Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00106
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 1981
Copyright Date: 1917
Subject: Florida Entomological Society
Entomology -- Periodicals
Insects -- Florida
Insects -- Florida -- Periodicals
Insects -- Periodicals
General Note: Eigenfactor: Florida Entomologist: http://www.bioone.org/doi/full/10.1653/024.092.0401
 Record Information
Bibliographic ID: UF00098813
Volume ID: VID00106
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Full Text

(ISSN 0015-4040)

(An International Journal for the Americas)

Volume 64, No. 1 March, 1981


LLOYD, J. E.-Pragmatic Insect Behavioral Ecology: A Not-So-Odd
Coupling .--.....-..------------------------------- 1
GREENFIELD, M. D.-Moth Sex Pheromones: An Evolutionary Perspec-
tive ......--....--..-..-----.............--------------------- 4
WALKER, T. J., AND S. A. WINERITER-Marking Techniques for Recog-
nizing Individual Insects ......-..-- .--.-------------------. 18
BURK, T.-Signaling and Sex in Acalyptrate Flies .............................----... 30
SULLIVAN, R. T.-Insect Swarming and Mating .........................-----. 44
ADAMS, G.-Search Paths of Fireflies in Two Dimensions ........................ 66
STRASSMAN, J. E.-Wasp Reproduction and Kin Selection: Repro-
ductive Competition and Dominance Hierarchies Among Polistes
annularis Foundresses ..----.......--.... ----------------. 74
LLOYD, J. E.-Sexual Selection: Individuality, Identification, and Rec-
ognition in a Bumblebee and Other Insects ...............................--- 89

GORDON, R. D., AND D. M. ANDERSON-The Species of Scarabaeidae
(Coleoptera) Associated with Sugarcane in South Florida ......... 119
FRANK, J. H., AND M. C. THOMAS-Myrmedoniini (Coleoptera, Staphy-
linidae, Aleocharinae) Associated With Army Ants (Hymeno-
ptera, Formicidae, Ecitoninae) in Florida ........--.....................- 138
DENMARK, H. A., AND K. L. ANDREWS-Phytoseiidae of El Salvador,
Central America (Acarina: Mesostigmata) ................................... 147
FAIRCHILD, G. B., AND R. C. WILKERSON-New Species of Esenbeckia
(Proboscoides) (Diptera: Tabanidae) With a Key to the Sub-
genus ---.. ...... ... ................................................................ ....... 158
PORTER, C. C.-Ecological Notes on Lower Rio Grande Valley Xylocopa
(Hymenoptera: Anthophoridae) .......-.......-- .....--- .......................... 175
TODD, E. L., AND R. W. POOLE-Moths of the parvula Complex of Laci-
nipolia McDunnough (Noctuidae: Hadenidae) ................................ 183
MUCHMORE, W. B.-Pseudoscorpions from Florida and the Caribbean
Area. 11. A New Parachelifer from the Virgin Islands (Cheli-
feridae) ...............-------.....--------.......... ...................... .---......... 189
Continued on Back Cover

Published by The Florida Entomological Society


President --. ----- -......- -.........---------------................. E. C. Beck
Vice-President ....-- ...............------------.. ............-----............ W. L. Peters
Secretary -----------............. ........................ .................. F. W. Mead
Treasurer ..........---------........................................ ......... D. P. W ojcik

R. E. Brown
N. C. Leppla
Other Members of Executive Committee ........... R. H. Maltby
C. A. Musgrave Sutherland
J. L. Taylor


Editor ...-------. .--.......................................-. C. A. Musgrave Sutherland
Associate Editors -...-...-- ...-......---------- ... ........... .. ... J. E. Lloyd
J. R. McLaughlin
C. W. McCoy
A. R. Soponis
H. V. Weems, Jr.
Business M manager ............................... ........ ....... ........ ..... D. P. W ojcik

FLORIDA ENTOMOLOGIST is issued quarterly-March, June, September,
and December. Subscription price to non-members is $15.00 per year in
advance, $5.00 per copy. Membership in the Florida Entomological Society,
including subscription to Florida Entomologist, is $10 per year for regular
membership and $2 per year for students. Inquiries regarding membership
and subscriptions should be addressed to the Business Manager, P. O. Box
12425, University Station, Gainesville, FL 32604. Florida Entomologist is
entered as second class matter at the Post Office in DeLeon Springs and
Gainesville, FL.
Authors should consult "Instructions to Authors" on the inside cover of
all recent issues while preparing manuscripts or notes. When submitting a
paper or note to the Editor, please send the original manuscript, original
figures and tables, and 3 copies of the entire paper. Include an abstract and
title in Spanish, if possible. Upon receipt, manuscripts and notes are ac-
knowledged by the Editor and assigned to an appropriate Associate Editor
who will make every effort to recruit peer reviewers not employed by the
same agency or institution as the authors(s). Reviews from individuals
working out-of-state or in nearby countries (e.g. Canada, Mexico, and
others) will be obtained where possible.'
Manuscripts and other editorial matter should be sent to the Editor,
C. A. Musgrave Sutherland, Rt. 3, Box 115H, Las Cruces, NM 88001.
Business matters for other Society officers can be sent to that individual at
the University Station address above.

This issue mailed April 22, 1981

Insect Behavioral Ecology-'80 Lloyd



"Art for art's sake makes no more sense
than gin for gin's sake"-Somerset Maugham

The intent and general philosophy behind these Symposia is that they
should couple "basic" and "applied" insect biology. These papers intend to
facilitate the transfer of concepts, ideas, facts, problems, technology, and
enthusiasm, from the basic-end to the mission-end (actually the raison d'etre-
end) of Entomology, and vice versa. It is usually from so-called basic research
that the fundamentally new ideas that lead to leaps in entomological problem-
understanding, if not problem-solving, have their origin; and it can be
argued that it is from ignorance of basic knowledge already available, that
Entomology has suffered its most acute embarrassments.
The relationship of basic information to practical application and prob-
lem-solution, is far from 1 to 1. This is at the core of, and primarily re-
sponsible for, our most painful carbuncle. Only a few of the thousands of
papers and ideas that are published in a truly basic vein ever have recog-
nized significance for (mission) Entomology. Moreover, which few and what
significance cannot generally be predicted "cost-effectively." The return from
investment in such basic research is not only low, but usually slow, measured
in decades, and often indirect. A practical Entomology accepts, and works
with, but tries to improve this, and recognizes the chancy, even whimsical
nature of the serendipity in it. Unfortunately, this is merely the culture
medium, and Entomology must also cope with the infectious organisms them-
selves: inept, incompetent, pusilanimous administrators and vandalizing
politicians who wouldn't survive or be tolerated in a logical, reasonable, and
truly cost-effective system.1 The Entomologists' carbuncle is of special sig-
nificance for behavioral ecologists. While they are often recognized by out-
siders and detractors as the enthusiastic, fun-having, eccentric (often sex-
oriented) naturalists that they are (and should be), they are not recognized
for their value and the contributions that they make to Entomology.
Pogo might have said, "We have met the problem, and it is us." We en-
courage and nurture our infection when we see ourselves from the perspec-
tives of the artificial dichotomy "basic versus applied." A major government
granting agency struggles with the problem of whether basic research can
be performed on an economically important (economic!) insect. (Oi gevalt!
Research should have the dollars they spend on their study.) The similarities
of their perceived dichotomy of "basic versus applied," and the moldy instinct
versus learned one, are obvious. Neither is real, neither is heuristic, neither
generates viable approaches to problems, and, in fact, both are counterpro-
ductive, scientifically and sociologically. Behaviorists now have a meaningful
and useful concept-heritability-to heal their formerly festering sore. Let's
find a suitable one for insect biology.
Let's be Pragmatic. Like the Venetian call-girl that became a World-class

1To say nothing of the pandering elements of a free-enterprising press.

Florida Entomologist 64(1)

swimmer from the practice she got working both sides of the canal, let's be-
come World-class Entomologists by spending more time going back and
forth between their basic and applied houses, than in them.
There are other aspects of Pragmatic Entomology. We should spend more
time communicating and working directly with the public, and especially in
encouraging real research by amateurs. And, we should recognize that it is a
point of honor as well as of necessity that accusations and attacks by
vandals, and the incompetence of those who are over as well as within our
groups, must be thoroughly and vigorously dealt with. In these Pragmatic
Insect Behavioral Ecology Symposia we will primarily be addressing the
coupling aspect, but will not neglect other problems. We welcome and en-
courage submitted outlines for papers that will rigorously address these
issues in a substantive way.
The participants of Symposium-80, like those of Symposium-79, have
made a valuable contribution to Pragmatic Entomology, and to the develop-
ment of insect behavioral ecology as a progressive, enlightened, scholarly, and
major analytical, synthetic, and philosophical subdiscipline of Entomology.
They have attempted to make their papers scientific yet interesting; stimulat-
ing and readable, without sacrificing professionalism. We are in agreement
with Fabre: "Others have reproached me with my style, which has not the
solemnity, nay, better the dryness of the schools. They fear least a page that
is read without fatigue should not always be the expression of the truth.
Were I to take their word for it, we are profound only on condition of being

Each participant has reprints of his own paper. Paper-bound reprints of
the entire Symposium-80 are available in quantity from E. 0. Painter
Printing Co., DeLeon Springs, Fla. 32028 (25 at $67.00). (Reprints of
Symposium-79 are also available from the printer, at 25 for $65.00.)
At the annual fall meeting of the Society in 1981, Symposium-81 will
consider insect behavioral ecology a' propos insect pest management.
I thank the participants of the Symposium, and the Executive and Pro-
gram Committees of F.E.S., especially E. C. Beck, N. C. Leppla, and D. P.
Wojcik, for their enthusiastic cooperation, guidance, and support; C. Mus-
grave-Sutherland, editor of the Florida Entomologist, and Dick Johnston,
printer of the Florida Entomologist, for their patience, understanding and
cooperation; Barbara Hollien and Sheila Eldridge for their untiring and
congenial cooperation at the electronic typewriters; and my Symposium
referees, H. J. Brockmann, T. G. Forrest, and T. J. Walker. I especially thank
Norm Leppla and Tom Walker for their sustaining enthusiasm, encourage-
ment, and counsel. Florida Agricultural Experiment Station Journal Series
No. 2928.

2Fabre, J. H. 1918. The wonders of instinct. The Century Co., New York.

March, 1981

Insect Behavioral Ecology-'80 Lloyd




*t 'i



~ .I* ~FZ. ;.j.:'
*, ;"



j i


Symposium Participants
Back Row: M. D. Greenfield, T. Burk, R. T. Sullivan, C. S. Barfield, J. E.
Strassmann, G. Adams. Front Row: J. E. Lloyd, T. J. Walker. Photograph
by F. W. Mead, FDACS-DPI. 28 August 1980, Daytona Beach, FL.

Florida Entomologist 64 (1)



"The greater peacock moths cross hills and valleys in the darkness, with
a heavy flight of wings spotted with inexplicable hieroglyphics. They
hasten from the remotest depths of the horizon to find their 'sleeping
beauties,' drawn thereto by unknown odours, inappreciable by our senses,
yet so penetrating that the branch of almond on which the female has
perched, and which she has impregnated with her effluvium, exerts the
same extraordinary attraction" (J. H. C. Fabre in Legros 1971: 131).
"In the obscurity of a dark chamber this splendid moth emits phantasmal
radiations, perhaps intermittent and reserved for the season of nuptials,
signals invisible to us, and perceptible only to those children of the night,
who may have found this means to communicate one with another, to call
one another in the darkness, and to speak with one another" (J. H. C.
Fabr6 in Legros 1971: 309).
The past 20 years have witnessed a rapid expansion in our knowledge of
insect sex pheromones. Ever since Butenandt et al. (1961a, b) chemically
isolated and identified a sex pheromone emitted by female silkworm moths,
entomologists and chemists have been engaged in a concerted effort to de-
termine the structures of these compounds. Much of this effort has focused
on various moth species because they display some of the most striking ex-
amples of long-distance chemical communication in the animal kingdom1 and
many are important pests. It had been suspected for several centuries that
olfaction mediated sex attraction in moths,2 but conclusive experimentation
demonstrating its occurrence was first performed by Mayer (1900). Between
1961 and 1975, however, long-range sex pheromones were chemically identified
in 61 moth species within 10 families (Mayer and McLaughlin 1975). In
another 80 species, including 5 additional families, certain chemicals were
found to elicit sex attraction. During the past 6 years, the entomological
literature on this subject has accumulated at a near exponential rate. These
recent reports substantiate that pair-forming communication in most moth
species is facilitated by a chemical signal (pheromone) emitted by the female
and perceived by conspecific males over great distances. Despite this vast
literature, little attention has been directed toward understanding the evolu-
tionary factors that maintain this typical system of pair-forming. Neglect of
such factors is unfortunate because an evolutionary approach could provide
the hypotheses needed to predict events in biological systems. This paper
presents an attempt to elucidate these factors by asking 2 basic questions:
1) Why does the sexual dichotomy in pair-forming, female signalers and
male searchers, exist in moths? 2) Why does the female-emitted signal
possess certain characteristics?

*Michael D. Greenfield is a postdoctoral fellow employed through a cooperative agreement
between the University of Florida and the U. S. Department of Agriculture, Insect Attractants,
Behavior, and Basic Biology Research Laboratory. He recently completed one year of field
work in Panama with the Smithsonian Tropical Research Institute. His research interests are
the evolution of communication in Lepidoptera and Orthoptera and the application of com-
munity ecology theory to reproductive isolation. Current address: Dept. of Entomology and
Nematology, Univ. of Florida, Gainesville, FL 32611 USA. Florida Agricultural Experiment
Station Journal Series No. 2770.
'Superscripts refer to notes in the appendix at the end of this paper.

March, 1981

Insect Behavioral Ecology-'80 Greenfield

Most prior explorations into the adaptive significance of pair-forming
systems dwelled on environmental or phylogenetic constraints on the sig-
naling mode (Wilson 1975, Alcock 1979). While such factors undoubtedly
affect adoption of one type of communicative channel over another, this in-
vestigative strategy can lead to inaccurate generalizations and diverts atten-
tion from more ultimate influences. Thus, the prevalence of olfaction in moth
communication was attributed, in part, to their nocturnal activity and lack
of hardened, sclerotized parts (Wilson 1975). Nocturnal activity presumably
predisposes an organism toward chemical communication because atmospheric
turbulence, common during daylight and potentially disruptive to odor trails,
would be infrequent. This explanation does not account for the widespread
occurrence of pheromonal signaling in diurnally active moths (Lymantriidae,
Saturniidae, Sesiidae, Tortricidae, Yponomeutidae). The lack of extensive
sclerotized parts in Lepidoptera was suggested to underlie the supposed ab-
sence of acoustic communication in this order, since stridulatory mechanisms
generally rely on such morphology. However, sound is produced with minimal
sclerotization in some moths,3 and in other lepidopterans heavily sclerotized
structures have evolved from which sound emanates.4
Only Alexander and Borgia (1979) and Thornhill (1979) have addressed
the topic of how natural selection affects pair-forming communication by
asking both basic questions above. Their emphasis on the first question,
sexual dichotomy between signalers and searchers, involves understanding
the basic "male-female phenomenon." In most organisms, females are ex-
pected to perform that role (signaling/searching) in pair-forming that is
less costly in terms of risk and energy expenditure. This situation arises be-
cause of the parsimonious relationship in which females usually direct more
energy into "parental effort" (production of large gametes, nourishment of
embryos, parental care), but less into "mating effort" (courtship, intra-
sexual competition) (Low 1978). These same economic considerations can be
applied to the second question to probe why a particular signaling mode (i.e.,
olfactory, acoustic, visual, tactile, electrical, surface wave) is employed by
an organism.


