of theWorld
Charles D. Michener
University of Kansas Natural History Museum
and Department of Entomology
© 2000 The Johns Hopkins University Press All rights reserved. Published 2000
Printed in the United States of America on acid-free paper 9 8 7 6 5 4 3 2 1
The Johns Hopkins University Press 2715 North Charles Street Baltimore, Maryland 21218-4363 www.press.jhu.edu
Library of Congress Cataloging-in Publication Data Michener, Charles Duncan, 1918–
The bees of the world / Charles D. Michener.
p. cm.
Includes bibliographical references. ISBN 0-8018-6133-0 (alk. paper) 1. Bees Classification. I. Title QL566.M53 2000
595.79⬘9—dc21 99-30198 CIP
A catalog record for this book is available from the British Library.
To my many students, now scattered over the world,
from whom I have learned much
Preface
ixNew Names
xivAbbreviations
xiv1. About Bees and This Book
12. What Are Bees?
23. The Importance of Bees
34. Development and Reproduction
45. Solitary versus Social Life
96. Floral Relationships of Bees
137. Nests and Food Storage
198. Parasitic and Robber Bees
269. Body Form, Tagmata, and Sex
Differences
3810. Structures and Anatomical Terminology
of Adults
4011. Structures and Terminology
of Larvae
5312. Bees and Sphecoid Wasps as a Clade
5413. Bees as a Holophyletic Group
5514. The Origin of Bees from Wasps
5815. Classification of the Bee-Sphecoid
Clade
6016. Bee Taxa and Categories
6117. Methods of Classification
7118. The History of Bee Classifications
7219. Short-Tongued versus Long-Tongued
Bees
7820. Phylogeny and the Proto-Bee
8321. The Higher Classification of Bees
8822. Fossil Bees
9323. The Antiquity of Bee Taxa
9424. Diversity and Abundance
9625. Dispersal
9926. Biogeography
10027. Reduction or Loss of Structures
10428. New and Modified Structures
10629. Family-Group Names
11130. Explanation of Taxonomic Accounts
in Sections 36 to 119
11231. Some Problematic Taxa
11432. The Identification of Bees
11533. Key to the Families, Based on Adults
11634. Notes on Certain Couplets in the Key
to Families (Section 33)
12035. Practical Key to Family-Group Taxa,
Based on Females
12136. Family Stenotritidae
12337. Family Colletidae
12638. Subfamily Colletinae
13039. Subfamily Diphaglossinae
16440. Tribe Caupolicanini
16541. Tribe Diphaglossini
16842. Tribe Dissoglottini
17043. Subfamily Xeromelissinae
17144. Tribe Chilicolini
17245. Tribe Xeromelissini
17746. Subfamily Hylaeinae
17847. Subfamily Euryglossinae
21048. Family Andrenidae
22549. Subfamily Alocandreninae
22850. Subfamily Andreninae
22951. Subfamily Panurginae
26052. Tribe Protandrenini
262Contents
viii
53. Tribe Panurgini
27354. Tribe Melitturgini
27855. Tribe Protomeliturgini
28156. Tribe Perditini
28257. Tribe Calliopsini
29258. Subfamily Oxaeinae
30159. Family Halictidae
30460. Subfamily Rophitinae
30761. Subfamily Nomiinae
31762. Subfamily Nomioidinae
33063. Subfamily Halictinae
33364. Tribe Halictini
33965. Tribe Augochlorini
37766. Family Melittidae
39667. Subfamily Dasypodainae
39968. Tribe Dasypodaini
40069. Tribe Promelittini
40570. Tribe Sambini
40671. Subfamily Meganomiinae
40972. Subfamily Melittinae
41273. Family Megachilidae
41774. Subfamily Fideliinae
41975. Tribe Pararhophitini
42076. Tribe Fideliini
42177. Subfamily Megachilinae
42478. Tribe Lithurgini
42779. Tribe Osmiini
43180. Tribe Anthidiini
47481. Tribe Dioxyini
52182. Tribe Megachilini
52683. Family Apidae
57084. Subfamily Xylocopinae
57585. Tribe Manueliini
57786. Tribe Xylocopini
57887. Tribe Ceratinini
59388. Tribe Allodapini
60089. Subfamily Nomadinae
61490. Tribe Hexepeolini
61891. Tribe Brachynomadini
62092. Tribe Nomadini
62493. Tribe Epeolini
62794. Tribe Ammobatoidini
63395. Tribe Biastini
63696. Tribe Townsendiellini
63997. Tribe Neolarrini
64098. Tribe Ammobatini
64199. Tribe Caenoprosopidini
646100. Subfamily Apinae
647101. Tribe Isepeolini
652102. Tribe Osirini
654103. Tribe Protepeolini
658104. Tribe Exomalopsini
660105. Tribe Ancylini
665106. Tribe Tapinotaspidini
667107. Tribe Tetrapediini
674108. Tribe Ctenoplectrini
676109. Tribe Emphorini
679110. Tribe Eucerini
686111. Tribe Anthophorini
720112. Tribe Centridini
731113. Tribe Rhathymini
739114. Tribe Ericrocidini
740115. Tribe Melectini
747116. Tribe Euglossini
754117. Tribe Bombini
761118. Tribe Meliponini
779119. Tribe Apini
806Literature Cited
809Addenda
871Index of Terms
873Index of Taxa
877In some ways this may seem the wrong time to write on the systematics
of the bees of the world, the core topic of this book. Morphological
in-formation on adults and larvae of various groups has not been fully
de-veloped or exploited, and molecular data have been sought for only a few
groups. The future will therefore see new phylogenetic hypotheses and
improvement of old ones; work in these areas continues, and it has been
tempting to defer completion of the book, in order that some of the new
information might be included. But no time is optimal for a systematic
treatment of a group as large as the bees; there is always significant
re-search under way. Some genera or tribes will be well studied, while others
lag behind, but when fresh results are in hand, the latter may well
over-take the former. I conclude, then, that in spite of dynamic current
activ-ity in the field, now is as good a time as any to go to press.
This book constitutes a summary of what I have been able to learn
about bee systematics, from the bees themselves and from the vast body
of literature, over the many years since I started to study bees, publishing
my first paper in 1935. Bee ecology and behavior, which I find fully as
fascinating as systematics, are touched upon in this book, but have been
treated in greater depth and detail in other works cited herein.
After periods when at least half of my research time was devoted to
other matters (the systematics of Lepidoptera, especially saturniid moths;
the biology of chigger mites; the nesting and especially social behavior of
bees), I have returned, for this book, to my old preoccupation with bee
systematics. There are those who say I am finally finishing my Ph.D.
thesis!
My productive activity in biology (as distinguished from merely
look-ing and belook-ing fascinated) began as a young kid, when I painted all the
native plants that I could find in flower in the large flora of Southern
California. When, after a few years, finding additional species became
difficult, I expanded my activities to drawings of insects. With help from
my mother, who was a trained zoologist, I was usually able to identify
them to family. How I ultimately settled on Hymenoptera and more
specifically on bees is not very clear to me, but I believe it had in part to
do with
Perdita rhois
Cockerell, a beautiful, minute, yellow-and-black
in-sect that appeared in small numbers on Shasta daisies in our yard each
summer. The male in particular is so unbeelike that I did not identify it
as a bee for several years; it was a puzzle and a frustration and through it I
Preface
x
became more proficient in running small Hymenoptera, including bees,
through the keys in Comstock’s
Introduction to Entomology
.
