. A. leg II, B. coxa, trochanter, femur, tibia IV, C. tarsus IV.

770

The University of Kansas Science Bulletin

Fig. 28. Parasitiis pcrthecatm . male. A. dorsal chaetotaxy, B. ventral side.

Parasitid Mitus Associatld with Bumblubees

771

Fk;. 29. Parasitiis pcrthecaliis, male. A. tarsus IV, B. coxa, trochanter, femur, tibia IV, C. leg II.

772

The University of Kansas Science Bulletin

Fig. 30. Parasittis perthecatits, protonymph. A. dorsal chactotaxy, B. ventral side.

Parasitid Mites Associated with Bumblebees

773

f vv>

Fi<;. 31. Parasiltis patheaittis,h\r\^. A. dorsal chactotaxy, B. ventral side.

%

iW^

iri'i-j''. MfJi

'.wM^K.''-'