To expose those factors that potentially maintain the typical pair-forming
system in moths, this paper will initially examine species that do not conform
to the typical scheme. The greater and lesser wax moths (Galleria mellonella
(L.), Achroia grisella (F.), respectively; Pyralidae: Galleriinae), inhabi-
tants of honeybee colonies, are among the better-known non-conformists. In
G. mellonella males emit copious amounts of a 2-component sex pheromone
(n-nonanal + n-undecanal) attractive to females over long distances (Roller
et al. 1968, Leyrer and Monroe 1973, Finn and Payne 1977). The function (s)
of this pheromone appears to differ considerably from the male "aphrodisiac
pheromones" common in moths (Birch 1974) and which only affect female
behavior during the final stages of courtship. Mating in G. mellonella is
known to occur outside beehives, but field observations and laboratory etho-
logical studies of courtship are rudimentary (Nielsen and Brister 1977,
R. Nielsen, personal communication).
Pair-forming in A. grisella follows the general system found in G.
mellonella. In the laboratory, male A. grisella have been observed to remain

Florida Entomologist 64 (1)

March, 1981

stationary in the upper parts of plexiglass cages, fanning their wings nearly
continuously during the scotophase (Kunike 1930, M. Greenfield, unpublished
data). During this stationary wing-fanning large quantities of a 2-com-
ponent sex pheromone (n-undecanal + cis-11-n-octadecenal) are emitted
(Dahm et al. 1971, M. Greenfield, unpublished data). As in G. mellonella, this
pheromone is attractive to females over long distances (Dahm et al. 1971),
but the tentatively attractive effects of sound and airflow cannot be ignored.
Preliminary laboratory observations showed that signaling (wing-fanning)
males tend to aggregate in small clusters and that agonistic behavior within
these clusters (attempts to dislodge neighbors from their signaling positions
by running directly at them) is common (M. Greenfield, unpublished data).
Eldana saccharina (Pyralidae: Galleriinae), found on papyrus in South
Africa, also exhibits male pheromonal signaling via stationary wing fanning
(P. Atkinson, personal communication).
Another group of pyralid moths, the sloth moths (Cryptoses choloepi
Dyar, Bradypodicola hahneli Spuler; Chrysauginae), possess equally aber-
rant pair-forming systems. Cryptoses choloepi larvae are coprophagous,
found only on sloth dung in Central and South America. Adults, primarily
males, are frequently found resting on the bodies of sloths. It is believed that
adult males locate sloths and remain there to intercept and mate with ar-
riving females (Waage and Montgomery 1976; J. K. Waage, personal com-
munication). Mating probably occurs on the body of the sloth, and when a
sloth defecates mated female moths leave the host to oviposit on its dung. No
characteristic signaling behavior has been observed in either sex of C.
choloepi (J. K. Waage, personal communication).
Bradypodicola hahneli exhibits similar pair-forming behavior, but its
larvae live on the sloth body, possibly feeding on oily skin secretions. Male
and female adults are found on sloths in approximately equal numbers.
Upon arriving at a sloth, these moths shed their wings. Mating occurs on the
sloth's body, and the duration of copulation is very long. As in C. choloepi,
no signaling behavior characteristic of moths was observed in either sex
(J. K. Waage, personal communication).
Other moths that may belong to this "non-conventional category," but for
which little information is available, are known. Males of the noctuid
Heliothis paradoxus (Grote) were observed forming small groups in the
field (J. E. Lloyd, D. Mays, and T. J. Walker, unpublished data). The ag-
gregating males produce a continuous snapping sound by striking together
costal swellings on their forewings (Hebard 1922; J. E. Lloyd, D. Mays, and
T. J. Walker, unpublished data). In the hepialid Hepialus humuli very dense,
populous aggregations of males, to which females may be attracted, have
been witnessed (Carolsfeld-Kraus6 1959). Males of other Hepialus species
produce an odor, detectable to observers, that is believed to attract females
(Barrett 1882, Robson 1887). Aggregations of males also occur in Adela bella
Chambers (Incurvariidae: Adelinae) (Heppner 1974). Gillmer (1922) de-
scribed attraction of females to males in Biston hirtarius Cl. (Geometridae).
Finally, instances of reversed sexual dimorphism (simple antennae in males;
pectinate antennae in females, supposedly adapted for detection of chemical
signals) are reported for several species in the genus Trilochana (Aegeriidae
= Sesiidae) (LeCerf 1920), suggesting an atypical pair-forming system in
these moths.

Insect Behavioral Ecology-'80 Greenfield

By comparing certain ecological parameters of moths possessing different
pair-forming systems, an attempt will be made to understand the adaptive
significance of the typical system (female signalers, male searchers). A
problem with this comparative technique is that information on species with
atypical systems is very scanty, and the few available examples may cer-
tainly bias interpretations. Undoubtedly, species other than the sloth moths
and wax moths exist in which females do not emit pheromones attractive to
males over long distances. Nevertheless, a strong correlation can be de-
tected between resource structure (spatial distribution of larval food re-
source) and the system of pair-forming.
In sloth moths, at one end of the spectrum (larval food resources are very
aggregated), males wait at the resource and scramble to mate with arriving
females. Males do not signal, the resource (sloth) probably attracting both
sexes to the courtship arena. Larval food resources of waxmoths (beehives)
are also very aggregated, but a risk factor exists. Bees are known to attack
adult G. mellonella (Nielsen and Brister 1977), rendering mating on the
resource very unsafe. The same danger and mating outside of beehives may
occur in A. grisella. Aggregations of signaling A. grisella males observed in
laboratory cages could reflect the formation of lekss" (Alexander 1975) in
the vicinity of beehives in field situations. Investigation of mating behavior
in other moths known to associate with social Hymenoptera5 may help
clarify the relationships between resource structure, risk, and pair-forming.
In moths exhibiting the typical pair-forming system, larval food re-
sources (i.e., host plants for phytophagous species) generally appear to be
dispersed regularly. Even species whose host plants are rare may encounter
a regularly dispersed resource distribution relative to the insect's normal
range of movement. Such resource structure may present a situation in
which searching is more costly (energy requirements, risk of predation) than
signaling. Alternatively, extremely aggregated resource structures, existing
as veritable point sources of dense resource scattered throughout the en-
vironment, should greatly lower the cost of searching in pair-forming. Fe-
males must locate larval food resources for oviposition and any additional
movement involved in mate seeking would usually be minimal. Waiting or
signaling males would be expected on or near most of these point sources,
respectively. Whether mating occurs on or near the point source may depend
principally on differential levels of danger between the 2 situations (Spieth
1973). For most moth species, however, the resource structure does not
present the insects with a finite number of point sources, and searching for
mates would represent a considerable cost in addition to locating larval food
resources for oviposition. Therefore, females assume the signaling role.
Oviposition behavior may also influence the sexual dichotomy of pair-
forming. In species that lay their eggs in clusters, as opposed to those that
deposit one egg per fruit, leaf, stem, etc., energy expenditure during oviposi-
tion is reduced. On the other hand, deposition of single eggs per resource
unit necessitates considerable movement and searching by the female. Search-
ing for males in such species would not considerably increase the mating
effort of a female, regardless of the resource structure.
Temporal stability of habitat can similarly affect pair-forming. If the
resource structure is stable from year-to-year, a female can often oviposit

Florida Entomologist 64 (1)

close to her emergence site. Alternatively, if habitat quality is continually
changing, an adult female may often emerge in a site lacking larval food
resources and must then engage in extensive movement in search of them. As
with single egg depositors, such females would not greatly increase their
mating effort through mate seeking.
Until now, I have considered the sexual dichotomy of pair-forming to be
directed only by the female. However, a situation of sexual conflict often
arises in nature (Parker 1979). The controversial cases of rape in various
animals (Thornhill 1980) may be examples of this. Essentially, sexual con-
flict arises when the interests of the male and female, related to maximiza-
tion of individual fitness, do not concur. The conflict can be of a much subtler
nature than rape. Regarding pair-forming behavior, such conflict may mani-
fest itself by the male's interests also providing input to the sexual
While females are expected to minimize risk and energy expenditure dur-
ing mating, males generally attempt to maximize the number of females
with which they copulate (Williams 1966, Trivers 1972). In certain circum-
stances a male might maximize this number by searching instead of sig-
naling. Organisms with a very short breeding season may present such a
situation. With only several days available for mating, a male's reproductive
success could be severely limited by awaiting attracted females. Data from
anuran amphibians support the above suggestion. In species that are ex-
plosive breeders (mating activity occurring only during several days an-
nually at temporary ponds) males generally do not vocalize, as most frogs
and toads, but obtain matings through intense searching (Wells 1977). Moth
life histories offer similar circumstances. The reproductive period in most
species is very compressed, adult life often lasting only 7 to 10 days in the
field. It should be noted that if the constraint imposed by a short breeding
period is operating, positive feedback into the moth pair-forming system
would result. This is because the interests of both sexes appear to coincide.
In all of the previous arguments it was implicit that male lepidopterans
provide a negligible amount of parental effort. However, Boggs and Gilbert
(1979) recently used radioactive labeling techniques to show that male
Heliconius butterflies (Nymphalidae: Heliconiinae) transferred a protein-
aceous substance to the female during copulation. Protein transfer during
mating was also found in Plodia interpunctella (Hiibner) (Pyralidae: Phy-
citinae) (M. Greenfield, unpublished data). These substances, passed via the
spermatophore, were discovered within unfertilized eggs dissected from the
females' ovaries after mating. A female's fecundity might be enhanced by
this male contribution which would then constitute a considerable degree of
male parental effort. If a male's material contribution is high enough, fe-
males would be expected to increase their mating effort, expending more
energy and assuming greater risks during pair-forming (Alexander and
Borgia 1979, Thornhill 1979). This possibility should not be overlooked in
moths, particularly among those species in which females typically mate
more than once.

Given the basic division between female signalers and male searchers,
what factors influence the characteristics of the female signal? Female

March, 1981

Insect Behavioral Ecology-'80 Greenfield

sexual signals in moths can be typified as chemical, restricted to a specific
group of compounds, and emitted at very low intensity (release rate of com-
pound). Economic factors related to risk and expenditure of energy in sig-
naling females should be of primary importance in the evolution of these 3
The chemical mode of signaling in female moths may principally be an
adaptation to avoid predators. That sexual signals are risky is supported by
evidence of predators attracted to various visual and acoustic signals of their
prey.6 However, only one instance of predators attracted to their prey's sex
pheromone is known.7 Why are predators (arthropod or vertebrate) ap-
parently not lured by these chemical signals? The answer may be found in
the relatively polyphagous habits of many predatory species. While sex
pheromones of moths and other insects are not as species specific as was orig-
inally believed (Takahashi 1973), closely related species often use struc-
turally different chemicals. Therefore, predators possessing receptors cap-
able of detecting a particular sex pheromone may be unable to support their
polyphagous existence through reliance on chemoreception. Additionally,
evolution of multiple chemoreceptors or a generalized chemoreceptory system
for detecting a wide range of chemical compounds could be impeded by cer-
tain morphological constraints. Perception of a broad spectrum of acoustic
or visual signals, however, may be a much simpler engineering problem for
an organism. Among moths, all known sex pheromones are aliphatic (straight
chain) compounds, but they differ interspecifically according to their func-
tional groups, length of carbon chain, and positions and orientations of
double bonds or asymmetric carbon atoms (Mayer and McLaughlin 1975,
Tamaki 1977). Detection of both an alcohol and aldehyde, for example, might
be vastly more difficult for a predator than evolving the machinery for locat-
ing cricket species singing at different sound frequencies or chirping rates.
Parasitoids must also be considered in weighing the risks of sexual sig-
naling. Unlike predators, most species are not polyphagous, possibly due to
being tightly interlocked with their host's biology (Vinson 1975). Both
acoustic and chemical signals are known to attract parasitoids to their
emitters,8 but moths appear to be immune from such attacks. All cases of
chemotaxis in parasitoids attacking moth species involve stimulation by
various larval, food, or fecal odors. Most probably, it would not be ad-
vantageous for a parasitoid to attack an adult female moth, a short-lived
organism soon to die. Unless a signaling female deposits eggs in her immedi-
ate vicinity, attraction to pheromone-emitting moths would reduce a para-
sitoid's fitness to zero.
Other possible advantages of chemical signaling include its effectiveness
at very low intensity and its low cost of production. Release rates of most
moth pheromones are probably 3-100 ng/h,9 yet they are detected by search-
ing males. The ability to communicate effectively in the chemical mode at
low signal intensity may increase predator/parasitoid avoidance, although no
evidence indicates that potential enemies cannot perceive moth pheromones
because of their low release rates. Many examples of extremely sensitive
host-locating mechanisms occur in parasitoids (Vinson 1976), and it appears
likely that chemotaxis to moth pheromones would have evolved were it
profitable. Similarly, pheromone production may be energetically cheaper

Florida Entomologist 64 (1)

March, 1981

than acoustic or visual bioluminescentt) signaling, but a comparison of this
sort is difficult to assess accurately.
Knowledge that all known sex pheromones of female moths are aliphatic
compounds, containing 10- to 18-carbon atom chains, invites further evolu-
tionary questions. Such compounds could be emitted due to biosynthetic
constraints and/or signaling efficiency. Because the biosynthesis of sex
pheromones is essentially unknown, this topic-more related to the evolu-
tionary origin rather than the maintenance of chemical communication-will
not be discussed here.10 Regarding signaling efficiency, Wilson and Bossert
(1963) pointed out that compounds within the molecular weight range of
moth pheromones possess the dual attributes of volatility and potential for
uniqueness. Heavier molecules would not diffuse easily whereas molecules
confined to a lower molecular weight range could not create a diversity of
structurally unique compounds high enough to allow for pheromonal spe-
cificity among the thousands of coexisting moth species. It must be recog-
nized, however, that this account was made prior to our present understand-
ing of the extreme chemical specificity in moth sex pheromones. Many of
these compounds are specific geometric or positional isomers, and most species
utilize combinations of various isomers in specific ratios as their pheromones
(Tamaki 1977). Therefore, sufficient diversity for reproductive isolation
probably could be achieved with compounds of lower molecular weight.
Lighter molecules, however, may be too volatile for maintenance of an aerial
odor trail. Although many reports assert that the specific chemical struc-
tures of moth pheromones evolved to enhance reproductive isolation (Roelofs
and Card4 1974), few studies rigorously test this claim."

Very low release rates (signal intensity) of sex pheromones are ubiqui-
tous among moths,9 a fact well known to anyone who has attempted quanti-
tative gas chromatography of these compounds. As previously discussed, this
may have evolved to enhance evasion from enemies or to minimize expendi-
ture of energy, but another possible explanation is the facilitation of sexual
selection. Essentially, if searching ability is heritable to any extent,12 fe-
males could select mates who are "better" searchers by emitting a very weak
signal. This modulation of signal intensity may represent a compromise be-
tween a release rate sufficiently high to be detected by males, but low enough
so that "inferior" searchers seldom reach the emitting female. Thus, sig-
naling females can stochastically implement mate choice13 analogous to the
manner in which searching females are believed to assess male signaling
power in many species. Lloyd (1979) referred to this (potential) ability of
signaling females as "active filtering."
Age dependent changes in the pheromonal release rates of virgin moths
indirectly support the role of low signal intensity in sexual selection. Female
signal intensity generally increases for several days immediately following
adult emergence, and in several cases older females were shown to be more
attractive to males.14 As senescence approaches, virgin female moths may
simply become less selective, and mating with an "inferior" male would be
preferable to not mating at all (Lloyd 1979). Despite the low signal in-
tensities in newly emerged females, their fecundity in some species was
found to be maximal if mating occurs at this time.15 Consequently, a dilemma

Insect Behavioral Ecology-'80 Greenfield 11

exists in that females are less attractive at an age when mating would pro-
duce maximal fecundity. This apparent incongruity could be partially ac-
counted for by the sexual selection hypothesis outlined above.


Superficially, searching male moths appear to be engaging in an activity
fundamentally different from a singing male cricket. However, the energy
expenditure and risk accrued in mate seeking and in stridulation may be
comparable. Additionally, intrasexual competition certainly exists in both
activities. Calling crickets attempt to outsignal one another (Alexander 1975)
and male moths participate in a race to locate females. Female choice can
likewise operate in both situations. This article suggests that ecological'
factors merely affect the specific activities (signaling/searching) of the sexes
in pair-forming, while retaining the fundamental economic balance. In
retrospect, the article could have been appropriately entitled "Pheromone
emission in female moths: why don't male moths sing?"


ISower et al. (1971) calculated the mean range for pheromonal communi-
cation in the cabbage looper [Trichoplusia ni (Hiibner), Noctuidae] as 100 m.
Earlier measures of communication distances for the gypsy moth, (Lyman-
tria dispar L., Lymantriidae), and various saturniid moths were several
kilometers (Collins and Potts 1932, Rau and Rau 1929), but these studies
greatly overestimated the distances by not considering the movement of male
moths prior to their perception of female pheromone. Kaissling and Priesner
(1970) showed that an olfactory sense cell in male silkworm moths [Bombyx
mori (L.), Bombycidae] responded neurophysiologically to the impingement
of a single sex pheromone molecule.
2Lepidopterists in 18th century England used caged female moths to at-
tract conspecific males (Card6 1976).
3The death's head sphinx moth (Acherontia atropos (L.), Sphingidae)
produces a piping sound by moving air through its pharynx (Busnel and
Dumortier 1959). It is believed that this sound allows the moth to gain
entrance to beehives where it feeds on honey, since it closely resembles the
piping of the queen bee.
4Many arctiid and ctenuchid moths emanate ultrasonic clicks with thoracic
microtymbals to deter insectivorous bats (Blest et al. 1963, Fenton and
Roeder 1974). Hamadryas butterflies (Nymphalidae) produce very loud
cracking sounds during flight, particularly during agonistic behavior (Sil-
berglied 1977).
5Several moths in the families Agrotidae, Cosmoptergyidae, Cyclotornidae,
Pyralidae, and Tineidae are reported to be symbionts of social Hymenoptera
and Isoptera (Wilson 1971).
GVarious acoustic Orthoptera, Homoptera, and anuran amphibians possi-
bly attract predators with their sexual signals (Walker 1964, Rentz 1975,
Bell 1979, Bonaccorso 1979, Tuttle 1979). 'Similarly, fireflies (Lampyridae)
may invite predators with their bioluminescence (Lloyd 1973).
7Bark beetles (Scolytidae) attract predatory clerid beetles and dolicho-
podid flies with their sex pheromones (Vit6 and Williamson 1970, Williamson
1971). One exceedingly bizarre case of exploitation of a moth's pheromonal
system by a predator exists. Bolas spiders (Mastophora, Araneidae) ap-
parently lure male Spodoptera moths (Noctuidae) with a chemical that is
either identical to or has effects similar to the female Spodoptera pheromone

Florida Entomologist 64 (1)