Southern California has a rich bee fauna, and as I collected more
species from the different flowers, of course I wanted to identify them to
the genus or species level. Somehow I learned that T. D. A.Cockerell at
the University of Colorado was the principal bee specialist active at the
time. Probably at about age 14 I wrote to him, asking about how to
iden-tify bees. He responded with interest, saying that Viereck’s
Hymenoptera
of Connecticut
(1916) (which I obtained for $2.00) was not very useful in
the West. Cresson’s
Synopsis
(1887) was ancient even in the 1930s, but
was available for $10.00. With these inadequate works I identified to
genus a cigar box full of bees, pinned and labeled, and sent them to
Cockerell for checking. He returned them, with identifications corrected
as needed, and some specimens even identified to species.
Moreover, Cockerell wrote supporting comments about work on bees
and invited me to meet him and P. H. Timberlake at Riverside,
Califor-nia, where the Cockerells would be visiting. Timberlake was interested in
my catches because, although I lived only 60 miles from Riverside, I had
collected several species of bees that he had never seen. Later, he invited
me to accompany him on collecting trips to the Mojave and Colorado
deserts and elsewhere.
Professor and Mrs. Cockerell later invited me to spend the next
sum-mer (before my last year in high school) in Boulder with them, where I
could work with him and learn about bees. Cockerell was an especially
charming man who, lacking a university degree, was in some ways a
second-class citizen among the university faculty members. He had never
had many students who became seriously interested in bees, in spite of
his long career (his publications on bees span the years from 1895 to
1949) as the principal bee taxonomist in North America if not the world.
Probably for this reason he was especially enthusiastic about my interest
and encouraged the preparation and publication of my first taxonomic
papers. Thus I was clearly hooked on bees well before beginning my
un-dergraduate work at the University of California at Berkeley.
As a prospective entomologist I was welcomed in Berkeley and given
space to work among graduate students. During my undergraduate and
graduate career, interacting with faculty and other students, I became a
comparative morphologist and systematist of bees, and prepared a
disser-tation (1942) on these topics, published with some additions in 1944.
The published version included a key to the North American bee genera,
the lack of which had sent me to Professor Cockerell for help a few years
before. Especially important to me during my student years at Berkeley
were E. Gorton Linsley and the late Robert L. Usinger.
xi
Until 1950, I had gained little knowledge of bee behavior and nesting
biology, having devoted myself to systematics, comparative morphology,
and floral relationships, the last mostly because the flowers help you find
the bees. In 1950, however, I began a study of leafcutter bee biology, and
a few years later I began a long series of studies of nesting biology and
so-cial organization of bees, with emphasis on primitively soso-cial forms and
on the origin and evolution of social behavior. With many talented
grad-uate students to assist, this went on until 1990, and involved the
publica-tion in 1974 of
The Social Behavior of the Bees.
Concurrently, of course,
my systematic studies continued; behavior contributes to systematics and
vice versa, and the two go very well together.
Across the years, I have had the good fortune to be able to study both
behavior and systematics of bees in many parts of the world. In addition
to shorter trips of weeks or months, I spent a year in Brazil, a year in
Aus-tralia, and a year in Africa. The specimens collected and ideas developed
on these trips have been invaluable building blocks for this book.
Without the help of many others, preparing this book in its present form
would have been impossible. A series of grants from the National Science
Foundation was essential. The University of Kansas accorded me
free-dom to build up a major collection of bees as part of the Snow
Entomo-logical Division of the Natural History Museum, and provided excellent
space and facilities for years after my official retirement. Students and
other faculty members of the Department of Entomology also
con-tributed in many ways. The editorial and bibliographic expertise of Jinny
Ashlock, and her manuscript preparation along with that of Joetta
Weaver, made the job possible. Without Jinny’s generous help, the book
manuscript would not have been completed. And her work as well as
Joetta’s continued into the long editorial process.
It is a pleasure to acknowledge, as well, the helpful arrangements made
by the Johns Hopkins University Press and particularly the energy and
enthusiasm of its science editor, Ginger Berman. For marvelously
de-tailed and careful editing, I thank William W. Carver of Mountain View,
California.
xii
D.C., USA (Halictini); Gabriel A. R. Melo, Ribeirão Preto, São Paulo,
Brazil (who read much of the manuscript); Robert L. Minckley, Auburn,
Alabama, USA (Xylocopini); Jesus S. Moure, Curitiba, Paraná, Brazil;
Christopher O’Toole, Oxford, England, UK; Laurence Packer, North
York, Ontario, Canada (Halictini); Alain Pauly, Gembloux, Belgium
(Malagasy bees, African Halictidae); Yuri A. Pesenko, Leningrad, Russia;
Stephen G. Reyes, Los Baños, Philippines (Allodapini); Arturo
Roig-Alsina, Buenos Aires, Argentina (phylogeny, Emphorini,
Tapinotaspi-dini, Nomadinae); David W. Roubik, Balboa, Panama (Meliponini);
Jerome G. Rozen, Jr., New York, N.Y., USA (Rophitini, nests and larvae
of bees, and ultimately the whole manuscript); Luisa Ruz, Valparaíso,
Chile (Panurginae); the late S. F. Sakagami, Sapporo, Japan (Halictinae,
Allodapini, Meliponini); Maximilian Schwarz, Ansfelden, Austria (
Coe-lioxys
); Roy R. Snelling, Los Angeles, California, USA (Hylaeinae);
Os-amu Tadauchi, Fukuoka, Japan (
Andrena
); Harold Toro, Valparaíso,
Chile (Chilicolini, Colletini); Danuncia Urban, Curitiba, Paraná, Brazil
(Anthidiini, Eucerini); Kenneth L. Walker, Melbourne, Victoria,
Aus-tralia (Halictini); V. B. Whitehead, Cape Town, South Africa (
Rediviva
);
Paul H.Williams, London, England, UK (
Bombus
); Wu Yan-ru, Beijing,
China; Douglas Yanega, Belo Horizonte, Minas Gerais, Brazil, and
Riverside, California.
The persons listed above contributed toward preparation or
comple-tion of the book manuscript, or the papers that preceded it, and also in
some cases gave or lent specimens for study; the following additional
persons or institutions lent types or other specimens at my request:
Josephine E. Cardale, Canberra, ACT, Australia; Mario Comba,
Cecchina, Italy (
Tetralonia
); George Else and Laraine Ficken, London,
England, UK; Yoshihiro Hirashima, Miyazaki City, Japan; Frank Koch,
Berlin, Germany; Yasuo Maeta, Matsue, Japan; the Mavromoustakis
Collection, Department of Agriculture, Nicosia, Cyprus (Megachilinae).
The illlustrations in this book are designed to show the diversity (or,
in certain cases, similarity or lack of diversity) among bees. It was entirely
impractical to illustrate each couplet in the keys—there are thousands of
them—and I made no effort to do so, although references to relevant text
illustrations are inserted frequently into the keys. Drs. R. J. McGinley
and B. N. Danforth, who made or supervised the making of the many
illustrations in Michener, McGinley, and Danforth (1994), have
permit-ted reuse here of many of those illustrations. The other line drawings are
partly original, but many of them are from works of others, reproduced
here with permission. I am greatly indebted to the many authors whose
works I have used as sources of illustrations; specific acknowledgments
accompany the legends. In particular I am indebted to J. M. F. de
Ca-margo for the use of two of his wonderful drawings of meliponine nests,
and to Elaine R. S. Hodges for several previously published habitus
drawings of bees. Modifications of some drawings, additional lettering as
needed, and a few original drawings, as acknowledged in the legends, are
the work of Sara L. Taliaferro; I much appreciate her careful work.
xiii
noting here that many other superb photographs by Westrich were
pub-lished in his two-volume work on the bees of Baden-Württemberg
(Westrich, 1989).
Svetlana Novikova and Dr. Bu Wenjun provided English translations
of certain materials from Russian and Chinese, respectively. Their help is
much appreciated.
The text has been prepared with the help of the bees themselves,
pub-lications about them, and unpublished help from the persons listed
above. I have not included here the names of all the persons responsible
for publications that I have used and from which I have, in many cases,
derived ideas, illustrations, bases for keys, and other items. They are
ac-knowledged in the text. Several persons, however, have contributed
pre-viously unpublished keys that appear under their authorship in this
book. Such contributions are listed below, with the authors’ affiliations.
“Key to the Palearctic Subgenera of
Hylaeus
” by H.H. Dathe, Deutsches
Entomologisches Institut, Postfach 10 02 38, D-16202 Eberswalde,
Germany.
“Key to the New World Subgenera of
Hylaeus
” by Roy R. Snelling, Los
Angeles County Museum of Natural History, 900 Exposition
Boule-vard, Los Angeles, California 90007, USA.
“Key to the Genera of Osmiini of the Eastern Hemisphere,” “Key to the
Subgenera of
Othinosmia,
” and “Key to the Subgenera of
Protosmia
”
by Terry L. Griswold, Bee Biology and Systematics Laboratory, UMC
53, Utah State University, Logan, Utah 84322-5310, USA.
“Key to the Genera of the Tapinotaspidini” by Arturo Roig-Alsina,
Museo Argentino de Ciencias Naturales, Av. A. Gallardo 470, 1405
Buenos Aires, Argentina.
I have modified the terminology employed in these keys, as necessary,
to correspond with that in use in other parts of this book (see Sec. 10).
Several contributions became so modified by me that the original
au-thors would scarcely recognize them. I have identified them by
expres-sions such as “modified from manuscript key by . . .”
Names of authors of species are not integral parts of the names of the
organisms. In behavioral or other nontaxonomic works I omit them
ex-cept when required by editors. But in this book, which is largely a
sys-tematic account, I have decided to include them throughout for the sake
of consistency.
xiv
Abstractors may note that five new names are proposed in this book, as
follows:
Acedanthidium,
new name (Sec, 80)
Andrena (Osychnyukandrena),
new name (Sec. 50)
Ceratina (Rhysoceratina),
new subgenus) (Sec. 87)
Fidelia (Fideliana),
new subgenus (Sec. 76)
Nomia (Paulynomia),
new subgenus (Sec. 61)
There are also numerous new synonyms at the generic or subgeneric
levels and, as a result, new combinations occur, as noted in the text.
The following are used in the text:
BP = before the present time
Code = International Code of Zoological Nomenclature
Commission = International Commission on Zoological
Nomemclature
L-T = long-tongued (see Sec. 19)
myBP = million years before the present
s. str. (sensu stricto)
= in the strict sense
s. l. (sensu lato)
= in the broad sense
S-T = short-tongued (see Sec. 19)
S1, S2, etc. = first, second, etc., metasomal sterna
scutellum = mesoscutellum
scutum = mesoscutum
stigma = pterostigma of forewing
T1, T2, etc. = first, second, etc., metasomal terga
Since ancient times, people have been drawn to the study of bees. Bees are spritely creatures that move about on pleasant bright days and visit pretty flowers. Anyone studying their behavior should find them attractive, partly because they work in warm sunny places, during pleasant seasons and times of day. The sights and odors of the fieldwork ambience contribute to the well-being of any researcher. Moreover, bees are important pollinators of both natural vegetation and crops, and certain kinds of bees make useful products, especially honey and wax. But quite apart from their practical importance, at least since the time of Aristotle people have been interested in bees because they are fascinating creatures. We are social ani-mals; some bees are also social. Their interactions and communications, which make their colonial life func-tion, have long been matters of interest; we wonder how a tiny brain can react appropriately to societal problems similar to those faced by other social animals, such as hu-mans. For a biologist or natural historian, bees are also fas-cinating because of their many adaptations to diverse flowers; their ability to find food and nesting materials and carry them over great distances back to a nest; their ability to remember where resources were found and re-turn to them; their architectural devices, which permit food storage, for example, in warm, moist soil full of bac-teria and fungi; and their ability to rob the nests of oth-ers, some species having become obligate robbers and others cuckoolike parasites. These are only a few of the in-teresting things that bees do.
I consider myself fortunate to work with such a bio-logically diverse group of insects, one of which is the com-mon honey bee, Apis melliferaLinnaeus. In terms of phys-iology and behavior, it is the best-known insect. Educated guesses about what happens in another bee species are of-ten possible because we know so much about Apis mellif-era.In this book, however, Apisis treated briefly, like all other bee taxa, its text supplemented by references to books on Apisbiology; the greater part of this book con-cerns bees (the great majority) that are not even social.
Sections 2 to 28, and what follows here, are intended to provide introductory materials important to an un-derstanding of all bees and aspects of their study. Some topics are outlined only briefly to provide background in-formation; others are omitted entirely; still others are dealt with at length and with new or little-known insights when appropriate.
This book is largely an account of bee classification and of phylogeny, so far as it has been pieced together, i.e., the systematics of all bees of the world. All families, subfam-ilies, tribes, genera, and subgenera are characterized by means of keys and (usually brief) text comments to facil-itate identification. I include many references to such re-visional papers or keys as exist, so that users can know where to go to identify species. About 16,000 species have been placed as to genus and subgenus (see Sec. 16); no at-tempt has been made even to list them here, although the approximate number of known species for each genus
and subgenus is given in Table 16-1, as well as under each genus or subgenus in Sections 36 to 119. Aspects of bee biology, especially social and parasitic behavior, nest ar-chitecture, and ecology, including floral associations, are indicated. Major papers on bee nesting biology and flo-ral relationships are also cited. The reader can thus use this book as a guide to the extensive literature on bee biology. Because the male genitalia and associated sterna of bees provide characters useful at all levels, from species to fam-ily, and because they are often complex and difficult to de-scribe, numerous illustrations are included, as well as ref-erences to publications in which others are illustrated.
Besides entomologists, this book should be useful to ecologists, pollination biologists, botanists, and other naturalists who wish to know about the diversity and habits of bees. Such users may not be greatly concerned with details of descriptive material and keys, but should be able to gain a sense of the taxonomic, morphological, and behavioral diversity of the bee faunas with which they work. As major pollinators, bees are especially important to pollination biologists. I hope that by providing infor-mation on the diversity of bees and their classification and identification, this book will in some mostly indirect ways contribute to pollination biology.
The title of this book can be read to indicate that the book should deal, to at least some degree, with all aspects of bee studies. It does not. All aspects of apiculture,the study and practice of honey bee culture, based on man-aged colonies of Apis melliferaLinnaeus and A. cerana Fabricius, are excluded. The findings about sensory phys-iology as well as behavioral interactions, including com-munication, foraging behavior, and caste control are vir-tually omitted, although they constitute some of the most fascinating aspects of biology and in the hands of Karl von Frisch led to a Nobel prize. A major work, principally about communication, is Frisch (1967).