March, 1981

(Eberhard 1977). Note that in this situation the male receiver, not the fe-
male transmitter, is the victim.
sparasitoid flies are attracted to acoustic sexual signals of several
Orthoptera and Homoptera (Cade 1975, Soper et al. 1976, Mangold 1978).
Mosquitoes were reported to orient to the calling songs of frogs on which
they fed (McKeever 1977). Sex pheromones of coccids and pentatomids at-
tract parasitoid Hymenoptera and Diptera, respectively (Mitchell and Mau
1971, Sternlicht 1973).
OActual pheromonal emission rates in female moths have been quantified
in only 3 species; Grapholitha molesta (Busck) (Tortricidae)-3.9 ng/h
(Baker et al. 1980), Plodia interpunctella (Pyralidae)-approximately 9
ng/h (Sower and Fish 1975), Trichoplusia ni (Noctuidae) approximately
1000 ng/h (Bjostad et al. 1980). However, based on the quantities of
pheromone extractable from female moths (Sower et al. 1973), most species
probably release at rates 3-100 ng/h.
I oKittredge and Takahashi (1972) proposed that sex pheromone communi-
cation in the Arthropoda evolved from searching males undergoing selection
pressure to detect incidental odors (molting fluids, excretory products) of
females. In time, selection acting on females created the specific sex
pheromone chemicals presently observed. The relationship between larval diet
and the constitution of moth sex pheromones is unclear. Hendry et al. (1975)
claimed that shifts in larval diet altered the chemical composition of sex
pheromones in oak leaf rollers (Archips semiferanus Walker, Tortricidae),
but this relationship was later disproved (Miller et al. 1976).
"Many sympatrically occurring moth species utilize slightly different sex
pheromones (Roelofs and Carde 1974), but these interspecific differences
could be due to random effects rather than selection to maintain a unique
communicatory channel. Greenfield and Karandinos (1979) used a resource
partitioning model to show that pheromonal differences, particularly those
involving "inhibition," among sesiid moths may have resulted from selection
to reduce interspecific interference.
12A controversy exists as to whether individuals ever select their mates
based on genetic (as opposed to material) criteria (Borgia 1979). If females
always choose mates of particular genotypes, other genotypes would rapidly
be depleted from the population, greatly decreasing genetic variance at cer-
tain loci. Without such variance, mate selection based on genetic differences
could not operate. However, the genetics of characters involved in mate selec-
tion (i.e., searching ability) are certainly complex, and the above scenario
may be unrealistic.
13Much of the close-range courtship behavior in moths may also have
evolved in the context of sexual selection (Baker and Card6 1979). Mate
choice can operate at various stages of the pair-forming sequence.
14The following references report an increase in pheromonal titer and/or
percentage of females emitting pheromone during the several days following
emergence (Shorey et al. 1968, Sower et al. 1971, Lawrence and Bartell
1972, Sanders and Lucuik 1972, Calvert and Corbet 1973, Nordlund and
Brady 1974, Marks 1976, Miller and Roelofs 1977, Swier et al. 1976, 1977).
In most of these studies, males showed greater responsiveness to older fe-
males or to chemical extracts of older females.
-OFemales in Ephestia cautella (Walker) (Pyralidae) (Barrer 1976),
Plodia interpunctella (Pyralidae) (M. Greenfield, unpublished data), and
Sitotroga cerealella (Olivier) (Gelechiidae) (Ayertey 1975) exhibit higher
total fecundity if mating occurs on the day of emergence rather than several
days later. The pheromonal release rate and frequency of release in P.
interpunctella females is lower during the initial 2 days following emergence
than on subsequent days.

Insect Behavioral Ecology-'80 Greenfield


I thank Ted Burk, Stuart Krasnoff, and members of the Insect Behavioral
Ecology Seminar at the University of Florida for providing stimulating dis-
cussion leading to the ideas expressed in this paper. Peter Atkinson (South
Africa Sugar Cane Association), Ross Nielsen (USDA, Baton Rouge, LA),
Jeffrey K. Waage (Imperial College, Silwood Park, Great Britain), and
Thomas J. Walker (University of Florida) provided unpublished data. My
findings on Achroia grisella and Plodia interpunctella reported here result
from research supported by a postdoctoral fellowship cooperatively admin-
istered by the University of Florida and the U. S. Department of Agricul-
ture, Insect Attractants, Behavior, and Basic Biology Research Laboratory.
Comments from James A. Coffelt (USDA, Gainesville, FL), James E. Lloyd
(University of Florida), and John R. McLaughlin (USDA, Gainesville, FL)
and typing by Elaine S. Turner greatly aided the manuscript's final prepara-

ALCOCK, J. 1979. Animal Behavior: An evolutionary approach, 2nd ed.
Sinauer Assoc. Inc., Sunderland, Massachusetts. 532 p.
ALEXANDER, R. D. 1975. Natural selection and specialized chorusing behavior
in acoustical insects. Pages 35-77 In D. Pimentel, Ed. Insects, science,
and society. Academic Press, New York.
AND G. BORGIA. 1979. On the origin and basis of the male-female
phenomenon. Pages 417-440 In M. S. Blum and N. A. Blum, Eds. Sex-
ual selection and reproductive competition in insects. Academic Press,
New York.
AYERTEY, J. N. 1975. Egg laying by unmated females of Sitotroga cerealella
(Lepidoptera: Gelechiidae). J. Stored Prod. Res. 11: 211-15.
BAKER, T. C., AND R. T. CARDf. 1979. Courtship behavior of the Oriental
fruit moth (Grapholitha molesta): experimental analysis and con-
sideration of the role of sexual selection in the evolution of courtship
pheromones in the Lepidoptera. Ann. Ent. Soc. America 72: 173-88.
- AND J. R. MILLER. 1980. Oriental fruit moth pheromone
component emission rates measured after collection by glass-surface
adsorption. J. Chem. Ecol. 6: 749-58.
BARRER, P. M. 1976. The influence of delayed mating on the reproduction of
Ephestia cautella (Walker) (Lepidoptera: Phycitidae). J. Stored
Prod. Res. 12: 165-9.
BARRETT, C. G. 1882. Odour emitted by the male of Hepialus hectus. Ent.
Month. Mag. 19: 90-1.
BELL, P. D. 1979. Acoustic attraction of herons by crickets. J. New York
Ent. Soc. 87: 126-7.
BIRCH, M. C. 1974. Aphrodisiac pheromones in insects. Pages 115-34 In M. C.
Birch, Ed. Pheromones. American Elsevier, New York.
BJOSTAD, L. B., L. K. GASTON, AND H. H., SHOREY. 1980. Temporal pattern
of sex pheromone release by female Trichoplusia ni. J. Insect Physiol.
26: 493-8.
BLEST, A. D., J. S. COLLETT, AND J. D. PYE. 1963. The generation of ultra-
sonic signals by a new world arctiid moth. Proc. Roy. Soc. Lond. (B).
158. 196-207.
BOGGs, C. L., AND L. E. GILBERT. 1979. Male contribution to egg production
in butterflies: First evidence for transfer of nutrients at mating.
Science 206: 83-4.

Florida Entomologist 64(1)

March, 1981

BONACCORso, F. J. 1979. Foraging and reproductive ecology in a Pana-
manian bat community. Bull. Florida State Mus. Biol. Ser. 24: 359-
BORGIA, G. 1979. Sexual selection and the evolution of mating systems. Pages
19-80 In M. S. Blum and N. A. Blum, Eds. Sexual selection and repro-
ductive competition in insects. Academic Press, New York.
BUSNEL, R. -G., AND B. DUMORTIER. 1959. Verification par des methods d'
analyse acoustique des hypotheses sur l'origine du cri du sphinx
Achenontia atropos (Linn6). Bull. Soc. Ent. France 64: 44-58.
BUTENANDT, A., R. BECKMANN, AND E. HECKER. 1961a. tber den sexuallock-
stoff des seidenspinners, I. Der biologische test und die isolierung des
reinen sexuallockstoffes bombykol. Hoppe-Seyler's Z. Physiol. Chem.
324: 71-83.
-- ---- AND D. STAMM. 1961b. Uber den sexuallockstoff des seiden-
spinners, II. Konstitution und configuration des bombykols. Hoppe-
Seyler's Z. Physiol. Chem. 324: 84-8.
CADE, W. 1975. Acoustically orienting parasitoids: fly phonotaxis to cricket
song. Science 190: 1312-3.
CALVERT, I., AND S. A. CORBET. 1973. Reproductive maturation and phero-
mone release in the flour moth Anagasta kuehniella (Zeller). J. Ent.
(A) 47: 201-9.
CARDE, R. T. 1976. Utilization of pheromones in the population management
of moth pests. Environ. Heal. Persp. 14: 133-44.
CAROLSFELD-KRAUSE, A. G. 1959. The mating of Hepialidae (Lep.). Ent. Rec.
J. Var. 71: 33-4.
COLLINS, C. W., AND S. F. POTTS. 1932. Attractants for the flying gypsy
moths as an aid in locating new infestations. United States Dept.
Agric. Tech. Bull. 336: 1-43.
The olfactory and auditory mediated sex attraction in Achroia grisella
(Fabr.). Naturwissenschaften 58: 265-6.
EBERHARD, W. G. 1977. Aggressive chemical mimicry by a bolas spider. Sci-
ence 198: 1173-5.
FENTON, M. B., AND K. D. ROEDER. 1974. The microtymbals of some Arcti-
idae. J. Lepid. Soc. 28: 205-11.
FINN, W. E., AND T. L. PAYNE. 1977. Attraction of greater wax moth fe-
males to male-produced pheromones. Southwestern Ent. 2: 62-5.
GILLMER, M. 1922. Mitteilunger z. entwicklungsgeschichte des Biston
hirtarius Cl. Int. Ent. Zeitschr. 16: 34-6.
GREENFIELD, M. D., AND M. G. KARANDINOS. 1979. Resource partitioning of
the sex communication channel in clearing moths (Lepidoptera:
Sesiidae) of Wisconsin. Ecol. Monogr. 49: 403-26.
HEBARD, M. 1922. The stridulation of a North American noctuid, Heliocheilus
paradoxus Grote (Lep.). Ent. News 33: 244.
M. E. ANDERSON. 1975. Evidence for origin of insect sex pheromones
-presence in food plants. Science 188: 59-63.
HEPPNER, J. B. 1974. Habitat: Adela bella in Florida (Incurvariidae;
Adelinae). J. Res. Lepid. 13: 67-8. '
KAISSLING, K. -E., AND E. PRIESNER. 1970. Die riechschwelle des seiden-
spinners. Naturwissenschaften 57: 23-8.
KITTREDGE, J. S., AND F. T. TAKAHASHI. 1972. The evolution of sex
pheromone communication in the Arthropoda. J. Theor. Biol. 35:
KUNIKE, G. 1930. Zur biologie der kleinen wachsmotte, Achroea grisella
Fabr. Z. Angew. Ent. 16: 304-56.

Insect Behavioral Ecology-'80 Greenfield

LAWRENCE, L. A., AND R. J. BARTELL. 1972. The effect of age on calling
behaviour of virgin females of Epiphyas postvittana (Lepidoptera)
and on their pheromone content and ovarian development. Ent. exp.
appl. 15: 455-64.
LE CERF, F. 1920. Probable heteromorphism of secondary sexual characters
in Trilochana. Proc. Ent. Soc. London: 84-5.
LEGROS, G. V. 1971. Fabr6 poet of science. Horizon Press, New York. 352 p.
LEYRER, R. L., AND R. E. MONROE. 1973. Isolation and identification of the
scent of the moth, Galleria mellonella, and a revaluation of its sex
pheromone. J. Insect Physiol. 19: 2267-71.
LLOYD, J. E. 1973. Firefly parasites and predators. Coleop. Bull. 27: 91-106.
-- 1979. Sexual selection in luminescent beetles. Pages 293-342 In M. S.
Blum and N. A. Blum, Eds. Sexual selection and reproductive com-
petition in insects. Academic Press, New York.
Low, B. S. 1978. Environmental uncertainty and the parental strategies of
marsupials and placentals. American Nat. 112: 197-213.
MANGOLD, J. R. 1978. Attraction of Euphasiopteryx ochracea, Corethrella
spp., and gryllids to broadcast songs of the southern mole cricket. Fla.
Ent. 61: 57-61.
MARKS, R. J. 1976. Mating behavior and fecundity of the red bollworm
Diparopsis castanea Hmps. (Lepidoptera, Noctuidae). Bull. Ent. Res.
66: 145-58.
MAYER, A. G. 1900. On the mating instinct in moths. Psyche 9: 15-20.
MAYER, M. S., AND J. R. Mc LAUGHLIN. 1975. An annotated compendium of
insect sex pheromones. Florida Agric. Exp. Sta. Monogr. Ser. 6, 83 p.
Mc KEEVER, S. 1977. Observations of Corethrella feeding on tree frogs
(Hyla). Mosq. News. 37: 522-3.
MILLER, J. R., T. C. BAKER, R. T. CARDf, AND W. L. ROELOFS. 1976. Rein-
vestigation of oak leaf roller sex pheromone components and the hy-
pothesis that they vary with diet. Science'192: 140-3.
AND W. L. ROELOFS. 1977. Sex pheromone titer correlated with
pheromone gland development and age in the redbanded leafroller
moth, Argyrotaenia velutinana. Ann. Ent. Soc. America 70: 136-9.
MITCHELL, W. C., AND R. F. L. MAU. 1971. Response of the female southern
green stink bug and its parasite, Trichopoda pennipes, to male stink
bug pheromones. J. Econ. Ent. 64: 856-9.
NIELSEN, R. A., AND D. BRISTER. 1977. The greater wax moth: adult be-
havior. Ann. Ent. Soc. America 70: 101-3.
NORDLUND, D. A., AND U. E. BRADY. 1974. Factors affecting release rate and
production of sex pheromone by female Plodia interpunctella (Hiib-
ner) (Lepidoptera: Pyralidae). Environ. Ent. 3: 797-802.
PARKER, G. A. 1979. Sexual selection and sexual conflict. Pages 123-66 In
M. S. and N. A. Blum, Eds. Sexual selection and reproductive competi-
tion in insects. Academic Press, New York.
RAU, P., AND N. RAU. 1929. The sex attraction and rhythmic periodicity in
the giant saturniid moths. Trans. Acad. Sci. St. Louis 26: 83-221.
RENTZ, D. C. 1975. Two new katydids of the genus Melanonotus from Costa
Rica with comments on their life history strategies (Tettigoniidae:
Pseudophyllinae). Ent. News 86: 129-40.
ROBSON, J. E. 1887. On the flight and pairing of Hepialus hectus and humuli.
Ent. Month. Mag. 23: 186-7.
ROELOFS, W. L., AND R. T. CARDI. 1974. Sex pheromones in the reproductive
isolation of lepidopterous species. Pages 96-114 In M. Birch, Ed.
Pheromones. American Elsevier, New York.
1968. Sex pheromones of pyralid moths. I. Isolation and identification

Florida Entomologist 64 (1)

of the sex-attractant of Galleria mellonella L. (greater waxmoth).
Acta Ent. Bohemoslovaca 65: 208-11.
SANDERS, C. J., AND G. S. LUCUIK. 1972. Factors affecting calling by female
eastern spruce budworm, Choristoneura fumiferana (Lepidoptera:
Tortricidae). Canadian Ent. 104: 1751-62.
SHOREY, H. H., S. U. MCFARLAND, AND L. K. GASTON. 1968. Sex pheromones
of noctuid moths. XIII. Changes in pheromone quantity, as related to
reproductive age and mating history, in females of seven species of
Noctuidae (Lepidoptera). Ann. Ent. Soc. America 61: 372-6.
SILBERGLIED, R. E. 1977. Communication in the Lepidoptera. Pages 362-402
In T. A. Sebeok, Ed. How animals communicate. Indiana Univ. Press,
SOPER, R. S., G. E. SHEWELL, AND D. TYRELL. 1976. Colcondamyia auditrix
Nov. spp. (Diptera: Sarcophagidae), a parasite which is attracted by
the mating song of its host, Okanagana rimosa (Homoptera: Cicad-
idae). Canadian Ent. 108: 61-8.
SOWER, L. L., J. A. COFFELT, AND K. W. VICK. 1973. Sex pheromone: a
simple method of obtaining relatively pure material from females of
five species of moths. J. Econ. Ent. 66: 1220-2.
-- AND J. C. FISH. 1975. Rate of release of the sex pheromone of the
female Indian meal moth. Environ. Ent. 4: 168-9.
-- L. K. GASTON, AND H. H. SHOREY. 1971. Sex pheromones of noctuid
moths. XXVI. Female release rate, male response threshold, and com-
munication distance of Trichoplusia ni. Ann. Ent. Soc. America 64:
SPIETH, H. T. 1973. Mating behavior and evolution of the Hawaiian Droso-
phila. Pages 94-101 In M. J. D. White, Ed. Genetic mechanisms of
speciation in insects. D. Reidel, Boston.
STERNLICHT, M. 1973. Parasitic wasps attracted by the sex pheromone of
their coccid host. Entomophaga 18: 339-42.
SWIER, S. R., R. W. RINGS, AND G. J. MUSICK. 1976. Reproductive behavior
of the black cutworm, Agrotis ipsilon. Ann. Ent. Soc. America 69:
AND 1977. Age-related calling behavior of the black
cutworm, Agrotis ipsilon. Ann. Ent. Soc. America 70: 919-24.
TAKAHASHI, F. 1973. Sex pheromones; are they really species specific? Mem.
Coll. Agric. Kyoto Univ. (Jap.). 104: 13-21.
TAMAKI, Y. 1977. Complexity, diversity, and specificity of behavior-modify-
ing chemicals in Lepidoptera and Diptera. Pages 253-286 In H. H.
Shorey and J. J. McKelvey, Jr., Eds. Chemical control of insect be-
havior. Wiley, New York.
THORNHILL, R. 1979. Male and female sexual selection and the evolution of
mating strategies in insects. Pages 81-121 In M. S. Blum and N. A.
Blum, Eds. Sexual selection and reproductive competition in insects.
Academic Press, New York.
1980. Rape and sexual selection in Panorpa scorpionflies and a gen-
eral rape theory. Anim. Behav. 28: 52-9.
TRIVERS, R. L. 1972. Parental investment and sexual selection. Pages 136-79
In B. Campbell, Ed. Sexual selection and the descent of man, 1871-
1971. Aldine, Chicago.
TUTTLE, M. D. 1979. A frog-eating bat and its probable impact upon calling
frogs. (Abstract) Tenth Annu. North American Symp. on Bat Res.
Washington Univ., St. Louis, Missouri.
VINSON, S. B. 1975. Biochemical coevolution between parasitoids and their
hosts. Pages 14-48 In P. W. Price, Ed. Evolutionary strategies of
parasitic insects and mites. Plenum, New York.