Whether the scientific study of communication in Apis is part of apiculture is debatable, but the study of all the other species of bees is not; such studies are subsumed un-der the term melittology.Persons studying bees other thanApisand concerned about the negative and awkward expression “non-Apisbees” would do well to call them-selves melittologists and their field of study melittology. I would include under the term “melittology” the taxo-nomic, comparative, and life history studies of species of the genus Apis,especially in their natural habitats. This book is about melittology.
Users of this book may wonder about the lack of a glos-sary. Definitions and explanations of structures, given mostly in Section 10, are already brief and would be largely repeated in a glossary. The terms, including many that are explained only by illustrations, are therefore in-cluded in the Index of Terms, with references to pages where they are defined, illustrated, or explained. Some terminology, e.g., that relevant only to certain groups of bees, is explained in other sections, and indexed accord-ingly.
1. About Bees and This Book
A major group of the order Hymenoptera is the Section Aculeata, i.e., Hymenoptera whose females have stings— modifications of the ovipositors of ancestral groups of Hymenoptera. The Aculeata include the wasps, ants, and bees. Bees are similar to one group of wasps, the sphecoid wasps, but are quite unlike other Aculeata. Bees are usu-ally more robust and hairy than wasps (see Pls. 3-15), but some bees (e.g., Hylaeus,Pl. 1; Nomada,Pl. 2) are slender, sparsely haired, and sometimes wasplike even in col-oration. Bees differ from nearly all wasps in their depen-dence on pollen collected from flowers as a protein source to feed their larvae and probably also for ovarian devel-opment by egg-laying females. (An exception is a small clade of meliponine bees of the genus Trigona, which use carrion instead of pollen.) Unlike the sphecoid wasps, bees do not capture spiders or insects to provide food for their offspring. Thus nearly all bees are plant feeders; they have abandoned the ancestral carnivorous behavior of sphecoid wasp larvae. (Adult wasps, like bees, often visit flowers for nectar; adult sphecoid wasps do not collect or eat pollen.)
Bees and the sphecoid wasps together constitute the superfamily Apoidea (formerly called Sphecoidea, but see Michener, 1986a). The Apoidea as a whole can be recog-nized by a number of characters, of which two are the most conspicuous: (1) the posterior pronotal lobe is dis-tinct but rather small, usually well separated from and be-low the tegula; and (2) the pronotum extends ventrally as a pair of processes, one on each side, that encircle or nearly encircle the thorax behind the front coxae. See Section 10
for explanations of morphological terms and Section 12 for more details about the Apoidea as a whole.
As indicated above, the Apoidea are divisible into two groups: the sphecoid wasps, or Spheciformes, and the bees, or Apiformes (Brothers, 1975). Structural charac-ters of bees that help to distinguish them from sphecoid wasps are (1) the presence of branched, often plumose, hairs, and (2) the hind basitarsi, which are broader than the succeeding tarsal segments. The proboscis is in gen-eral longer than that of most sphecoid wasps. The details, and other characteristics of bees, are explained in Sec-tion 12.
A conveniently visible character that easily distin-guishes nearly all bees from most sphecoid wasps is the golden or silvery hairs on the lower face of most such wasps, causing the face to glitter in the light. Bees almost never exhibit this characteristic, because their facial hairs are duller, often erect, often plumose, or largely absent. This feature is especially useful in distinguishing small, wasplike bees such as Hylaeusfrom similar-looking sphe-coid wasps such as the Pemphredoninae.
The holophyletic Apiformes is believed to have arisen from the paraphyletic Spheciformes. Holophyletic is used here to mean monophyletic in the strict sense. Such a group (1) arose from a single ancestor that would be considered a member of the group, and (2) includes all taxa derived from that ancestor. Groups termed Para-phyleticalso arose from such an ancestor but do not in-clude all of the derived taxa. (See Sec. 16.)
2. What Are Bees?
Probably the most important activity of bees, in terms of benefits to humans, is their pollination of natural vegeta-tion, something that is rarely observed by nonspecialists and is almost never appreciated; see Section 6. Of course the products of honey bees—i.e., wax and honey plus small quantities of royal jelly—are of obvious bernefit, but are of trivial value compared to the profoundly im-portant role of bees as pollinators. Most of the tree species of tropical forests are insect-pollinated, and that usually means bee-pollinated. A major study of tropical forest pollination was summarized by Frankie et al. (1990); see also Jones and Little (1983), Roubik (1989), and Bawa (1990). In temperate climates, most forest trees (pines, oaks, etc.) are wind-pollinated, but many kinds of bushes, small trees, and herbaceous plants, including many wild flowers, are bee-pollinated. Desertic and xeric scrub areas are extremely rich in bee-pollinated plants whose preser-vation and reproduction may be essential in preventing erosion and other problems, and in providing food and cover for wildlife. Conservation of many habitats thus de-pends upon preservation of bee populations, for if the bees disappear, reproduction of major elements of the flora may be severely limited.
Closer to our immediate needs, many cultivated plants are also bee-pollinated, or they are horticultural varieties of pollinated plants. Maintenance of the wild, bee-pollinated populations is thus important for the genetic diversity needed to improve the cultivated strains. Gar-den flowers, most fruits, most vegetables, many fiber crops like flax and cotton, and major forage crops such as alfalfa and clover are bee-pollinated.
Some plants require bee pollination in order to pro-duce fruit. Others, commonly bee-pollinated, can self-pollinate if no bees arrive; but inbreeding depression is a frequent result. Thus crops produced by such plants are usually better if bee-pollinated than if not; that is, the numbers of seeds or sizes of fruits are enhanced by polli-nation. Estimates made in the late 1980s of the value of insect-pollinated crops (mostly by bees) in the USA ranged from $4.6 to $18.9 billion, depending on various assumptions on what should be included and how the es-timate should be calculated. Also doubtful is the eses-timate that 80 percent of the crop pollination by bees is by honey bees, the rest mostly by wild bees. But whatever estimates one prefers, bee pollination is crucially important (see O’Toole, 1993, for review), and the acreages and values of insect-pollinated crops are increasing year by year.
Wild bees may now become even more important as pollinators than in the past, because of the dramatic de-crease in feral honey bee populations in north-temperate climates due to the introduction into Europe and the Americas of mites such as Varroaand tracheal mites, which are parasites of honey bees. Moreover, there are var-ious crops for which honey bees are poor pollinators pared to wild bees. Examples of wild bees already com-mercially used are Osmia cornifrons (Radoszkowski), which pollinates fruit trees in Japan, Megachile rotundata (Fabricius), which pollinates alfalfa in many areas, Bom-bus terrestris(Linnaeus), which pollinates tomatoes in Eu-ropean greenhouses, and other Bombusspecies that do the same job elsewhere. O’Toole (1993) has given an account of wild bee species that are important in agriculture, and the topic was further considered by Parker, Batra, and Te-pedino (1987), Torchio (1991), and Richards (1993). Since honey bees do not sonicate tubular anthers to ob-tain pollen (i.e., they do not buzz-pollinate; see Sec. 6), they are not effective pollinators of Ericaceae, such as blueberries and cranberries, or Solanaceae such as egg-plants, chilis, and tomatoes.
Many bees are pollen specialists on particular kinds of flowers, and even among generalists, different kinds of bees have different but often strong preferences. There-fore, anyone investigating the importance of wild bees as pollinators needs to know about kinds of bees. The clas-sification presented by this book can suggest species to consider; for example, if one bee is a good legume polli-nator, a related one is likely to have similar behavior. Pro-boscis length is an important factor in these considera-tions, for a bee with a short proboscis usually cannot reach nectar in a deep flower, and probably will not take pollen there either, so is unlikely to be a significant pollinator of such a plant.
In many countries the populations of wild bees have been seriously reduced by human activity. Destruction of natural habitats supporting host flowers, destruction of nesting sites (most often in soil) by agriculture, roadways, etc., and overuse of insecticides, among other things, ap-pear to be major factors adversely affecting wild bee pop-ulations. Introduction or augmentation of a major com-petitor for food, the honey bee, has probably also affected some species of wild bees. Recent accounts of such prob-lems and some possible solutions were published by Banaszak (1995) and Matheson et al. (1996); see also O’Toole (1993).
3. The Importance of Bees
As in all insects that undergo complete metamorphosis, each bee passes through egg, larval, pupal, and adult stages (Fig. 4-1).
The haplodiploid system of sex determination has had a major influence on the evolution of the Hymenoptera. As in most Hymenoptera, eggs of bees that have been fer-tilized develop into females; those that are unferfer-tilized de-velop into males. Sex is controlled by alleles at one or a few loci; heterozygosity at the sex-determining locus (or loci) produces females. Development without fertiliza-tion, i.e., with the haploid number of chromosomes, pro-duces males, since heterozygosity is impossible. Inbreed-ing results in some diploid eggs that are homozygous at the sex-determining loci; diploid males are thus pro-duced. Such males are ordinarily reproductively useless, for they tend to be short-lived (those of Apisare killed as larvae) and to have few sperm cells; moreover, they may produce triploid offspring that have no reproductive po-tential. Thus for practical purposes the sex-determining mechanism is haplodiploid.
When she mates, a female stores sperm cells in her spermatheca; she usually receives a lifetime supply. She can then control the sex of each egg by liberating or not liberating sperm cells from the spermatheca as the egg passes through the oviduct.
Because of this arrangement, the female (of species whose females are larger than males) is able to place fe-male-producing eggs in large cells with more provisions, male-producing eggs in small cells. In Apis,the males of which are larger than the workers, male-producing cells are larger than worker-producing cells and presumably it is the cell size that stimulates the queen to fertilize or not to fertilize each egg. Moreover, among bees that construct cells in series in burrows, the female can place male-pro-ducing eggs in cells near the entrance, from which the re-sultant adults can escape without disturbing the slower-developing females. The number of eggs laid during her lifetime by a female bee varies from eight or fewer for some solitary species to more than a million for queens of some highly social species. Females of solitary bees give care and attention to their few offspring by nest-site se-lection, nest construction, brood-cell construction and provisioning, and determination of the appropriate sex of the individual offspring. Of course, it is such atttention to the well-being of offspring that makes possible the low reproductive potential of many solitary bees.
The eggs of nearly all bees are elongate and gently curved, whitish with a soft, membranous chorion (“shell”) (Fig. 4-1a), usually laid on (or rarely, as in Lithur-gus, within) the food mass provided for larval consump-tion. In bees that feed the larvae progressively (Apis, Bom-bus,and most Allodapini), however, the eggs are laid with little or no associated food. Eggs are commonly of mod-erate size, but are much smaller in highly social bees, which lay many eggs per unit time, and in Allodapula (Al-lodapini), which lays eggs in batches, thus several eggs at about the same time. Eggs are also small in many clep-toparasitic bees (see Sec. 8) that hide their eggs in the
brood cells of their hosts, often inserted into the walls of the cells; such eggs are often quite specialized in shape and may have an operculum through which the larva emerges (see Sec. 8). Conversely, eggs are very large in some sub-social or primitively eusub-social bees like Braunsapis (Allo-dapini) and Xylocopa(Xylocopini). Indeed, the largest of all insect eggs are probably those of large species of Xylo-copa,which may attain a length of 16.5 mm, about half the length of the bee’s body. Iwata and Sakagami (1966) gave a comprehensive account of bee egg size relative to body size.
The late-embryonic development and hatching of eggs
4. Development and Reproduction
4
Figure 4-1.Stages in the life cycle of a leafcutter bee, Megachile brevis Cresson.a,Egg;b-d,First stage, half-grown, and mature lar-vae;e,Pupa;f,Adult. From Michener, 1953b.
a
b
c
d
e
has proved to be variable among bees and probably rele-vant to bee phylogeny. Torchio, in various papers (e.g., 1986), has studied eggs of several different bee taxa im-mersed in paraffin oil to render the chorion transparent. Before hatching, the embryo rotates on its long axis, ei-ther 90˚ or 180˚. In some bees (e.g., Nomadinae) the chorion at hatching is dissolved around the spiracles, then lengthwise between the spiracles; eventually, most of the chorion disappears. In others the chorion is split but oth-erwise remains intact.
Larvaeof bees are soft, whitish, legless grubs (Fig. 4-1b-d). In mass-provisioning bees, larvae typically lie on the upper surface of the food mass and eat what is below and in front of them, until the food is gone. They commonly grow rapidly, molting about four times as they do so. The shed skins are so insubstantial and hard to observe that for the great majority of bees the number of molts is un-certain. For the honey bee (Apis)there are five larval in-stars (four molts before molting into the pupal stage); and five is probably the most common number in pub-lished reports such as that of Lucas de Olivera (1960) for Melipona.In some bees, e.g., most nonparasitic Apinae other than the corbiculate tribes (i.e., in the old An-thophoridae), the first stage remains largely within the chorion, leaving only four subsequent stages (Rozen, 1991b); such development is also prevalent in the Mega-chilidae. In the same population of Megachile rotundata (Fabricius) studied by Whitfield, Richards, and Kveder (1987), some individuals had four instars and others five. The first to third instars were almost alike in size in the two groups, but the terminal fourth instar was intermediate in size between the last two instars of five-stage larvae.
Markedly different young larvae are found in most cuckoo bees, i.e., cleptoparasitic bees. These are bees whose larvae feed on food stored for others; details are presented in Section 8. Young larvae of many such para-sites have large sclerotized heads and long, curved, pointed jaws with which they kill the egg or larva of the host (Figs. 82-5, 89-6, 103-3). They then feed on the stored food and, after molting, attain the usual grublike form of bee larvae.
Other atypical larvae are those of allodapine bees, which live in a common space, rather than as a single larva per cell, and are mostly fed progressively. Especially in the last instar, they have diverse projections, tubercles, large hairs, and sometimes long antennae that probably serve for sensing the movements of one another and of adults, and obviously function for holding masses of food and re-taining the larval positions in often vertical nest burrows (Fig. 88-6). Many of the projections are partly retracted when the insect is quiet, but when touched with a probe or otherwise disturbed, they are everted, probably by blood pressure.
It has been traditional to illustrate accounts of bee lar-vae (unfortunately, this is largely not true for adults). The works of Grandi (culminating in Grandi, 1961), Mich-ener (1953a), McGinley (1981), and numerous papers by Rozen provide drawings of mature larvae of many species. Various other authors have illustrated one or a few larvae each. Comments on larval structures appear as needed later in the phylogenetic and systematic parts of this book. Unless otherwise specified, such statements always
concern mature larvae or prepupae. Accounts of larvae are listed in a very useful catalogue by McGinley (1989), or-ganized by family, subfamily, and tribe. It is therefore un-necessary except for particular cases to cite references to papers on larvae in this book, and such citations are mostly omitted to save space.