March, 1981

Insect Behavioral Ecology-'80 Greenfield 17

1976. Host selection by insect parasitoids. Annu. Rev. Ent. 21:
VITE, J. P., AND D. L. WILLIAMSON. 1970. Thanasimus dubius: prey per-
ception. J. Insect. Physiol. 16: 233-9.
WAAGE, J. K., AND G. G. MONTGOMERY. 1976. Cryptoses choloepi: a copro-
phagous moth that lives on a sloth. Science 193: 157-8.
WALKER, T. J. 1964. Experimental demonstration of a cat locating or-
thopteran prey by the prey's calling song. Fla. Ent. 47: 163-5.
WELLS, K. D. 1977. The social behavior of anuran amphibians. Anim. Behav.
25: 666-93.
WILLIAMS, G. C. 1966. Adaptation and natural selection. Princeton Univ.
Press, Princeton, NJ. 307 p.
WILLIAMSON, D. L. 1971. Olfactory discernment of prey by Medetera
bistriata (Diptera: Dolichopodidae). Ann. Ent. Soc. America 64:
WILSON, E. 0. 1971. The insect societies. Harvard Univ. Press, Cambridge,
Massachusetts. 548 p.
-- 1975. Sociobiology. Harvard Univ. Press, Cambridge, Massachusetts.
697 p.
--- AND W. H. BOSSERT. 1963. Chemical communication among animals.
Rec. Prog. Horm. Res. 19: 673-716.

Florida Entomologist 64 (1)



The current revolution in ecological theory is based on the preeminence of
individual reproductive success. Testing such theory requires that individuals
be recognized and monitored as they play out their reproductive lives. Insects
have the diversity, abundance, and accessibility that make them especially
attractive as research subjects, but getting to know individuals is often dif-
ficult. Their small size, short lives, and great mobility complicate the task
of the field observer. It is not surprising that lizards, ground squirrels, and
hyenas are better known as individuals under field conditions than are their
insect counterparts.
Southwood (1966, 1978) extensively reviewed the literature of insect
marking methods, including those suited to identifying individuals. We will
build on his contribution by listing and discussing techniques that are broadly
applicable and by outlining the principles important in developing a marking
method or in deciding which existing method to adopt.
Marking systems that permit marking 50 or more insects for individual
recognition fall into three principal categories: (1) Mutilation-changing
the insect itself, (2) Labeling-attaching a label to the insect, (3) Direct
marking-using the insect as a blank label.


The grosser forms of mutilation-such as amputation of all or parts of
appendages or wings-should be avoided because they are likely to disrupt
or alter behavior. Gangwere et al. (1964) found pronotal notching to be
satisfactory for marking saltatorial Orthoptera and cockroaches and pointed
out that the notches, unlike conventional marks, remained readable through
one or more molts.
The hard, thick, often smooth cuticle of beetles provides a suitable sub-
strate for scratching or burning coded spots or numbers. For example,
Murdoch (1963) used a piece of safety razor blade to mark the elytra of
carabids, and J. C. Schuster (pers. comm., 1974) used an insect pin to en-
grave numbers into passalid pronota.
The chief advantage to mutilation techniques is permanence; the chief
disadvantage is damage to the insect or infection.

*T. J. Walker is a professor in the Department of Entomology and Nematology, University
of Florida. He teaches a graduate course in Insect Ecology in which one of the laboratory
periods is devoted to marking techniques and how to improve and evaluate them. His research
interests include butterfly migration and mating strategies in crickets, two fields that have
benefited greatly from marking techniques that permit individual recognition.
S. A. Wineriter is an Assistant in Entomology and scientific illustrator in the Department
of Entomology and Nematology at the University of Florida. She earned her M.S. in Biology
at Ball State University and has since studied art and scientific illustration at the University
of South Carolina and the University of Florida.
Current address: Dept. of Entomology and Nematology, University of Florida, Gainesville,
FL 32611. Florida Agricultural Experiment Station Journal Series No. 2809.

March, 1981

Insect Behavioral Ecology-'80 Walker & Wineriter

In this procedure a prepared label, often printed, is attached to the
insect. In some cases the writing may be small enough and the label large
enough to allow the insect to carry instructions for the finder and a return
address. In an entomological equivalent of bird banding, Urquhart and
Urquhart (e.g. 1978, 1979) labeled Monarch butterflies to learn their migra-
tion routes. The alar tags they used were 9x13 mm self-adhesive labels
folded over the de-scaled costal margin of the forewing (Fig. 1A). The tags
remained in place as the Monarchs traveled 1000's of kilometers on their way
to overwintering sites. Few insects can carry a label as heavy as those used
by the Urquharts. Roer (1957, 1969) devised a much lighter butterfly label,
composed of two 6 mm discs glued together through a hole in the forewing
(Fig. 1B). The upper disc was aluminum foil that glinted in the sun making
it easier to spot marked individuals.
If a label has only a number on it, it can be very small and light-weight.
German apiculturists developed plastic "sign-platelets" (Zeichenplittehen)
to be glued to the pronota of queen bees making them easy to see and
identify'. The 2 mm-diameter platelets, numbered 00-99, come in 5 colors and
weigh only 1.3 mg each. They are suited to labeling a variety of insects (Fig.
1C) and are legible nearly a meter away. Gary (1971) added a disc of shim
steel to each numbered plastic disc (total weight 11 mg) and retrieved the
labels from returning honeybees with powerful magnets mounted at the hive


This is by far the largest and most varied set of techniques for marking
for individual recognition. The variety is only partly a result of different


dorsal ventrol
A-URQUHART'S self-adhesive 1
"alar tag" S
actual size

Sa C-Glue-on numbered disc
6 mm .aluminum foil "Zeichenplattchen"
./. ,i-I| 2 mm 0 hole

B-ROER'S glue-through Mu
GQJE "Etiketten" Bu 3
I^ actual size
( 6mm 0 airmail paper actu size

Fig. 1. Labeling techniques. Weights of labels are (A) 11-14 mg (includ-
ing self-adhesive), (B) 1.2 mg (excluding glue), (C) 1.3 mg (excluding

Florida Entomologist 64 (1)

constraints imposed by different insect sizes, surfaces, habits, and habitats,
and by different requirements of the researcher. Much of the variety results
from scientists settling on the first technique they try that works fairly well
for them. They are, after all, not concerned with developing optimal marking
techniques but with learning something about "their" insects. If buying a
can of paint at the variety store and applying it as coded dots with the head
of an insect pin works, then no more time need be wasted. Southwood (1978)
reviews a large number of ad hoc solutions to direct marking problems. We
will develop below a framework for evaluating and improving direct mark-
ing methods.

Marking Materials

The variety of substances available to the researcher as potential mark-
ing materials continually increases. Generally, one or more of the first few
materials a researcher tries will work, so the problem is not in finding a
workable material but in failing to find an equally available material that is
A perfect marking material would combine these properties (and all
prospective marking materials should be evaluated in these respects) :
Durable. The material must resist wear and abrasion for the duration of
the study.
Adhesive. The material must not flake or chip from the insect. Some
materials that are durable are prone to chip-for example, butyrate dope
sold for painting model cars. If the entire mark2 flakes off, the insect reverts
to unmarked status; if part of a mark flakes off, the insect may be made the
identical twin of another marked insect (Fig. 4C).
Non-toxic. Neither the material nor its solvent, in the amounts applied,
should kill or permanently alter the behavior of the insect3.
Easy to apply. Ease of application is a function of the applicator (see
below) as well as the material and becomes more important as the number
of insects to be marked and the number of persons doing the marking in-
Quick-drying. Insects may need to be held after marking until the ma-
terial can no longer smear or act as an adhesive for organic debris. There-
fore, slow-drying materials are undesirable.
Light-weight. Some insects are so small (e.g. mosquitoes) that the weight
of the mark may prove an important criterion.
Available in several easy-to-distinguish colors4. In many instances more
than one color must be used in order to produce enough unique marks.
Invisible except to researcher. Although desirable so that predators will
be no more or less likely to take marked insects than unmarked ones, this
criterion is seldom met. Using pigments that can be seen only in the dark
under UV illumination is generally impractical for studies of diurnal insects.
For cave insects, or for nocturnal insects that stay in the dark during the
day, pigment invisibility is of no consequence. Only for nocturnal insects that
remain exposed during daylight is the use of UV-fluorescent materials worth
Materials that have been successfully used to mark insects for individual
recognition are paints (e.g. artist's oils, enamels, tempera, and those in Fig.
2-Tech-Pen "ink," acrylics, Liquid Paper), lacquers (e.g. nailpolish, bu-

March, 1981

Insect Behavioral Ecology-'80 Walker & Wineriter

Fig. 2. Some especially useful marking materials (left to right) Tech-Pen
"ink," developed for marking laboratory glassware, is a thick paint that
comes in 11 colors; acrylic paint with day-light fluorescent pigments7
(acrylics can be thinned with water but are waterproof when dry); Liquid
Paper correction fluid comes in 9 colors8; Pentel pens deliver a fine line of
opaque-white, oil-base ink that drys quickly; Sharpie marking pens have
"fine" or "extra fine" points that deliver quick-drying, water-resistant inks
(8 colors available) o; technical pens'1 filled with india ink produce uniform
lines or dots that can be applied to transparent wings or blots of Liquid
Paper (Fig. 3H).

tyrate dope, nitrocellulose lacquers), inks (e.g. stamp pad ink, india ink,
permanent inks in marking pens as in Fig. 2), copper wirel3. Most of these
materials have been tried by 50 or more graduate students on 10 or more
kinds of insects. Although all will serve at least to some extent on some in-
sects, only a few mark a variety of insects well. We took the four materials
with the highest student ratings and tested them on three difficult-to-mark
species. Tech-Pen ink and Liquid Paper won the competition (Table 1). We
are presently testing a much wider variety of materials in a similar fashion
(Wineriter and Walker, in preparation).

Application Methods
Techniques for applying a mark maybe as important as the choice of
marking material.
Holding the insect firmly yet without injury during marking is often
difficult. When fingers fail, devices employing netting or suction14 can be used
(Fig. 3A-D). Anesthetizing the insect should be avoided if at all possible
because of potential effects on the physiology or behavior of the insect.
Chilling the insect-for example, by placing the insect in a vial in an ice-
filled vacuum jug-is apparently safer than using carbon dioxide3 or ether.

22 Florida Entomologist 64(1) March, 1981

(SMALL SIZE), PASSALID BETTLE, Odontotaenius disjunctus (HARD,

T. castaneum 0. disjunctus P. americana
Marking material* first** mediant first median first median

Tech-Penink 3 6 7 13 5 8
Liquid Paper (could not apply) 3 4 6 12
Pentel pen 1 1 (would not adhere) 2 3
Hyplar acrylic 1 1 1 1 1 1
(lemon yellow)

*Application methods were as follows: Liquid Paper, in-bottle brush or bristle of brush;
Tech-Pen ink and acrylic, bristle attached to swab stick (for T. castaneum) and splintered
end of swab stick (for others); Pentel pen, point or portion of the point splayed to one side.
**Average no. of weeks before loss of first mark (4 replications of 5 individuals each).
fNo. of weeks before loss of median mark by those individuals still living. For example, a
value of 5 means that not until the sixth week did more than half of the surviving individuals,
in the four replications combined, lose their mark. In no species were fewer than 15 individuals
alive when the median mark was lost.
$M. Grumbacker, Inc., New York, N.Y. 10001.
An important criterion in the choice of an applicator for a marking ma-
terial is often the minimum size of spot it can make consistently. Therefore
we have arranged the following list of examples from fine to coarse: single
bristle, minutin pin, headless insect pin, technical pen (Fig. 2, 3H), shaft of
pin (Fig. 3E), head of fine pin, grass stem, pine needle, syringe (Fig. 3F),
paper clip (Fig. 3G), swab stick, marking pen (Fig. 2), rubber stamp (Fig.
With many marking materials, proper viscosity is important but difficult
to maintain. If the material is too thin, it spreads out-of-control on the insect,
interfering with sensory areas and increasing exposure to often-toxic sol-
vents; if the material is too viscous, it adheres poorly to the insect and makes
a high-profile spot or line, increasing the likelihood that the material will
flake off. A paint pot, such as the one designed by W. D. Hamilton (Fig. 3E),

Fig. 3. A-D Holding techniques. A-B. Devices using netting (mark is
applied through the mesh of the net). C-D. Devices using suction. E-G. Mark-
ing techniques's. E. Paint pot keeps paint from drying out and assures a
standard amount of paint with each withdrawal of the pin (W. D. Hamilton,
pers. comm., 1980). F. A small quantity of Tech-Pen ink6 can be squeezed
directly into the barrel of disposable plastic syringe with needle removed, the
plunger reinserted forcing the paint to the needle-end of the syringe, and the
needle replaced. So long as the ink is not allowed to dry and clog the needle,
uniform droplets can be produced by gentle pressure on the plunger. Inserting
the needle into a cork retards the drying12. G. Paper clip. H. Technical pen"1
used to write on spot of Liquid Papers. I. Rubber stamp used to mark wing
of cloudless sulphur butterfly with return address.

Insect Behavioral Ecology-'80 Walker & Wineriter 23

A ring with

nylon net

S-1 diophram

(Conwoy et al 1974)

- brass boll, drilled in
center to take pin

- aluminum rod hollowed
and threaded

F-tuberculin syringe

- cork to receive
needle while
not in use

Slightly concove
r tipholdsdrop
of point


Florida Entomologist 64 (1)

or a syringe (Fig. 3F) can be used to maintain optimal viscosity during a
marking period. (Hamilton's device also solves another problem-that of
getting a uniform amount of material on the applicator.)

Coding Systems
In addition to selecting a marking material and a way to apply it, the
researcher desiring to uniquely mark a large number of insects must decide
upon a code.
These are four important standards a code should meet:
Enough unique marks2. A system adequate for identifying 1000 indi-
viduals must be more complex than one that permits identifying only 100
individuals. The coding system selected should be as simple as the study
permits yet easily expandable (say by adding another color) if the study
need be increased in scope.
Marks easy to apply. Application is facilitated by keeping the marks
simple and by using only one color per individual. Two dots are generally
easier to make than are one triangle or one star (for an example of overly
complex marks, see White 1970). When more than one color is used in mark-
ing an individual, the process is seriously complicated by having to switch
Easy to read and verbalize. If the mark can be seen all at once and
quickly translated into a few words (or a figure15), note-taking and record
keeping are simplified (Fig. 4D).
Failsafe from misidentification. The most serious problem that can arise
in a study requiring recognition of individuals is that one marked individual
is mistaken for another. The most likely way for this to occur is through
partial loss of a mark (Fig. 4C). If every mark has the same number of
components, partial loss can generally be detected and misidentification
avoided. For example, 364 ("three sixty-four") cannot become "thirty-six"
through loss of the 4 if "thirty-six" is coded 036. With systems using posi-
tioned dots, the same number of dots should be used for each individual (Fig.
4A,C) (Michener et al. 1955). A disadvantage of using complex symbols as
components of marks, in addition to the difficulty of application, is that
partial loss may transform the symbol: a star becomes a circle if it loses its
points, and an 8 becomes a 3 if it loses its left side.
For large insects, coding systems present few problems. An arabic num-
ber (of constant number of digits) or even a return address (Fig. 31) can
be inscribed on a wing, elytron, or pronotum. If a less conspicuous mark is
desired, enough easily identifiable positions are available to make positioned
dots easily translated into arabic numbers. For example, the system in Fig.
4A permits marking 100 individuals per color by positioning one dot on each
forewing. However, 20 identifiable positions are required. For insects having
fewer positions for symbols, some other coding system must be used". The
use of symbols more complicated than' dots or varying the numbers of
symbols per mark violates important principles explained above. The two
remaining ways of expanding a code of positioned spots are (1) using more
colors and (2) allowing more than one color per mark'7 (Table 2). The
former is limited by the number of colors that remain easily identifiable under
field conditions5; the latter makes marking a more complicated process.
Both are subject to error from the fact that pigments are apt to differ in

March, 1981

Insect Behavioral Ecology-'80 Walker & Wineriter

00 0

C0 I

.., ..... ij

R ( CF, R B C'" -R '34
W 20 2 .





R-6 R-12:20, R-2,6

Fig. 4. Positioned dots. A-B. Effect of number of positions. A. Using two
dots (one on each wing) 100 individuals can be marked distinctively for each
color used. B. Using two dots per individual only 15 can be marked for each
color used. C. Need for a constant number of dots per mark. Without this
safeguard, loss of a dot can cause the researcher to misidentify the indi-
vidual. (Individual "R-two-six" can become indistinguishable from "R-two"
or "R-six.") D. Different systems for naming marked individuals. "R-CF,RB"
(short for "red dot at center front and at right back") is unnecessarily
cumbersome; "R-thirty-four" requires adding the values of the positions
(e.g. Sheppard et al. 1969) and is of no benefit if a constant number of dots
per individual is used; "R-twelve-twenty" depends on imagining the
pronotum as the face of a clock with the first dot starting clockwise at "12"
being the hour hand and the second dot being the minute hand (adopted from
Jackson 1933) ; "R-two-six" is the easiest, most direct way to name the indi-
vidual bearing the pronotum pictured16.

their properties-especially in their likelihood of fading or flaking. For
example, consider the effects on the results of a study when a color used
early in a study adheres well, while another, used late, does not.

Hasty adoption of a less-than-optimal marking technique may seriously
limit what can be learned from a study or even compromise its results. We
have tried to identify some serious pitfalls and to give tips on how to avoid

26 Florida Entomologist 64 (1) March, 1981


No. of unique marks
No. of colors 1 color/indiv. 1 or 2 colors/indiv.
used* (A) (B)

1 15 -
2 30 60
4 60 240
8 120 960

*In many instances no more than 4 or 5 colors remain easily distinguishable under field


Persons who have generously shared their ideas relative to marking
insects are too numerous for us to list here; furthermore, the ideas they
shared were generally offered not as original but as useful methods they had
learned in conversation or research with their colleagues, students, or pro-
fessors18. Our greatest debt is to 240 graduate students who took Insect
Ecology at University of Florida 1966-1980. Their enthusiasm and verve for
marking arthropods of nearly every description seemed too valuable to
seclude in the pages of a bulging loose-leaf compendium labeled "Insect-
Ecology Marking Manual."
For special help with various aspects of the manuscript we thank J. E.
Lloyd, R. C. Littell, Jane Brockmann, Frank Robinson, and Susan Jungreis.
Finally, we must thank T. R. E. Southwood, who published his extensive
survey of marking methods (1966) and then updated it (1978). We, as
others, have benefited greatly.

'Zeichenplittehen are not available in the United States but can be ob-
tained in Germany from apicultural suppliers-e.g. Hamman Bienenzucht-
gerite, Postfach 225, 6633 Hassloch. Smith (1972) described a technique
developed by N. E. Gary for making numbered discs manually. Fresneau and
Charpin (1977) described how to make similar labels photographically.
2Throughout this article we use mark to denote all marking material used
to make one insect identifiable: a mark consists of one or more dots, figures,
or characters.
3The possibility of subtle, chronic effects of chemicals is amply demon-
strated by effects of CO2 anesthetization (Edwards and Patton 1965).
4Colors that are easy to distinguish in large amounts in good light may
be difficult or impossible to tell apart on the insect under field conditions.
Some daylight fluorescent colors5 become indistinguishable when fluorescing
under UV.
"Fluorescence is the emission of light of some hue (i.e. a particular wave-
length) upon absorption of radiation of shorter wavelengths. Ultraviolet
(UV) radiation will cause numerous substances to fluoresce (e.g. calcium
fluoride, zinc silicate) and these can be applied in amounts that are difficult
to see in daylight but easy to detect under UV.

Insect Behavioral Ecology-'80 Walker & Wineriter

Daylight fluorescent materials (Day-Glo Color Corp., 4732 St. Clair
Ave., Cleveland, Ohio 44103) appear unusually bright in daylight by reflect-
ing some wavelengths of visible light and transforming some of the absorbed
wavelengths (via fluorescence) into emitted light of the same hue as that
being reflected. These pigments, widely used in group-marking, can be added
to various paint bases to produce colors that are conspicuous in daylight and
brilliant under UV. When dilute, they are inconspicuous in daylight and
bright under UV.
6Mark-Tex Corp., 161 Coolidge Ave., Englewood, N.J. 07631.
7The brand illustrated has been discontinued. Day-light fluorescent pig-
ments5 can be mixed with acrylic base to produce extra-bright colors.
8Liquid Paper Corp., 9130 Markville Drive, Dallas, Texas 75243.
9Pentel of America, 1100 Arthur Ave., Elk Grove Village, Ill. 60007.
10Sanford Corp., 2740 W. Washington Blvd., Bellwood, Ill. 60104.
1Pen shown is a size 00 Rapidograph, Koh-I-Noor Rapidograph Inc.,
100 North St., Bloomsbury, N.J. 08804.
"Another technique for keeping a marking syringe operable is to squeeze
a drop onto the point after each marking. The drop dries but can be brushed
off to open the needle prior to marking the next individual (J. E. Lloyd, pers.
comm., 1980).
13Dave Synder (pers. comm., technique developed in Insect Ecology lab,
spring 1975) and Mirenda and Vinson (1979) independently discovered that
if small pieces of fine copper wire from a lamp cord are tied around various
portions of an ant's anatomy, the ant is identifiable and neither it nor its
nestmates can remove the mark-as they generally do for more conventional
marking materials.
14Hand-held battery operated vacuum cleaners are available in increasing
variety, making practical the use of suction devices in the field.
"1A distinction must be made between oral and written note taking. It is
easy to write 768, but if one is talking into a tape recorder, the same mark
becomes 'seven hundred sixty-eight" (or the shorter and more ambiguous
"seven, six, eight").
16Yet another way to read the six positions illustrated is to picture them
as three columns of two to be read from the cricket's left to right. If the
anterior positions are 1's and the posterior positions are 2's, the marked
cricket becomes "R-0,1,2." A cricket marked red in positions CF and CB
would be read "R-0,12,2" (or "R-0,3,2") (J. E. Lloyd, pers. comm., 1980).
"The formula for calculating the number of unique marks (N) for a
coding system using a fixed number of positions (n) and a fixed number of
dots (of one color) per mark (k) is
(n),kC k!(n-k)
For example with 6 positions and 2 dots per mark, N(6),2 = 15; with 6 posi-
tions and 3 dots per mark, N(6,,3 = 20.
If the number of dots per mark is allowed to vary, the number of unique
marks is the sum of the N's for each number of dots permitted. For example,
with 6 positions and 1, 2, or 3 dots (of one color) per mark,
N(6),1,2,3 = N(6 + N(6),2 + N(6),3 = 6 + 15 + 20 = 41.
If more than one color is used, but only one color is applied to each in-
dividual, the total number of unique marks for the coding system becomes
where C is the number of colors available. For examples, see Column (A) of
Table 2.
If more than one color is used and one or more is applied to each indi-
vidual, the total number of unique marks for the coding system becomes

Florida Entomologist 64 (1)

March, 1981

For examples, see column (B) of Table 2.
IsWe realize that we cannot properly acknowledge the originators of most
of the marking methods we describe. We not only are unable to back-track an
idea from our informer to its source but we are also aware that good mark-
ing techniques may have more than one original source-they may have been
independently invented two or more times (for example, copper bands for
ants13). Nonetheless we wish to acknowledge our sources for these ideas:
paper clips as applicators (Fig. 3G) (F. A. Lawson, pers. comm., 1980),
Tech-Pen ink as a marking agent and tuberculin syringe as an applicator
(Fig. 2 and 3F) (W. A. Banks, pers. comm., 1972) (see also Freeman 1964),
pen-and-Liquid-Paper technique (Fig. 3H) (Alan Bolton, pers. comm., 1977,
who learned the technique from Donald Windsor), rubber stamping ad-
dresses on butterfly wings (Fig. 31) (Richard Mankin and Basilios
Mazomenos, pers. comm., 1977) (see also Neilsen 1961). Those who know
earlier sources for these ideas, especially if published, should send us the
information if they wish that more proper credit be given in the future.

CONWAY, G. R., M. TRPIS, AND G. A. H. MCCLELLAND. 1974. Population
parameters of the mosquito Aedes aegypti (L.) estimated by mark-
release-recapture in a suburban habitat in Tanzania. J. Anim. Ecol.
43: 289-304.
EDWARDS, L. J., AND R. L. PATTON. 1965. Effects of carbon dioxide anesthesia
on the house cricket, Acheta domesticus (Orthoptera: Gryllidae).
Ann. Ent. Soc. Am. 58: 828-32.
FREEMAN, B. E. 1964. A population study of Tipula species (Diptera,
Tipulidae). J. Anim. Ecol. 33: 129-40.
FRESNEAU, D., AND A. CHARPIN. 1977. Une solution photographique au
problem du marquage individual des petits insects. Ann. Soc. Ent.
Fr. (N.S.) 13: 423-6.
GANGWERE, S. K., W. CHAVIN, AND F. C. EVANS. 1964. Methods of marking
insects, with especial reference to Orthoptera (sens. lat.). Ann. Ent.
Soc. Am. 57: 662-9.
GARY, N. E. 1971. Magnetic retrieval of ferrous labels in a capture-recapture
system for honey bees and other insects. J. Econ. Ent. 64: 961-5.
HEWLETT, P. S. 1954. A micro-drop applicator and its use for the treatment
of certain small insects with liquid insecticide. Ann. Appl. Biol. 41:
45-64. P1. 5-6.
JACKSON, C. H. N. 1933. On a method of marking tsetse flies. J. Anim. Ecol.
2: 289-90.
A. WILLE. 1955. Additional techniques for studying the behavior of
wild bees. Insectes Soc. 2: 237-46.
MIRENDA, J. T., AND S. B. VINSON. 1979. A marking technique for adults of
the red imported fire ant (Hymenoptera: Formicidae). Fla. Ent. 62:
MORRIS, G. K. 1965. Vacuum cleaner restraining device. Turtox News 43:
MURDOCH, W. W. 1963. A method for marking Carabidae (Col.). Ent. Mon.
Mag. 99: 22-4.
NIELSEN, E. T. 1961. On the habits of the migratory butterfly Ascia monuste
L. Biol. Medd. Kgl. Dan. Vidensk. Selsk. 23(11) :1-81.
ROER, H. 1957. Aluminiumfolie im Dienste der Erforschung der Schmet-
terlingswanderfluge. Aluminium 33: 267.
ROER, H. 1969. Zur Biologie des Tagpfauenauges, Inachis io L. (Lep.,
Nymphalidae), unter besonderer Berucksichtigung der Wanderungen

Insect Behavioral Ecology-'80 Walker & Wineriter 29

im mitteleuropaischen Raum. Zool. Anz. 183: 177-94.
dynamics of an adult population of Aedes aegypti in relation to dengue
haemorrhagic fever in Bangkok. J. Anim. Ecol. 38: 661-702.
SMITH, M. V. 1972. Marking bees and queens. Bee World 53: 9-13.
SOUTHWOOD, T. R. E. 1966. Ecological methods, with particular reference to
the study of insect populations. Methuen, London. 391 p.
SOUTHWOOD, T. R. E. 1978. Ecological methods, with particular reference to
the study of insect populations, 2nd ed. John Wiley and Sons, New
York. 524 p.
URQUHART, F. A., AND N. R. URQUHART. 1978. Autumnal migration routes
of the eastern population of the monarch butterfly (Danaus p.
plexippus L.); overwintering site in the Neovolcanic Plateau of Mex-
ico. Can. J. Zool. 56: 1759-64.
URQUHART, F. A., AND N. R. URQUHART. 1979. Vernal migration of the mon-
arch butterfly (Danaus p. plexippus, Lepidoptera: Danaidae) in North
America from the overwintering site of the Neo-volcanic Plateau of
Mexico. Can. Ent. 111: 15-8.
WHITE, E. G. 1970. A self-checking coding technique for mark-recapture
studies. Bull. Ent. Res. 60: 303-7.

Florida Entomologist 64(1)

March, 1981



"It is curious that these little flies (Sepsidae) seem to need so many
aids to success in copulation-spotted wings and male display, aphrodisiac
odour by the females, special male clasping organs-when other flies
manage without any of these."
"A male of the Mediterranean fruit-fly, Ceratitis capitata. . The
elaborate patterns and vivid colour of this fly seem to go far beyond the
demands of mimicry or mutual recognition."
(Oldroyd 1964)


The study of animal communication has always been one of the corner-
stones of ethology. The pioneers of ethology, Lorenz, Tinbergen, and von
Frisch, demonstrated that many of the unusual actions of animals have
communicative functions, and that these behaviors have evolved in the same
manner as morphological characters (Tinbergen 1951, Lorenz 1958). The
behavioral ecology "revolution" that Lloyd (1980) described has given new
impetus to studies of animal communication. Behavioral ecologists have built
on classical ethological studies by looking in detail at the environmental and
social factors that shape and constrain animal communication systems. Some
behavioral ecologists have examined the ways in which animal signals are
adapted to aspects of the physical environment (Cullen 1957, Morton 1975,
Bennet-Clark 1970, Lall et al. 1980, Paul and Walker 1979). Another ap-
proach has been to look at the selection pressures on animal signals deriving
from the types of social interactions taking place between signalers and re-
ceivers (Dawkins and Krebs 1978). In this paper I will consider a particular
problem, that of why signaling is relatively simple in some species, but ex-
ceedingly complex in others. To get at this problem, I will look at signaling
and mating behavior in a particular group of insects, the acalyptrate flies.
The acalyptrates are a large and diverse group of fly families which, with
the bristly calyptrate flies such as the common housefly, constitute the "higher
flies" (Oldroyd 1964), in contrast to "lower flies" such as mosquitoes and
gnats. The best known acalyptrate flies are probably the fruit flies of the
families Drosophilidae and Tephritidae. We know more about the communi-
cation of these two families than the dozens of other acalyptrate families
combined. But in demonstrating the diversity of acalyptrate signals and the
evolutionary factors which shape them, I will give a variety of examples. I
have been studying the behavior of the tephritid Caribbean fruit fly,

*Theodore Burk is a postdoctoral associate employed through a cooperative agreement at
the Department of Entomology and Nematology, University of Florida, and the USDA Insect
Attractants, Behavior, and Basic Biology Laboratory, Gainesville, FL. He obtained his D. Phil.
in 1979 in the Animal Behaviour Research Group, Oxford University, where he was a Rhodes
Scholar from Kansas. His research interests are in insect social behavior, especially aggressive
behavior and acoustic communication.
Florida Agricultural Experiment Station Journal Series No. 2816. Reprint requests should
be directed to P.O. Box 14565, Gainesville, FL 32604.

Insect Behavioral Ecology-'80 Burk

Anastrepha suspense, so it will be the feature example. The "caribfly" ex-
cellently demonstrates many of the points I wish to make.

Flies probably produce a greater variety of signals than any other insect
order. Ewing (1977) has reviewed dipteran signals, so I will not attempt a
comprehensive listing. Instead, a few examples will be given which demon-
strate this variety. Then I will look at some ways that ecology, via the type
of mating system present, influences this signal variety.
Signals may be defined as particular morphological characters or be-
havioral actions which have been naturally selected because of the modifying
effects they have on the behavior of conspecific individuals (Dawkins and
Krebs 1978). This definition excludes such phenomena as mimicry or the
perception of host or prey cues by parasites and predators. Signals are prob-
ably best viewed as advertisements by individuals, which were selected be-
cause they induced specific effects. Acalyptrate flies use many different
channels in their signaling, as the following representative examples show.


The body of the caribfly is colored shiny gold; its clear wings are marked
with distinctive brown patterns. When alert to interactions with other
caribflies, it stands stiffly up on its legs; when facing another fly it holds its
wings perpendicular to its body, sometimes waving them slowly back and
forth, either synchronously or in alternating fashion (Dodson 1978). An
early German student of the tephritid fruit flies, of which the caribfly is one,
spoke of their "delicate markings" and "graceful behavior" (H. Loew 1862,
quoted in Zwolfer 1974a). The visual signals of acalyptrates as a group live
up to these descriptions.
In many families of acalyptrates, individuals have bright body colors,
usually glossy blacks or purples or metallic greens, golds, or browns. Pat-
terned wings are also widespread, as are distinct markings on other body
parts such as head and tarsi. In many cases these visual markings are dis-
played and emphasized by particular movements. Most other tephritids, like
the caribfly, wave their wings back and forth or up and down during court-
ship or aggressive encounters (Zwolfer 1974a). Males of the platystomatid
Euprosopia subula display their distinctive black fore tarsi by waving them
in the female's face during courtship (McAlpine 1973). Not only do such
movements emphasize morphological characters, but, since types of move-
ment and rates of presentation differ among sexes and species, they consti-
tute visual signals in themselves. Special postures in particular contexts also
display visual patterns and provide information about species, sex, size, or
other individual characteristics of the signaler. For example, the apple mag-
got fly, Rhagoletis pomonella adopts an "active posture" during feeding and
investigation, and a "pawing posture" during aggressive displays (Biggs
1972). Whole body movements and "dances" complete the list of ways that
visual signals can be produced. Hawaiian Drosophila males "joust" (Ringo
1976), courting pairs of the otitid Physiphora demandata perform a "spiral
dance" (Alcock and Pyle 1979) and advertising and courting tephritids spin,
circle, and weave about (Dodson 1978).