As in other aculeate Hymenoptera, the young larvae of bees have no connection between the midgut and the hindgut, so cannot defecate. This arrangement probably arose in internal parasitoid ancestors of aculeate Hy-menoptera, which would have killed their hosts prema-turely if they had defecated into the host’s body cavity. In some bees defecation does not begin until about the time that the food is gone; in others, probably as a derived con-dition, feces begin to be voided well before the food sup-ply is exhausted. After defecation is complete the larva is smaller and often assumes either a straighter or a more curled form than earlier and becomes firmer; its skin is less delicate, and any projections or lobes it may have are commonly more conspicuous (Fig. 4-2). This last part of the last larval stage is called the prepupaordefecated larva; this stage is not shown in Figure 4-1. Most studies of larvae, e.g., those by Michener (1953a) and numerous studies by Rozen, are based on such larvae, because they are often available and have a rather standard form for each species; feeding larvae are so soft that their form fre-quently varies when preserved. Prepupae are often the stage that passes unfavorable seasons, or that survives in the cell for one to several years before development re-sumes. Houston (1991b), in Western Australia, recorded living although flaccid prepupae of Amegilla dawsoni (Rayment) up to ten years old; his attempts to break their diapause were not successful. Such long periods of devel-opmental stasis probably serve as a risk-spreading strat-egy so that at least some individuals survive through long periods of dearth, the emergence of adults being some-how synchronized with the periodic blooming of vegeta-tion. Even in nondesertic climates, individuals of some species remain in their cells as prepupae or sometimes as adults for long periods. Thus Fye (1965) reported that in a single population and even in a single nest of Osmia atriventrisCresson in Ontario, Canada, some individuals emerge in about one year, others in two years.
Mature larvae of many bees spin cocoons, usually at about the time of larval defecation, much as is the case in sphecoid wasps. The cocoons are made of a framework of silk fibers in a matrix that is produced as a liquid and then solidifies around the fibers; the cocoon commonly con-sists of two to several separable layers. Various groups of bees, including most short-tongued bees, have lost co-coon-spinning behavior and often are protected instead by the cell lining secreted by the mother bee. Cocoon spinning sometimes varies with the generation. Thus in Microthurge corumbae(Cockerell), even in the mild cli-mate of the state of São Paulo, Brazil, the cocoons of the overwintering generation are firm and two-layered but those of the other generation consist of a single layer of silk (Mello and Garófalo, 1986). Similar observations were made in California by Rozen (1993a) on Spheco-dosoma dicksoni(Timberlake), in which larvae in one-lay-ered cocoons pupated without diapausing, whereas those in two-layered cocoons overwintered as prepupae. In
other cases, in a single population, some individuals make cocoons and others do not. Thus in Exomalopsis nitens Cockerell, those that do not make cocoons pupate and eclose promptly, but those that make cocoons diapause and overwinter (Rozen and Snelling, 1986).
When conditions are appropriate, pupation occurs; for all eusocial species and many others this means soon after larval feeding, defecation, and prepupal formation are completed. In other species pupation occurs only af-ter a long prepupal stage. Pupaeare relatively delicate, and their development proceeds rapidly; among bees the pupa is never the stage that survives long unfavorable pe-riods. Because they are delicate and usually available for short seasons only, fewer pupae than larvae have been pre-served and described. Pupal characters are partly those of the adults, but pupae do have some distinctive and useful characters of their own (see Michener, 1954a). Most con-spicuous are various spines, completely absent in adults, that provide spaces in which the long hairs of the adults develop. Probably as a secondary development, long spines of adults, like the front coxal spines of various bees, arise within pupal spines.
Adultsfinally appear, leave their nests, fly to flowers and mate, and, if females, according to species, either re-turn to their nests or construct new nests elsewhere. Many bees have rather short adult lives of only a few weeks. Some, however, pass unfavorable seasons as adults; if such periods are included, the adult life becomes rather long. For example, in most species of Andrena, pupation and
adult maturation ccur in the late summer or fall, but the resulting adults remain in their cells throughout the win-ter, leaving their cells and coming out of the ground in the spring or summer to mate and construct new nests. In most Halictinae, however, although pupation of repro-ductives likewise occurs in late summer or autumn, the resulting adults emerge, leave the nest, visit autumn flow-ers for nectar, and mate. The males soon die, but the fe-males dig hibernaculae (blind burrows), de novo or inside the old nest, for the winter. A few bees live long, relatively active adult lives. These include the queens of eusocial species and probably most females of the Xylocopinae and some solitary Halictinae. Among the Xylocopinae, a female Japanese Ceratinain captivity is known to have laid eggs in three different summer seasons, although only one was laid in the last summer (for summary, see Mich-ener, 1985b, 1990d). Females of some solitary Lasioglos-sum(Halictinae), especially in unfavorable climates (only a few sunny days per summer month, as in Dartmoor, En-gland) provision a few cells, stop by midsummer, and pro-vision a few more cells the following year (Field, 1996). Like the variably long inactivity of prepupae described above, this may be a risk-spreading strategy.
The male-female interactions among bees are diverse; they must have evolved to maximize access of males to re-ceptive females and of females to available males. The mating system clearly plays a major role in evolution. Re-views are by Alcock et al. (1978) and Eickwort and Gins-berg (1980); the following account lists only a few exam-Figure 4-2.Change of a ma-ture larva to a prepupa shown by last larval stadium of Neff-apis longilongua Ruz.a, Pre-defecating larva; b, Postdefe-cating larva or prepupa. (The abdominal segments are num-bered.) From Rozen and Ruz, 1995.
a
ples selected from a considerable literature. Many male bees course over and around flowers or nesting sites, pouncing on females. In other species females go to par-ticular types of vegetation having nothing to do with food or nests and males course over the leaves, pouncing on females when they have a chance. In these cases mating occurs quickly, lasting from a few seconds to a minute or two, and one’s impression is that the female has no choice; the male grasps her with legs and often mandibles and mates in spite of apparent struggles. The male, however, may be quite choosy. In Lasioglossum zephyrum(Smith), to judge largely by laboratory results, males over the nest-ing area pounce on small dark objects includnest-ing females of their own species, in the presence of the odor of such females, but do so primarily when stimulated by un-familiar female odor, thus presumably discriminating against female nestmates, close relatives of nestmates, and perhaps females with whom they have already mated (Michener and Smith, 1987). Such behavior should pro-mote outbreeding. Conversely, it would seem, males are believed to fly usually over the part of the nesting area where they were reared; they do not course over the whole nesting aggregation (Michener, 1990c). Such behavior should promote frequent inbreeding, since males would often encounter relatives, yet they appear to discriminate against their sisters. The result should be some optimum level of inbreeding.
In communal nests of Andrena jacobiPerkins studied in Sweden, over 70 percent of the females mated within the nests with male nestmates (Paxton and Tengö, 1996). Such behavior, with its potential for inbreeding, may be common in communal bees. Given the rarity with which one sees mating in most species of bees, it may be that mating in nests is also common in some solitary species. In species that have several sex-determining loci, in-breeding may not be particularly disadvantageous, be-cause deleterious genes tend to be eliminated by the hap-loid-male system.