Florida Entomologist 64 (1)

March, 1981


Flies cannot match the Lepidoptera in the sophistication of their chemical
communication (Greenfield, this symposium), but pheromones are important
in several groups of acalyptrates. Chemical communication in flies has been
reviewed by Fletcher (1977) and Chambers (1977).
Sexually active male caribflies distend pouches at the end of their ab-
domen; at this time a clear droplet can be seen protruding from the anus.
Such males wave their wings, occasionally vibrating them rapidly (see next
section). Between bursts of wing fanning they often spin around, touching
the tip of their abdomen where the droplet is hanging to the substrate on
which they are resting (Nation 1972). What these males are doing is dis-
persing a pheromone which attracts receptive females (and also other males
-see below). Such long-distance attractants are found in several tephritids
(Chambers 1977, Pritchard 1967), in Hawaiian species of Drosophila (Spieth
1968), and in the otitid Physiphora demandata (Alcock and Pyle 1979).
With only one known exception, these long-distance attractants are produced
by males (Haniotakis 1974). Males of the fruit flies Dacus orientalis and D.
cucurbitae even release "smoke"-along with sex pheromone, sodium and
potassium phosphate particles are expelled from males, probably acting as
carriers of the pheromone molecules (Ohinata et al. 1981).
Contact, or short-range, pheromones are important signals in many
calyptrate flies (Ewing 1977, Fletcher 1977)-males are stimulated to mount
by the "taste" of females. Contact pheromones are also present in some
acalyptrates. In Drosophila they have been shown to be involved in recogni-
tion of species, mating status, and strain, including the "rare male effect",
which females prefer to mate with males of rare types in a population (re-
viewed by Ewing 1977).
Females of fruit-infesting tephritids of several genera mark host fruit
in which chemicals which deter other females from further egg-laying in the
same fruit (Prokopy et al. 1978), thus preventing competition between the
offspring of the marking female and the younger larvae of newcomers. Males
of some species "eavesdrop" on these signals, remaining in marked areas and
searching for the females whose proximity is thereby indicated (Katsoyannos

Although most people think of crickets and cicadas as the only "singing
insects," males of several groups of acalyptrate flies also produce compli-
cated, stereotyped acoustic signals as part of their sexual displays.
The best studied of these signals are those of Drosophila (Bennet-Clark
and Ewing 1970). Courting males within a few cm of a female extend one
wing and vibrate it, producing species-typical acoustic signals. Because sev-
eral different Drosophila species may be present at the same place and time,
these signals are important in species recognition. Other effects are to
stimulate sexual motivation in females, inhibit locomotion by females, and to
increase activity and sexual behavior in nearby males (von Schilcher 1976a
and b). Females of the drosophilids Drosophila and Zaprionus also produce
acoustic signals when rejecting or accepting males (Bennet-Clark et al.

Insect Behavioral Ecology-'80 Burk

1980). These female songs lack the regularity and species-specificity of male
In the last section, I mentioned that pheromone-calling male caribflies
produce bursts of rapid wing vibration. When males fight, they also vibrate
their wings rapidly, although more continuously. Also, after mounting fe-
males, both before copulation and when females become restless, males pro-
duce long sequences of rapid wing vibration (Webb et al. 1976). During these
rapid wing vibrations, characteristic acoustic signals are produced: calling,
aggressive, and pre-copulatory songs, each with a distinct pulse pattern.
Other tephritid species also produce species-specific sound signals (Monro
1953, Rolli 1976). The behavioral effects of tephritid songs are currently
being investigated, especially by a group led by Dr. D. L. Chambers and Dr.
J. C. Webb in the USDA Insect Attractants, Behavior and Basic Biology
Laboratory in Gainesville. In caribflies, the calling song increases female
activity (Burk unpublished) and, in conjunction with the pheromone, at-
tracts females (Webb 1973). The pre-copulatory song prolongs copulations
(Burk unpublished) and probably increases female sexual motivation.
An unusual method of acoustic signaling is found in the chloropid genus
Lipara (Mook and Bruggemann 1968, Chvala et al. 1974). Females oviposit
in reed stems and males must move about from stem to stem searching for
them. Males vibrate reed stems, different species vibrating in different pat-
terns. Receptive females reply with "good vibes" of their own (female signals
lacking species-specificity), and males move up the stem to mate. The system
works over distances up to 2 m. This system is a marvelous example of the
naturally selected fit between a species' environment and its signaling system.

Tactile signals are important in the courtship of many insects, and play
a role in species recognition in some calyptrate flies (Tobin and Stoffolano
1973). Tactile signals might also function in evaluations of size or other
qualities of interacting individuals. Various acalyptrates lick, kick, or probe
with their legs during aggression and courtship (Biggs 1972, Ewing 1977,
Wangberg 1978). Some of these actions probably help in the transmission of
visual or chemical signals, but tactile signals are probably also involved.
Examples in the caribfly include fighting males butting heads and males
mounted on females probing and licking with their proboscis (Dodson 1978,
Nation 1972). No detailed studies have been made of the effects of such
One set of structures that may constitute visual and tactile signals are
the long eyestalks of some acalyptrate flies. Such characters, with the eyes
located on the end of long stalks, are found in at least 8 families (McAlpine
1979). Usually the character is sexually dimorphic, males having more highly
developed eyestalks. McAlpine (1979) has shown that in several cases, males
butt heads during aggressive displays and utilize the eyestalks in the shoving
matches. Some interactions end without actual fighting, suggesting that the
length of eyestalks might be used as a visual cue about the size of opponents.
In last year's symposium, Thornhill (1980) gave examples of courtship
feeding of female insects by males. Some examples of this phenomenon can
be seen in the acalyptrate flies (see review in Dodson 1978). Males may pro-
vide salivary droplets to females during copulation (Wheeler 1924) or con-

34 Florida Entomologist 64 (1) March, 1981

struct mounds of foam on which females feed while males copulate (Stoltzfus
and Foote 1965, Pritchard 1967, Foote 1967, Novak and Foote 1975). These
nutritional offerings or mating inducements qualify as signals to the extent
that they act as chemical or visual attractants to females, especially in cases
where males build froth masses or blow bubbles in the absence of females
(Stoltzfus and Foote 1965, Novak and Foote 1975).

After this brief survey of acalyptrate signals, some generalizations can
be drawn. First, as has been pointed out for Diptera generally by Ewing
(1977), almost all signals are produced in a sexual context. In flies, it is only
during sexual encounters that the kinds of social interactions occur in which
signaling behavior would be beneficial. This is not surprising since, as Lloyd
(1979) has emphasized, sex is the most important thing there is for most
organisms. Second, most signals are male-produced. This is in line with the
idea that males are generally the more sexually aggressive sex, because of
the male-female dichotomy in parental roles (Trivers 1972, Thornhill 1980;
see Greenfield, this symposium). Third, there are vast differences in the
complexity of sexual signals exhibited by different species. Some hardly
signal at all while others, such as the caribfly, bombard the environment with
visual, chemical, acoustic, and other signals. This last observation is the one
I want to pursue in the remainder of this paper. To do so requires a slight
diversion-we must look at the mating systems of acalyptrate flies.

The links between patterns of parental investment, resource distribution
and availability, and form of mating system have been well demonstrated in
several recent papers (Trivers 1972, Emlen and Oring 1977, Alcock 1980).
Mating systems can be classified in a variety of ways. For our purpose, two
of the most useful are those of Alexander (1975) and Borgia (1979). The
first is based on what Parker (1978) calls the "encounter site convention"-
the encounter site being the place where males and females meet and mating
takes place. Alexander distinguishes between encounter sites in immediate
proximity to important resources needed by females for successful reproduc-
tion (food sites, refuges, oviposition sites, etc.) and encounter sites not so
located ("resource-based sites" vs "nonresource-based sites"). Borgia classi-
fies mating systems according to the interactions between males and females,
particularly which sex can best be said to control the decision whether or not
a mating will take place. A combination of these two approaches seems to
work best in classifying acalyptrate mating systems and in providing in-
sights into the effects of mating systems on signaling behavior.
In many acalyptrate flies, sexual encounter sites are resource-based. This
occurs when female-required resources are distributed in discrete, scattered,
predictable patches. In some such cases, males are able to control access to the
resource, only allowing access to females in return for copulation: such
matings might even be called rapes (Smith and Prokopy 1980). Sometimes
males establish individual territories on the resource. Males of fruit-infesting
and gall-forming tephritids establish exclusive territories on host plants, ex-
pelling other males and copulating with females who arrive to oviposit (Bush

Insect Behavioral Ecology-'80 Burk

1969, Zwolfer 1974b, Batra 1979, Prokopy 1980). Males of the neriid cactus
fly, Odontoloxozus longicornis, defend decaying patches on cacti, the female
oviposition site (Mangan 1979). Males of the dryomyzid Oedoparena glauca
defend areas of barnacle beds at low tide, seizing and copulating with females
who come to the barnacle beds to deposit their barnacle-consuming larvae
(Burger et al. 1980). In other cases, males search actively over resource
patches for females, not defending particular areas but still interacting ag-
gressively with other males. Searching replaces territoriality where the
number of competing males is very high or where resources are short-lived,
less discrete, or otherwise less economic to defend (see review by Davies
1978). Examples include the sepsid fly, Sepsis cynipsea, whose males search
for ovipositing females on the surface of fresh cow pats (a resource only
good for a few minutes) (Parker 1972), and another sepsid, Themira putris,
whose males form vast aggregations over sewage sludge, the larval habitat
(Oldroyd 1964) 1.
In some acalyptrates, sexual encounters take place on female-required
resources, but control of the mating decision is maintained by females. This
may be due to physical or behavioral adaptations which allow females to
fend off males or because resources are more available or evenly distributed,
thereby less monopolizable by males. Courtship and mating in most species
of Drosophila occurs on food sites, where males approach and court feeding
females (Spieth 1974). The searching behavior of male reed flies (Lipara),
who vibrate reed stems hoping for a reply from a female, has already been
described (Chvala et al. 1974). Even in some of the gall-infesting tephritids,
territory-establishing males do not monopolize all potential oviposition sites,
so that females are presented with a choice of ovipositing in sites with or
without resident males.
The third basic type of acalyptrate mating system is that in which en-
counter sites are not found in proximity with female-required resources.
Mating away from resources can be thought of as a system that males only
adopt when the resource-based alternative is too costly or otherwise un-
available to them. The reasons for this high cost or unavailability vary, and
in some cases are imperfectly understood. I will present some examples, try-
ing to account for the lack of resource-based mating in each case.
One type of non-resource based system, common in other groups of flies,
is the aerial swarm, so well discussed by R. Sullivan in this symposium (see
also Downes 1969). Aerial swarms are uncommon in acalyptrates, but a few
examples are reported from three families (McAlpine and Munroe 1968).
Best studied are several species of Lonchaeidae, which form swarms that in
all respects fit the patterns described for lower Diptera by Sullivan. The
retention of this mating system in these acalyptrates is probably related to
the usual low population densities and unpredictable distribution of larval
habitats (injured plant tissues), which would tend to lower the encounter
rates with females that could be expected by a male occupying a resource.
Unlike most other Drosophila, mating in Hawaiian Drosophila does not
take place on food sites. Instead, males form aggregations on vegetation,
where males joust with each other, perform sexual displays, and mate with
incoming females (Spieth 1968, Ringo 1976). By analogy with the mating
aggregations of vertebrates, these male mating aggregations have been
called lekss," a name which can be applied to all aggregations of displaying
1Superscripts refer to notes in the appendix.

Florida Entomologist 64(1)

males away from female-required resources. Lekking in Hawaiian Drosophila
may be due to predation pressure: the native insectivorous birds of Hawaii
may have taken a heavy toll of flies behaving conspicuously at exposed food
sites, necessitating mating away from such sites. Borgia (1979) has described
a similar case. Small, dung-ovipositing flies are forced to mate away from
dung pats because of the presence on dung pats of the larger, predatory dung
fly, Scatophaga stercoraria (which does have a resource-based mating sys-
The caribfly is one of a number of species of fruit-infesting tephritids
which often do not mate on host fruits, but on vegetation (sometimes, but
not always, in host trees) (Dodson 1978). Other, better-studied examples
include the Mediterranean fruit fly, Ceratitis capitata (the "medfly")
(Prokopy and Hendrichs 1979) and the Queensland fruit fly, Dacus tryoni
(Tychsen 1977). In these species, aggregations of males form on vegetation,
often with males establishing small territories consisting of a single leaf.
From these territories, males interact with each other, perform sexual dis-
plays, and court arriving females. Lekking in these tephritids is probably
related to biogeographical and ecological factors. Lekking species tend to be
tropical or sub-tropical; like many such groups, they tend to have wide host
ranges, unlike temperate species that tend to specialize on particular hosts.
Populations persist for much or all of the year, and host fruits are not
seasonally synchronized. The caribfly has been reported from over 100 fruits
(Swanson and Baranowski 1972); the medfly from over 200 (Christenson
and Foote 1960). Male resource monopolization in such polyphagic species
is unlikely to lead to high encounter rates with females (see Prokopy 1980) 2.

Behavioral ecologists look for relationships between ecological factors
and the behavior patterns of animals. A clear connection exists between
ecological circumstances and mating systems in the acalyptrates. This con-
necting thread also links up mating systems and signaling behavior.
In those species with resource-based, male-controlled mating systems,
sexual signaling is usually very simple. In some cases, males do not signal
at all but merely jump on and attempt to copulate with any object of ap-
proximately the correct size and shape (Berube 1978). In such mating sys-
tems, males who control resources are assured of a high encounter rate with
females. Therefore, male reproductive efforts are better directed toward adap-
tations for locating resources quickly, defending them against other males,
and seizing and retaining hold on encountered females (Parker 1970, 1974a).
Where signaling is present, as in the gall-forming tephritids, it usually con-
sists of simple repetitive actions involving a single channel, probably acting
mainly as a species-identifying cue. The repetitive wing-waving of many
tephritids falls into this category (Zwolfer 1974a).
When high encounter rates on resources are not so certain, or where
mating decisions are made by females, sexual signaling by males is more
highly developed. Female Drosophila must be courted and induced to mate;
male Drosophila have evolved courtship songs that not only identify the
species of the male but also have arousing effects on females. Male tephritids
in some species must advertise their resource-based territories with
pheromones or induce females to oviposit in their territories with nutritional

March, 1981

Insect Behavioral Ecology-'80 Burk

offerings (Stoltzfus and Foote 1965). Females may respond to such
pheromones or offerings because they allow rapid location of suitable ovi-
position sites or because the nutritional offerings help them to produce more
eggs: in effect, they trade a copulation for reproductive assistance by the
male. (An additional benefit may be reduced risk of predation while search-
ing for oviposition sites or food.)
The extremes in male signal complexity, represented by such species as
the caribfly, are reached when mating takes place in leks, away from re-
sources and with control of the mating decision firmly in female hands. Un-
like other Drosophila, Hawaiian Drosophila are large, have patterned wings,
produce elaborate sexual displays and aggressive behaviors, and disperse
long-distance pheromonal attractants (Spieth 1974). The lekking tephritids
retain bright colors and wing-waving, are more likely to produce long-
distance pheromones, and include all the species that are now known to
produce acoustic signals (Dodson 1978). The otitid Physiphora demandata,
another lekking species, produces attractant phermones and exhibits a
complicated spiral courtship dance (Alcock and Pyle 1979). It is worthwhile
to consider the interactions that account for this clear connection between
lek mating systems and conspicuous, elaborate sexual signals.
As Labeyrie (1971) and Zwolfer (1974b) have emphasized, mating on
the larval host plant often eliminates the need for attractant signals. In
lekking species mating away from host plants, males are selected to produce
conspicuous, high intensity attractants. Males who call in groups may be
able to attract females from greater distances, resulting in a tendency for
males to aggregate (Otte 1977). This probably accounts for the high response
of males to male-produced sex pheromones in such tephritids as the caribfly
(Perdomo et al. 1976-male responses are typically equal to those of virgin
females) : males must locate leks, either to establish their own calling stations
or to become "satellites" of calling males (Cade 1980).
The proximity of competing males in leks adds other selection pressures.
Males may fight to establish calling stations in favorable locations or to
keep other males at a distance-aggressive interactions are common in
groups of male caribflies (Dodson 1978). Parker (1974b) and Maynard Smith
and Parker (1976) have shown that selection often favors the substitution
of "war propaganda" signaling (Borgia 1979) in place of actual fighting,
and Barnard and Burk (1979) have shown that in such a competitive situa-
tion signals may become more complex due to contrasting selection pressures
on males to evolve ever more intimidating displays and ever less gullible
responses to the displays of other males.
The presence of females in leks adds more complexity to the situation. To
point signals at individual females, males may be selected to switch from
omni-directional calling signals to more restricted courtship displays (for
example to switch from long-distance pheromones to short-distance visual
displays). To reduce competition from neighboring males, males may perform
courtship displays less intensely than calling displays. Alexander (1975)
considers the variety of signaling interactions that may occur between males
in leks.
Female mate choice adds another selection pressure. Where females can
not obtain parental assistance or nutrition, they are expected to sexually
select males of high genetic quality (Trivers 1972), as indicated by pheno-

Florida Entomologist 64 (1)

March, 1981

typic vigor3. The signals of individual males may differ in ways which indi-
cate differences in male quality: females may observe the "war propaganda"
signals of males (Borgia 1979) ; mere establishment of a calling station may
indicate successful fights against other males (Burk 1979); large males may
produce more intense signals (Forrest 1980); or dominant males may be
more likely to signal or may produce more continuous signals (Burk 1979,
Crankshaw 1979). If such signal differences exist, females will be selected to
utilize them in mating decisions (Fisher 1958). However, such female choice
of male signals may set up another "arms race" (Dawkins and Krebs 1979),
similar to the one between fighting males. All males will be selected to
produce signals which induce females to mate, thereby mimicking the signals
of high-quality males. Females will be selected to become ever choosier and
coyer to avoid being deceived by low-quality males. In such an arms race,
signals may replace one another over evolutionary time, or signals may be-
come "layered" (Barnard and Burk 1979, West Eberhard 1979). This may
account for cases where courting males must perform a series of actions,
sometimes even involving different signal channels, before females will mate.
Thus, three types of selection pressure occur in leks which all tend to
increase the elaborateness and complexity of male sexual signals: the neces-
sity of attracting females, competition with neighboring males, and the de-
mands of female choice. These factors account for the wealth of signals in
such lekking acalyptrates as the caribfly.

Through the course of this paper, I have tried to show how ideas from
behavioral ecology help one to understand some aspects of insect communica-
tion. Such understanding can have practical benefits. (1) Some predators
and parasites intercept dipteran signals and use them to locate prey or hosts
-for example, the hymenopteran parasite Opius lectus locates its apple
maggot hosts by responding to the oviposition deterring pheromones of adult
female apple maggot flies (Prokopy and Webster 1978). (2) Sex attractants
can be used in the monitoring and control of pest populations. Field trials
using the male sex pheromones of the caribfly and medfly have shown them
to be at least as effective as traditional non-pheromonal attractants;
pheromones have the advantage that both sexes respond in large numbers
(Perdomo et al. 1976, Chambers 1977). The oviposition-deterrent pheromones
of fruit-infesting female tephritids have shown great promise as a means of
reducing attacks by ovipositing females (Katsoyannos and Boiler 1980). (3)
Because they are conspicuous and often stereotyped, and because control
techniques such as the release of sterilized males rely on their effective pro-
duction, sexual signals are particularly appropriate as indicators of the
quality of mass-reared insects (Huettel 1976, Sharp and Webb 1977).
In a wider sense, though, the behavioral ecology approach to the study of
insect communication is likely to be important to pragmatic entomologists
for the same reason that it is to ethologists. Thorough understanding of an
insect's behavior, whether from a theoretical or pragmatic standpoint, de-
pends on knowledge of the total biology of the species. The behavioral
ecology approach is the method best suited to give us such an understanding.

Insect Behavioral Ecology-'80 Burk

Acalyptrate flies produce a great variety of visual, chemical, acoustic,
tactile, and other signals. Most signals are produced by male flies as part of
sexual behavior, and species differ greatly in the degree of complexity of
these sexual signals. Patterns of parental behavior and resource structure
lead to three basic types of mating systems in acalyptrate flies: (1) Male-
controlled matings on female-required resources, (2) Female-controlled mat-
ings on female-utilized resources, and (3) Female-controlled matings in male
aggregations away from resources (lek mating system). Differences between
species in complexity of sexual signals are the result of these differences in
mating systems. Signals are simple in male-controlled systems, intermediate
in complexity in female-controlled resource-based systems, and most complex
in lek systems (because of the complexity of male-male and male-female
intractions in leks).

1In cases where mating decisions are controlled by males, it is often found
that females mate many times over their lifetime, as multiple visits are made
to resources (Foote 1967, Berube 1978). Where mating occurs away from
resources, multiple matings are rare (often only occurring when females run
low on stored sperm) (Cunningham et al. 1971). However, in species where
females are forced to copulate, they may still retain ultimate control, through
mechanical or physiological adaptations, over whether or not insemination
occurs following a particular copulation (Lloyd 1979, Walker 1980).
2This classification of acalyptrate mating systems is, of course, an over-
simplification in some respects. One complication is the presence in a single
species of alternative mating "strategies," as discussed by Cade (1980).
Smith and Prokopy (1980) and Prokopy and Hendrich (1979) have given
examples of alternative patterns in tephritid fruit flies. In Rhagoletis
pomonella, female-initiated matings take place on vegetation early in the
fruiting season of their hosts; later, when most females have already mated,
males attempt to force copulations on fruit with ovipositing females (Smith
and Prokopy 1980). A similar vegetation/host encounter site dichotomy is
found in Ceratitis capitata; in this case male mating strategies differ accord-
ing to the time of day (Prokopy and Hendrichs 1979).
3A controversy exists as to whether such female selection for high-quality
males can really be effective. It is argued that any genetic variance in char-
acters selected by females will be quickly eliminated (see Borgia 1979 and
Thornhill 1980). A useful way of considering this problem would be the
"E.S.S." method (see Cade 1980). Would such a female choice "strategy" be
invadable by a strategy in which females mated randomly?

I thank Dr. J. E. Lloyd for inviting me to speak in the symposium and for
suggesting the topic; Drs. T. J. Walker and' M. Greenfield for many valuable
discussions and insights; Drs. D. L. Chambers and J. C. Webb for introduc-
ing me to the caribfly; and the staff of the USDA Insect Attractants, Be-
havior, and Basic Biology Laboratory in Gainesville for all their help.

40 Florida Entomologist 64 (1) March, 1981

ALCOCK, J. 1980. Natural selection and the mating systems of solitary bees.
American Sci. 68: 146-53.
-- AND D. W. PYLE. 1979. The complex courtship behavior of Physi-
phora demandata (F.) (Diptera: Otitidae). Z. Tierpsychol. 49:
ALEXANDER, R. D. 1975. Natural selection and specialized chorusing behavior
in acoustical insects. Pages 35-77 in D. Pimental, ed. Insects, science,
and society. Academic Press, N.Y.
BARNARD, C. J., AND T. BURK. 1979. Dominance hierarchies and the evolu-
tion of "individual recognition." J. Theoret. Biol. 81: 65-73.
BATRA, S. W. T. 1979. Reproductive behavior of Euaresta bella and E.
festival (Diptera: Tephritidae), potential agents for the biological
control of adventive North American ragweeds (Ambrosia spp.) in
Eurasia. J. New York Ent. Soc. 87: 118-25.
BENNET-CLARK, H. C. 1970. The mechanism and efficiency of sound produc-
tion in mole crickets. J. Exp. Biol. 52: 619-52.
-- AND A. W. EWING. 1970. The love song of the fruit fly. Sci. American
223(1) : 84-92.
Y. LEROY, AND L. TSACAS. 1980. Species and sex-specific songs and
courtship behaviour in the genus Zaprionus (Diptera-Drosophilidae).
Anim. Behav. 28: 230-55.
BERUBE, D. E. 1978. Larval descriptions and biology of Tephritis dilacerata
(Dipt.: Tephritidae), a candidate for biocontrol of Sonchus arvensis
in Canada. Entomophaga 23: 69-82.
BIGGS, J. D. 1972. Aggressive behavior in the adult apple maggot (Diptera;
Tephritidae). Canadian Ent. 104: 349-53.
BORGIA, G. 1979. Sexual selection and the evolution of mating systems. Pages
19-80 in M.S. and N.A. Blum, eds. Sexual selection and reproductive
competition in insects. Academic Press, N.Y.
BURGER, J. F., J. R. ANDERSON, AND M. F. KNUDSEN. 1980. The habits and
life history of Oedoparena glauca (Diptera: Dryomyzidae), a predator
of barnacles. Proc. Ent. Soc. Wash. 82: 360-77.
BURK, T. 1979. An analysis of social behaviour in crickets. D. Phil. thesis,
Oxford Univ., Oxford, England. 335 p.
BUSH, G. L. 1969. Mating behavior, host specificity, and the ecological sig-
nificance of sibling species in frugivorous flies of the genus Rhagoletis
(Diptera-Tephritidae). American Nat. 103: 669-72.
CADE, W. 1980. Alternative male reproductive behaviors. Florida Ent. 63:
CHAMBERS, D. L. 1977. Attractants for fruit fly survey and control. Pages
327-44 in H. H. Shorey and J. J. McKelvey, Jr., eds. Chemical control
of insect behavior: theory and application. John Wiley and Sons, N.Y.
CHRISTENSON, J. D., AND R. H. FOOTE. 1960. Biology of fruit flies. Annu.
Rev. Ent. 5: 171-92.
CHVALA, M., J. DOSKOCIL, J. H. MOOK, AND V. POKORNY. 1974. The genus
Lipara Meigen (Diptera, Chloropidae), systematics, morphology, be-
haviour, and ecology. Tijd. Ent. 117: 1-25.
CRANKSHAW, O. S. 1979. Female choice in relation to calling and courtship
songs in Acheta domesticus. Anim. Behav. 27: 1274-5.
CULLEN, E. 1957. Adaptations in the kittiwake to cliff-nesting. Ibis 99: 275-
Reproduction in the Mediterranean fruit fly: depletion of stored sperm
in females. Ann. Ent. Soc. America 64: 312-3.
DAVIES, N. B. 1978. Ecological questions about territorial behavior. Pages

Insect Behavioral Ecology-'80 Burk

317-50 in J. R. Krebs and N. B. Davies, eds. Behavioural ecology: an
evolutionary approach. Blackwell, Oxford, England.
DAWKINS, R., AND J. R. KREBS. 1978. Animal signals: information or manip-
ulation? Pages 282-309 in J. R. Krebs and N. B. Davies, eds. Be-
havioural ecology: an evolutionary approach. Blackwell, Oxford, Eng-
AND J. R. KREBS. 1979. Arms races between and within species. Proc.
R. Soc. London, B. 205: 489-511.
DODSON, G. N. 1978. Behavioral, anatomical, and physiological aspects of
reproduction in the Caribbean fruit fly. Anastrepha suspense (Loew).
M.S. thesis, Univ. of Florida, Gainesville. 68 p.
DOWNES, J. A. 1969. The swarming and mating flight of Diptera. Annu. Rev.
Ent. 14: 271-98.
EMLEN, S. T., AND L. W. ORING. 1977. Ecology, sexual selection, and the
evolution of mating systems. Science 179: 215-23.
EWING, A. W. 1977. Communication in Diptera. Pages 403-17 in T. A.
Sebeok, ed. How animals communicate. Indiana Univ., Bloomington.
FISHER, R. A. 1958. The genetical theory of natural selection. Dover, N.Y.
291 p.
FLETCHER, B. S. 1977. Behavioral responses of Diptera to pheromones,
allomones, and kairomones. Pages 129-48 in H. H. Shorey and J. J.
McKelvey, Jr., eds. Chemical control of insect behavior: theory and
application. John Wiley, and Son, N.Y.
FOOTE, B. A. 1967. Biology and immature states of fruit flies: the genus
Icterica (Diptera: Tephritidae). Ann. Ent. Soc. America 60: 1295-305.
FORREST, T. G. 1980. Phonotaxis in mole crickets: its reproductive signif-
icance. Florida Ent. 63: 45-53.
HANIOTAKIS, G. E. 1974. Sexual attraction in the olive fruit fly, Dacus oleae
(Gmelin). Environ. Ent. 3: 82-6.
HUETTEL, M. D. 1976. Monitoring the quality of laboratory-reared insects:
a biological and behavioral perspective. Environ. Ent. 5: 807-14.
KATSOYANNOS, B. I. 1975. Oviposition-deterring, male-arresting, fruit- mark-
ing pheromone in Rhagoletis cerasi. Environ. Ent. 4: 801-7.
AND E. F. BOLLER. 1980. Second field application of oviposition-
deterring pheromone of the European cherry fruit fly, Rhagoletis
cerasi L. (Diptera: Tephritidae). Z. ang. Ent. 89: 278-81.
LABEYRIE, V. 1971. Trophic relations and sex meetings in insects. Acta
Phytopathol. Acad. Sci. Hungary 6: 229-34.
LALL, A. B., H. H. SELIGER, W. H. BIGGLEY, AND J. E. LLOYD. 1980. Ecology
of colors of firefly luminescence. Science 210: 560-2.
LLOYD, J. E. 1979. Mating behavior and natural selection. Florida Ent. 62:
1980. Insect behavioral ecology: coming of age in bionomics or
compleat biologists have revolutions too. Florida Ent. 63: 1-4.
LORENZ, K. Z. 1958. The evolution of behavior. Sci. American 199(6) : 67-78.
MANGAN, R. L. 1979. Reproductive behavior of the cactus fly, Odontoloxozus
longicornis, male territoriality and female guarding as adaptive
strategies. Behav. Ecol. Sociobiol. 4: 265-78.
MAYNARD SMITH, J., AND G. A. PARKER. 1976. The logic of asymmetric con-
tests. Anim. Behav. 24: 159-75.
McALPINE, D. K. 1973. Observations on sexual behavior in some Australian
Platystomatidae (Diptera, Schizophora). Rec. Australian Mus. 29:
1979. Agonistic behavior in Achias australis (Diptera, Platy-
stomatidae) and the significance of eyestalks. Pages 221-30 in M.S.
and N.A. Blum, eds. Sexual selection and reproductive competition in

Florida Entomologist 64(1)

March, 1981

insects. Academic Press, N.Y.
MCALPINE, J. F., AND D. D. MUNROE. 1968. Swarming of lonchaeid flies and
other insects, with descriptions of four new species of Lonchaeidae
(Diptera). Canadian Ent. 100: 1154-78.
MONRO, J. 1953. Stridulation in the Queensland fruit fly, Dacus (Strumeta)
tryoni (Frogg.). Australian J. Sci. 16: 60-2.
MOOK, J. H., AND C. G. BRUGGEMANN. 1968. Acoustical communication by
Lipara lucens (Diptera, Chloropidae). Ent. Exp. and Appl. 11: 397-
MORTON, E. S. 1975. Ecological sources of selection on avian sounds. Amer-
ican Nat. 109: 17-34.
NATION, J. L. 1972. Courtship behavior and evidence for a sex attractant in
the male Caribbean fruit fly, Anastrepha suspense. Ann. Ent. Soc.
America 65: 1364-7.
NOVAK, J. A., AND B. A. FOOTE. 1975. Biology and immature stages of fruit
flies: the genus Stenopa (Diptera: Tephritidae). J. Kansas Ent. Soc.
48: 42-52.
MOTO, AND H. HIGA. 1981. Oriental fruit fly and melon fly: biological
and chemical studies on smoke produced by males. Environ. Ent. (in
OLDROYD, H. 1964. The natural history of flies. Weidenfield and Nicolson,
London. 324 p.
OTTE, D. 1977. Communication in Orthoptera. Pages 334-61 in T. A. Sebeok,
ed. How animals communicate. Indiana Univ., Bloomington.
PARKER, G. A. 1970. Sperm competition and its evolutionary consequences
in the insects. Biol. Rev. 45: 525-68.
1972. Reproductive behaviour of Sepsis cynipsea (L.). I. A pre-
liminary analysis of the reproductive strategy and its associated be-
haviour patterns. Behaviour 41: 172-206.
1974a. Courtship persistence and female-guarding as male time-
investment strategies. Behaviour 48: 157-84.
.1974b. Assessment strategy and the evolution of animal conflicts.
J. Theoret. Biol. 47: 223-43.
1978. Evolution of competitive mate searching. Annu. Rev. Ent.
23: 173-96.
PAUL, R. C., AND T. J. WALKER. 1979. Arboreal singing in a burrowing
cricket, Anurogryllus arboreus. J. Comp. Physiol. 132: 217-23.
PERDOMO, A. J., J. L. NATION, AND R. M. BARANOWSKI. 1976. Attraction of
female and male Caribbean fruit flies to food-baited and male-baited
traps under field conditions. Environ. Ent. 5: 1208-10.
PRITCHARD, G. 1967. Laboratory observations on the mating behaviour of
the island fruit fly Rioxa pornia (Diptera: Tephritidae). J. Australian
Ent. Soc. 6: 127-32.
PROKOPY, R. J. 1980. Mating behavior of frugivorous Tephritidae in nature.
Pages 37-46 in Proc. Symp. Fruit Fly Problems, XVI Int. Congr. Ent.,
AND J. HENDRICHS. 1979. Mating behavior of Ceratitis capitata on a
field caged host tree. Ann. Ent. Soc. America 72: 642-8.
SAND R. P. WEBSTER. 1978. Oviposition-deterring pheromone of
Rhagoletis pomonella: a kairomone for its parasitoid Opius lectus.
J. Chem. Ecol. 4: 481-94.
J. R. ZIEGLER, AND T. T. Y. WONG. 1978. Deterrence of repeated ovi-
position by fruit-marking pheromone in Ceratitis capitata (Diptera:
Tephritidae). J. Chem. Ecol. 4: 55-63.
RINGO, J. M. 1976. A communal display in Hawaiian Drosophila (Diptera:

Insect Behavioral Ecology-'80 Burk

Drosophilidae). Ann. Ent. Soc. America 69: 209-14.
ROLLI, K. 1976. Die akustischen Sexualsignale von Ceratitis capitata Wied.
und Dacus oleae Gmel. Z. ang. Ent. 81: 219-23.
SCHILCHER, F. VON. 1976a. The role of auditory stimuli in the courtship of
Drosophila melanogaster. Anim. Behav. 24: 18-26.
1976b. The function of pulse song and sine song in the courtship of
Drosophila melanogaster. Anim. Behav. 24: 622-5.
SHARP, J. L., AND J. C. WEBB. 1977. Flight performance and signaling sound
of irradiated or unirradiated Anastrepha suspense. Proc. Hawaiian
Ent. Soc. 22: 525-32.
SMITH, D. C., AND R. J. PROKOPY. 1980. Mating behavior of Rhagoletis
pomonella (Diptera: Tephritidae). VI. Site of early-season encounters.
Canadian Ent. 112: 585-90.
SPIETH, H. T. 1968. Evolutionary implications of sexual behavior in
Drosophila. Evol. Biol. 2: 157-93.
SPIETH, H. T. 1974. Courtship behavior in Drosophila. Annu. Rev. Ent. 19:
STOLTZFUS, W. B., AND B. A. FOOT. 1965. The use of froth masses in court-
ship of Eutreta. Proc. Ent. Soc. Washington. 67: 263-4.
SWANSON, R. W., AND R. M. BARANOWSKI. 1972. Host range and infestation
by the Caribbean fruit fly, Anastrepha suspense (Diptera: Tephrit-
idae) in south Florida. Proc. Florida State Hort. Soc. 85: 271-4.
THORNHILL, R. 1980. Competitive, charming males and choosy females: was
Darwin correct? Florida Ent. 63: 5-30.
TINBERGEN, N. 1951. The study of instinct. Oxford Univ., Oxford, England.
228 p.
TOBIN, E. N., AND J. G. STOFFOLANO, JR. 1973. The courtship of Musca
species found in North America. 2. The face fly, Musca autumnalis,
and a comparison. Ann. Ent. Soc. America 66: 1329-34.
TRIVERS, R. L. 1972. Parental investment and sexual selection. Pages 136-79
in B. Campbell, ed. Sexual selection and the descent of man, 1871-
1971. Aldine, Chicago.
TYCHSEN, P. H. 1977. Mating behavior of the Queensland fruit fly, Dacus
tryoni, in field cages. J. Australian Ent. Soc. 16: 459-65.
WALKER, W. F. 1980. Sperm utilization strategies in nonsocial insects.
American Nat. 115: 780-99.
WANGBERG, J. K. 1978. Biology of gall-formers of the genus Valentibulla
(Diptera: Tephritidae) on rabbitbrush in Idaho. J. Kansas Ent. Soc.
51: 472-83.
WEBB, J. C. 1973. Analysis and identification of specialized sounds possibly
used by the Caribbean fruit fly, Anastrepha suspense for communica-
tive purposes. Ph.D. diss., Univ. of Tennessee, Knoxville. 169 p.
----, J. L. SHARP, D. L. CHAMBERS, J. J. McDow, AND J. C. BENNER. 1976.
The analysis and identification of sounds produced by the male Carib-
bean fruit fly, Anastrepha suspense (Loew). Ann. Ent. Soc. America
69: 415-20.
WEST EBERHARD, M. J. 1979. Sexual selection, social competition, and evolu-
tion. Proc. Phil. Soc. America 123: 222-34.
WHEELER, W. M. 1924. Courtship of the calobatas. J. Hered. 15: 485-95.
ZWOLFER, H. 1974a. Innerartliche Kommunikationssysteme bei Bohrfliegen.
Biol. unser. Zeit 4(5) : 147-53.
1974b. Das Treffpunkt-Prinzip als Kommunikationsstrategie und
Isolationsmechanismus bei Bohrfliegen (Diptera: Trypetidae). Ent.
Germ. 1: 11-20.