In some bees, females tend to mate only once. Males in such species attempt to mate with freshly emerged young females, even digging into the ground to meet them, as in Centris pallidaFox (Alcock, 1989) or Colletes cunicular-ius(Linnaeus) (Cane and Tengö, 1981). In other species females mate repeatedly. The behavior of males suggests that there is sperm precedence such that sperm received from the last mating preferentially fertilize the next egg. Males either (1) mate again and again with whatever fe-males they can capture, as in Dianthidium curvatum (Smith) (Michener and Michener, 1999), or (2) remain in copula for long periods with females as they go about their foraging and other activities, thus preventing the fe-males from mating with other fe-males (many Panurginae, personal observation).
In Colletes cunicularius (Linnaeus), Lasioglossum zephyrum(Smith), Centris pallidaFox, and many others, female-produced pheromones seem to stimulate or attract males, but in Xylocopa varipuncta Patton a male-pro-duced pheromone attracts females to mating sites (Alcock and Smith, 1987). Some male Bombusscent-mark a path that they then visit repeatedly for females (Haas, 1949). In other species of Bombus,those with large-eyed males, the males wait on high perches and dash out to passing objects includingBombusfemales (Alcock and Alcock, 1983).
Al-though playing a role in all cases, vision is no doubt espe-cially important also in other bees with large-eyed males, such as Apis melliferaLinnaeus, the males of which fly in certain congregating areas and mate with females that come to those areas; see also the comments on mating swarms of large-eyed males in Section 28.
Most male bees can mate more than once, but in Meliponini and Apini the male genitalia or at least the en-dophallus is torn away in mating, so that after the male mates he soon dies.
Males in many species of bees in diverse families have enlarged and modified legs, especially the hind legs (see Sec. 28), or broad heads with long, widely separated mandibles. These are features that help in holding females for mating, and may be best developed in large males. Many males of Megachilehave elaborately enlarged, flat-tened, pale, fringed front tarsi (Fig. 82-19). Wittmann and Blochtein (1995) found epidermal glands in the front basitarsi; at mating these tarsi hold the female’s an-tennae, or cover her eyes. This behavior and gland prod-uct are presumably associated with successful mating or mate choice.
Large-headed males occur especially in some An-drenidae—both Andreninae and Panurginae—and in some Halictinae. Large heads appear to be characteristic of the largest individuals of certain species, no doubt as an allometric phenomenon. In two remarkable examples, one an American Macrotera (Panurginae) (Danforth, 1991b) and the other an Australian Lasioglossum (Chilal-ictus)(Halictinae) (Kukuk and Schwarz, 1988; Kukuk, 1997), the large-headed males (Figs. 4-3, 56-3, 56-4) have relatively short wings and are flightless nest inhabi-tants in communal colonies. The large-headed males also have large mandibles and fight to the death when more than one is present in a nest. Smaller males of each species have normal-sized wings and fly. Great size variation among males and macrocephaly may be most frequent in, or even limited to, communal species. Unlike most male bees that leave the nest permanently and mate elsewhere, short-winged males mate with females of their own colony. Thus such a male is often the last to mate with a female before she lays an egg.
In some other bees the male mating strategy also varies greatly with body size. Large males usually fly about the nesting sites, finding young females as they emerge from the ground or even digging them out of the ground, pre-sumably guided by odor. Small males seek females on flowers or in vegetation near the nesting area. Such dual behavior is documented for Centris pallidaFox (Alcock, 1989) in the Centridini, and for Habropoda depressa (Fowler) (Barthell and Daly, 1995) and Amegilla dawsoni (Rayment) (Alcock, 1996), both in the Anthophorini. Such behavior seems akin to that of Anthidium manica-tum(Linnaeus), in which large males have mating terri-tories that include flowers visited by females (Severing-haus, Kurtak, and Eickwort, 1981), whereas small ones are not territorial, and to that of certain Hylaeus(Alcock and Houston, 1996), in which large males with a strong ridge or tubercle on S3 are territorial whereas small ones with reduced ventral armature or none are not territorial. The ventral armature is apparently used to grasp an ad-versary against the thoracic venter by curling the meta-soma.
An interesting and widespread feature in Hymenop-tera is the prevalence of yellow (or white) coloration on the faces of males. If a black species has any pale col-oration at all, it will be on the face (usually the clypeus) of males. Species with other yellow markings almost al-ways have more yellow on the face of the male than on that of the female, although on the rest of the body yel-low markings often do not differ greatly between the sexes. Groups like Megachilethat lack yellow integumen-tal markings frequently have dense yellow or white hairs on the face of the male, but not on that of the female. In mating attempts males usually approach females from above or behind, so that neither sex has good views of the face of the other. Therefore I do not suppose that the male’s yellow face markings have to do with male-female recognition or mating. Rather, I suppose that they are in-volved in male-male interactions, when males face one another in disputes of various sorts. Sometimes, males of
closely related species, such as Xylocopa virginica (Lin-naeus) and californicaCresson, differ in that one (in this casevirginica) has yellow on the face but the other does not. Someone should study the male-male interactions in such species pairs. Presumably, male behavior linked to yellow male faces is found in thousands of species of Hy-menoptera.
Obviously, the variety of mating systems in bees de-serves further study, both because of its interest for bee evolution and for evolutionary theory. Moreover, because of the frequency of morphological or chromatic corre-lates, mating systems and such correlates are important for systematists.
Figure 4-3.Male morphs of Lasioglossum (Chilalictus) hemichal-ceum (Cockerell) from Australia. a,Ordinary male; b, c,Heads of same;d,Large, flightless male; e,Head of same. From Houston, 1970.
a
b
c
Many works treat aspects of behavior of diverse kinds of bees. Specialized papers are cited throughout this book; some more general treatments are the books by Friese (1923), with its interesting colored plates of nests of Eu-ropean bees; Iwata (1976), with its review of previous work on the behavior of bees and other Hymenoptera; and O’Toole and Raw (1991), which offers readable ac-counts and fine illustrations of bees worldwide. A major aspect of behavior involves intraspecific interactions, i.e., social behavior in a broad sense. Courtship and mating are treated briefly in Section 4. Here we consider colonial behavior—its origin as well as its loss.
Some female bees are solitary; others live in colonies. Asolitarybee constructs her own nest and provides food for her offspring; she has no help from other bees and usu-ally dies or leaves before the maturation of her offspring. Sometimes such a female feeds and cares for her offspring rather than merely storing food for them; such a rela-tionship is called subsocial.
Acolonyconsists of two or more adult females, irre-spective of their social relationships, living in a single nest. Frequently the females constituting a colony can be di-vided into (1) one to many workers, which do most or all of the foraging, brood care, guarding, etc., and are often unmated; and (2) one queen, who does most or all of the egg laying and is usually mated. The queen is often, and in some species always, larger than her workers, but some-times the difference is only in mean size. In some social halictines, the largest females have extraordinarily large heads, often with toothed genae or other cephalic modi-fications probably resulting from allometry.
For many people, bees are thought of as stinging, honey-producing social insects living in perennial colonies, each of which consists of a queen and her many daughter workers. This is indeed the way of life for the honey bees (genus Apis) and the stingless honey bees (Trigona, Melipona, etc.) of the tropics. Queens and workers in these cases are morphologically very different, and the queen is unable to live alone (e.g., she never for-ages); nor do workers alone form viable colonies (they cannot mate and therefore cannot produce female off-spring). These are the highly eusocialbees. Such bees al-ways live in colonies, and new colonies are established so-cially, by groups or swarms. Only two tribes, the Apini and the Meliponini (family Apidae), consist of such bees. Most bumble bees (Bombini) and many sweat bees (Halictinae) and carpenter bees and their relatives (Xylo-copinae) may live in small colonies, mostly started by sin-gle females working as solitary individuals performing all necessary functions of nest construction, foraging, provi-sioning cells or feeding larvae progressively, and laying eggs. Later, on the emergence of daughters, colonial life may arise, including division of labor between the nest foundress (queen) and workers. These are primitively eu-socialcolonies. Queens and workers are essentially alike morphologically, although often differing in size; they differ more distinctly in physiology and behavior. Such colonies usually break down with production of
repro-ductives; thus the colonies are obligately temporary rather than potentially permanent like those of highly so-cial bees.
Since in primitively social bees the individual that be-comes the queen cannot always be recogized until she has workers, the word gynehas been introduced for both po-tential queens and functional queens. The word is most frequently used for females that will or may become queens (Michener, 1974a), but have not yet done so. Thus it is proper to say that a gyne establishes her nest by herself in the spring, and becomes a queen when the colony develops.
Both permanent honey bee colonies and temporary bumble bee or halictine colonies are called eusocial,
meaning that they have division of labor (egg-layer vs. foragers) among cooperating adult females of two gener-ations, mothers and daughters. Such a definition is ade-quate for most bees; there is currently much discussion of modifying the definition of “eusocial” and relevant terms to make them useful across the board for all groups of so-cial animals or, alternatively, eliminating them in favor of a system of terms that addresses the social levels among diverse animal species as well as the variability within species (see Crespi and Yanaga, 1995; Gadagkar, 1995; Sherman et al., 1995; Costa and Fitzgerald, 1996; and Wcislo, 1997a).
Not all bees that live in colonies are eusocial. Some-times a small colony consists of females of the same gen-eration, probably often sisters, that show division of la-bor, with a principal egg-layer or queen and one or more principal foragers or workers. Such colonies, called semi-social, may not be worth distinguishing from primitively eusocial colonies. As noted below, they often arise when the queen of a primitively eusocial colony dies and her daughters carry on, one of them commonly mating and becoming the principal egg-layer or replacement queen. Some bee colonies lack division of labor or castes: all colony members behave similarly. Some such colonies are
communal;two or more females use the same nest, but each makes and provisions her own cells and lays an egg in each of them. In most or all species that have commu-nal colonies, other individuals in the same populations nest alone, and are truly solitary. Thus colonial life is fac-ultative. A possible precursor of communal behavior arises when a nest burrow, abandoned by its original oc-cupant, is then occupied by another bee of the same species (Neff and Rozen, 1995). Such behavior is rarely reported, because without marked bees, one does not know of it. A condition that appears to promote com-munal behavior is very hard soil or other substrate, be-cause it is much easier to join other bees in a preexisting nest than to excavate a new nest starting at the surface (Michener and Rettenmeyer, 1956; Bennett and Breed, 1985).
A little-used additional term is quasisocial.It applies to the relatively rare case in which a few females occupy-ing a nest cooperate in buildoccupy-ing and provisionoccupy-ing cells, but different individuals (as opposed to a single queen)
5. Solitary versus Social Life
lay eggs in cells as they are completed. That is, all the fe-males have functional ovaries, mate, and can lay eggs. This may not be the terminal or most developed social state for any species of bees, but at times some colonies exhibit this condition.
When one opens a nest containing a small colony of bees, it is often impossible to recognize the relationships among the adult female inhabitants. The colony might be communal, quasisocial, or semisocial. Only observations and dissections will clarify the situation. Such colonies can be called parasocial,a noncommittal umbrella term used for a colony whose members are of a single genera-tion and interact in any of the three ways indicated or in some as yet unrecognized way. At first, primitively euso-cial colonies may look like parasoeuso-cial colonies, but one in-dividual, the queen (mother), is older, more worn, and sometimes larger than the others, which are workers (daughters). The queen commonly has enlarged ovaries and sperm cells in the spermatheca; workers usually do not.
Because many species pass through ontogenetic stages of sociality or are extremely variable in this regard, terms like “eusocial” should be applied to colonies, not species, except when dealing with permanently highly social forms like Apisand the Meliponini. For example, a nest may contain a single female, a gyne who has provided for and is protecting her immature progeny in a subsocial re-lationship. After emergence of the first adult workers, however, the nest contains a eusocial colony. There are species of Halictinae that have eusocial colonies in warmer climates but are solitary in cold climates (Eick-wort et al., 1996). A single population may consist of some individuals functioning like solitary bees while oth-ers are eusocial, as observed in a New York population of Halictus rubicundus(Christ) (Yanega, 1988, 1989). In most Allodapini and some Ceratinini (in the Xylo-copinae), although nests harboring colonies of two or more cooperating adult females exist, most nests contain a lone adult, rearing her young subsocially without ben-efit of a worker or other adult associate (Michener, 1990b). In the nests containing two or more adults, one is often the principal layer, thus a queen, and the others (or one of them), principal foragers and often unmated, and thus workers. Later, if the queen dies, one of the workers may become a queen; the result is a semisocial colony of sisters. But if several or all of the sisters become reproductive, the result is a quasisocial colony. Of course there are sometimes intergradations or mixtures. In such bees eusocial and other social relationships have arisen even though most individuals of the species never experi-ence cooperative behavior among adult females.
The terminology summarized above is not always helpful; I introduce it here because some of the terms are often found in the literature and are used later in this book. A case in which the terminology (“communal,” “semisocial,” etc.) is not useful is found in the autumnal colonies of Exoneura bicolorSmith in Australia (Melna and Schwarz, 1994). The bees in such colonies can be di-vided into four classes, according to their activities. Yet there is no reproductive activity at this time; thus there is no queenlike or workerlike division of labor, but rather division along other lines. It may be that in the
Allodap-ini, whenever two or more adults nest together, some sort of division of labor ensues.
Many kinds of bees that nest in the ground construct numerous nests in limited areas; a patch of earth, a path, or an earthen bank may be peppered with their holes (Fig. 5-1). Such groupings of individual nests are called aggre-gations.Each burrow may be made and inhabited by one female or may contain some sort of small colony (Fig. 5-2). Some aggregations doubtless result from the avail-ability of local patches of suitable soil, but often the bees choose to aggregate in only part of an extensive area that appears uniform. Sometimes, gregarious behavior seems to be a response to the presence of other bees or bee nests—thus a social phenomenon; see Michener (1974a). In other cases bees may be returning to the site of their own emergence or “birth.” In the literature, aggregations are sometimes called “colonies.” I think it is best to avoid this usage and to limit the word “colony” as indicated above.
The above is a brief account of a large topic, the social diversity found among bees. Additional information and sources can be found in Michener, 1974a, 1985b, 1990c, d. The great abundance of the highly social forms (honey bees, stingless bees) almost wherever they occur suggests that such sociality itself is an enormous advantage in the presumed competition with other bees. The great body of literature on the theory of eusocial behavior of insects mostly addresses in one way or another the problem of how it is possible for attributes like those of workers to evolve and be passed on from generation to generation, even though they decrease the probability of their bearer’s leaving progeny. Briefly expressed, an individual’s overall