Florida Entomologist 64 (1)

March, 1981



The active aggregations of small animals are often called "swarms." The
term has been applied to the mass flight of workers and queen from a fission-
ing beehive, and to the dispersal flights of termite reproductive. Swarming
in insects otherwise is generally a mating activity characterized by pair
formation in flight.
There is great variation among insect mating swarms with regard to the
degree of aggregation. Most swarms fit into one of two classes, which are the
extremes of a continuum. In one class, males fly throughout an area of
female availability and are, more or less, evenly spaced. In the other, males
are strongly clumped within a small portion of the volume where females
are available and they tend to fly at a station for a prolonged period.
I use the modifer "dispersed" to refer to swarming of the first class, while
the unmodified swarm is reserved for the second class. Gruhl (1955) pro-
posed the terms "synhesmia" and "synorchesia" for the same purpose. The
adoption of particular terms by other workers is not as desirable as is a
general realization that there are important behavioral differences among
various mating swarms, and that it is therefore often worthwhile to describe
swarms in considerable detail.
Dispersed swarming describes situations in which males individually
search for and pair with females while in flight and in sufficient density that
someone has called the phenomenon a "swarm." This activity occurs in areas
where females are emerging, feeding, ovipositing, or dispersing when flight
is integral to these pursuits. Males of some mayflies and caddisflies fly in
great numbers over water and shoreline vegetation where females are
emerging or resting, and intercept females as they fly (e.g. Needham et al.
1935, Solem 1978). Mass mating flights of ants may be synchronized over a
large area (Sudd 1967). The "swarms" of lovebugs, Plecia nearctica
(Bibionidae), seem to be dispersed swarms (Thornhill 1980a) 1.
Small dense swarms of the second variety occupy a volume of a cubic
meter or so and usually keep station with reference to some object. This
object, the "swarm marker" (Downes 1969), generally has no value to the
swarming species except its use in swarming, but flying males of some spe-
cies may aggregate at resources where females may predictably occur, such
as the hosts of blood feeding flies (e.g. Gubler and Bhattacharya 1972).
Swarms may be closely spaced or separated by hundreds of meters, but in
any case the greatest part of the volume within which females fly has a very
low male density in comparison to the swarms.
The process of assembly at a marker occurs in many insect species that
do not mate in flight. Shields (1967) provides an extensive review of hill-
topping behavior, which is the best known form of the phenomenon.

*Robert T. Sullivan is a graduate student at the Department of Entomology and Nema-
tology, University of Florida. His research interests include sexual selection, the evolution and
ecology of mating systems, and the biology of crane flies. Florida Agricultural Experiment
Station Journal Series No. 2826. Current Address: Department of Entomology and Nema-
tology, University of Florida, Gainesville, FL 32611.
1Superscripts refer to notes in the appendix.

Insect Behavioral Ecology-'80 Sullivan

Many observers have noted that mating occurs in connection with swarm-
ing, and the promotion of pair formation has been generally regarded as the
function of male swarming (e.g. Downes 1969). The idea that mating is not
the function of swarms has been advanced to account for swarms in which
few or no matings were observed (Nielsen and Greve 1950, Nielsen and
Haeger 1960). These workers note that the lack of a plausible alternative
purpose for swarms is not, in itself, evidence for a mating function. Reports
of swarming without mating may in part be due to poor conditions for ob-
serving. Corbet (1964) found that efforts to improve the methods of observa-
tion allowed the detection of frequent matings in the swarms of two mos-
quito species, and the observations of Nielsen and Haeger (1960) indicate
that swarm mating was detected where special efforts at observation were
made. Swarming without mating may actually occur as an occasional effect
of the unpredictable nature of female emergences in some species, as the
mosquito, Aedes taeniorhynchus (Nielsen and Haeger 1960). Some other
factor may be involved in Blacus ruficornis (Braconidae) where prolonged
observations and numerous captures of swarms failed to demonstrate the
presence of females (Syrjamaki 1976, but see van Achterberg 1977). Male
swarming without mating does require explanation, and swarming may have
effects aside from mating, but the traditional view that swarming and mating
are functionally related seems to be generally correct.
It seems logical that females are attracted to swarms and not accidentally
intercepted, because in the latter case males that flew outside of swarms
would have higher reproductive success than males in swarms due to less
competition for mates. Females have often been observed flying into swarms,
but the problem of their attraction to swarms as opposed to accidental inter-
ception has not been addressed by field biologists.
A female that flies to a swarm spends energy and incurs a risk of preda-
tion that might be avoided if she were located by a male and mated while
resting or feeding. The advantages of flying to a swarm must exceed these
costs, on the average.
One hypothesis for the existence of non-resource based male mating ag-
gregations such as swarms has involved a type of female choice. In this
view, females benefit from comparing the available males before choosing a
mate, and they prefer to choose from within groups of males (Otte 1974,
Alexander 1975). Female choice must be based on male differences, either in
genetic quality or in the material investment that a male might provide
during copulation (Borgia 1979).
Males of certain species of Empididae give the female an insect prey or
a ball of froth or silk on which she may feed while mating. These flies ap-
proach each other cautiously in swarms (e.g. Downes 1970) and although
discrimination has not been studied it is probably characteristic of both
sexes. Females might discriminate between males on the basis of the size of
the prey (see Thornhill 1977). This discrimination might be less important
when males with prey are in short supply and swarms of waiting females
occur (Steyskal 1941, Alcock 1973). Male discrimination and caution in
mating could result from fecundity differences between females or from the
presence in swarms of males without prey (Downes 1970, Alcock 1973) that
might adopt female behavior in an effort to steal prey (see Thornhill 1979).

Florida Entomologist 64 (1)

Investment through the male ejaculate may occur in swarming insects, but I
am not aware of supportive evidence (see Hinton 1974, Lloyd 1979, Sivinski
Genetically based female discrimination between males, if it occurs, must
rely on characters of the phenotype to indicate genetic quality. Swarms have
not been studied with regard to differential male mating success and the
possibility of genetic correlates. Large male size and fast flight with complex
maneuvers in the swarm may be good general indicators of genetic quality,
and if males with these qualities have above average mating success this
could be due to the exercise of female preference. But this pattern would not
necessarily mean that female choice is acting, because male mating com-
petition in the swarms could also result in large, fast males being most suc-
cessful. Physical combat between swarming males for possession of a female
has been frequently reported (Knab 1906; Howard et al. 1915; Cambournac
and Hill 1940; Belkin et al. 1951; Frohne and Frohne 1952, 1954; Blickle
1959; Nielsen and Haeger 1960; Haeger 1960; Chiang 1962; Quraishi 1965;
Hilsenhoff 1966; Chiba 1967; Savolainen and Syrjamaki 1971; Moorhouse
and Colbo 1973). These occurrences may be due solely to male competition
and so be detrimental to females, which lose time and risk injury from bat-
tling males. Or, a female may actively invite such combat as a means of
acquiring a large male that might contribute material resources to her
progeny and whose genetic quality with regard to his ability to gather re-
sources as a larva is evinced by his size. Females that are first captured in
a swarm by small males might resist leaving the swarm and so increase the
chance of interference by other males (see Cox and LeBoeuf 1977).
The importance of intersexual selection in swarming insects may be
clarified by studies of female preference. Choice would be supported by show-
ing that (i) females prefer certain males and (ii) they benefit by the prefer-
ence (Thornhill 1980b). Theoretical considerations of genetically based fe-
male choice suggest that male genetic variability would be rapidly depleted,
so that there would be no substrate for genetic choice (Williams 1975,
Maynard Smith 1978, but see Borgia 1979).
Swarms may occur because they afford a female the most efficient means
of becoming inseminated. Swarming species probably evolved from species
in which males were the searching sex (see Williams 1975; Greenfield, this
symposium). I noted above that a female might incur energy costs and risk
of predation in flying to a swarm, and should prefer to mate without these
costs. But environmental rhythms result in certain times being most suitable
for mating. For example, female mosquitoes of some species oviposit at night
(e.g. Haddow and Gillett 1957, 1958). An unmated female has reduced
prospects of mating as night approaches, to the extent that mate location is
visually mediated. To remain unmated might result in the loss of an entire
day with consequent increased risk of death by predation or disease before
reproduction. In such situations females might take flight to enhance their
conspicuousness to males or to search for males. A genetic tendency for in-
dividuals of both sexes to fly to conspicuous environmental features would
be favored if it led to a reduction of average search time. In certain situa-
tions then both sexes might fly to these conspicuous features, the swarm
markers. Males would presumably only fly at markers if their mating success
was greater here than elsewhere, due to the attraction of females from the

March, 1981

Insect Behavioral Ecology-'80 Sullivan

surrounding area. This evolutionary route to swarming might be called
"female desperation" to emphasize the contrasts with female choice.
Low male density may result in poor mating prospects for females that
wait to be found by males, and may thus lead to swarming and mating at
markers. Sibling species or populations of a species that differ in density
might adopt different behaviors. The degree of synchrony in adult eclosion
affects male density, and swarming at markers may be less likely in syn-
chronous species than in species where emergence is diffuse. Fluctuations in
adult male density in a population within a season or from year to year
might induce shifts between swarming and other behaviors. I am not aware
of evidence that supports these proposed relationships of density and male
swarming. But Scott (1968) has shown that butterfly species that typically
occur at low densities are more likely than common species to engage in hill-
topping behavior, which is analogous to swarming, as noted above. In addi-
tion, hilltopping in two species "breaks down when these species are dense
and mating occurs on hillsides and sloping ridges as well as hilltops."
Male search efficiency may influence the evolution of swarming. Tipuline
crane fly males in general locate females by flying low over leaf litter and
repeatedly touching the substrate with their tarsi. Limoniinae are often
active in the same habitats, but male search of the Tipuline sort is uncomm-
mon in this subfamily, from which swarming is the most frequently recorded
adult behavior (Alexander 1919, Sullivan in preparation). Limoniinae in
general are much smaller than Tipulinae. The difference in behavior may
result from this size difference, since tactile search must rapidly become less
effective as the sizes of the searcher and the search object decrease.
The "female choice" and "female desperation" hypotheses for the ex-
istence of swarms need not be mutually exclusive. Species that limit mating
to swarms due to intersexual selection may also be served by swarming as
the most efficient way for the sexes to meet. In addition, a number of insect
species are known to mate both in and out of swarms (Cambournac and Hill
1940; Gibson 1945; Nielsen and Nielsen 1953; Downes 1958; Thew 1958;
Nielsen and Haeger 1960; Nielsen 1962; Syrjamaki 1964, 1965, 1967; Peters
and Peters 1977). Thus, in at least some swarming species, female choice at
swarms is not always essential to mating, and since non-swarm mating and
male swarming tend to be temporally separated in these species the "female
desperation" explanation may apply2. Yet females of these dual behavior
species that mate at swarms may exercise choice, while females mating out-
side of swarms may judge males by an internal standard or may mate op-
portunistically to ensure a sperm supply, and later search for swarms to
find a more desirable mate or to diversify the genotypes of their offspring
through additional matings.


Most swarms are associated with a visual marker3. As Downes (1969)
notes, "Swarm markers, though of many different forms, are usually objects
of a kind that human beings would also regard as useful landmarks; the
relatively infrequent objects of large size, or notable contrast against the
ground or sky, or sharp boundaries, or conspicuous angles." Markers range
in size from copies and stones to church steeples. Ephemeral spots of sun-
light are often used as markers. The criteria used by insects to select markers

Florida Entomologist 64 (1)

March, 1981

may be quite subtle. Klassen and Hocking (1964) have noted that certain
objects are used repeatedly by the mosquito Aedes cataphylla while similar
objects nearby are neglected. Selection for pre-mating reproductive isolation
between sibling species could result in a shift of swarm markers. Savolainen
(1978) reports the use of trees and pale earth patches by the mayfly
Leptophlebia marginata (Leptophlebiidae). Two sibling species may actually
be involved since marked individuals retained their marker preference when
transferred to habitats where swarming over the other marker predominated.
Two sympatric species of Cricotopus (Chironomidae) differ in their swarm
markers, one preferring to fly over shrubbery while the second chooses light
colored objects in grassy areas (Titmus 1980). Swarms of Aedes caspius
(Culicidae) disperse when light markers are placed beneath them, while
swarms of the sympatric sibliing species Aedes dorsalis concentrate above
these markers (Nielsen and Nielsen 1963).
Sometimes swarms have no apparent reference to a marker. In these
cases the swarm alone must serve as the cue for female attraction. Marker-
less swarms may occur where no marker is available, such as over a lake or
savannah (Southwood 1957, Nielsen 1962, Thompson 1967, McGowan 1975),
or they may compete with marker swarms, in which event they disperse in
poor light while marker swarms persist (Nielsen and Greve 1950, Southwood
1957). Both a scarcity of markers and competition with other swarms con-
tributed to the occurrence of markerless swarms in the ceratopogonid
Culicoides brevitarsis (Campbell and Kettle 1979).
Swarms of some species occasionally consist of thousands of males (e.g.
Weber 1963, Hubbard and Nagell 1976), but it is also proper to refer to a
single male as "swarming" if his behavior resembles that of individuals in
larger swarms (Downes 1969). Swarms are usually larger when male density
is high, but there may be an upper limit at which additional males form a
new and competing swarm (Syrjamaki 1965). Large swarms, with more
males, tend to persist longer than small swarms (Nielsen and Nielsen 1958,
Haddow and Corbet 1961). This is readily explained to the extent that larger
swarms tend to be over larger markers (Downes 1955, Chiang 1961) since
larger markers will generally be more visible, to both sexes, in the poor light
of dusk and dawn. Savolainen (1978) reported experimental studies with a
crepuscularly active mayfly that provided a direct correlation of large
swarms, large markers and long duration.
The shape of a swarm is presumably the result of marker size, the num-
ber of swarming males, and the heights at which females approach the
swarm. The size of the marker may have a directly proportional effect on
the diameter of the swarm (e.g. Cambournac and Hill 1940, Chiba 1967).
A markerless swarm of chironomids that was 4 feet in diameter condensed
to 2 feet when an observer walked beneath it (Gibson 1945). An increase in
the number of males in a spherical swarm may cause it to increase in
diameter (Hoy and Anderson 1978) or to become taller and so columnar
(Syrjamaki 1964). These different responses presumably reflect the statistical
distribution of the heights of females approaching the swarms.
The shape of swarms of the ceratopogonid Culicoides brevitarsis is cor-
related with their height above the cow manure marker (Campbell and
Kettle 1979). "Swarms formed at two levels. High swarms were essentially
columnar, 20-50 cm in diameter, with base at 0.6-1.2 m above ground and

Insect Behavioral Ecology-'80 Sullivan

top at 1.5-2.1 m. Low swarms were spherical, 30-60 cm in diameter, with
base at 10-40 cm and top at 60-120 cm. Swarms occasionally moved en masse
between levels without dispersal, in 5 s or less. The swarm always assumed
the shape appropriate to its height." Swarms near cattle are more likely to
be low. Females oviposit in cow manure and since fresh manure is more likely
to be found about cows, it is logical that females near cows would be flying
low, and male swarms might be low and relatively dense. Females flying at
greater heights might not be so predictably concentrated at a uniform level,
and high male swarms would be elongate and less dense. It is not evident
what factors induce the rapid movement of swarms between levels, or why
males do not become distributed throughout the entire range of heights. The
mosquito Mansonia fuscopennata when swarming above a human marker
forms a low, flat swarm close to the head and a higher columnar swarm
(Corbet 1964). The low swarm might intercept females arriving to feed.
Downes (1955) recorded a difference of shape between low swarms over
dung and high swarms over other markers. High swarms fluctuated in height.
Vertical movements of the swarms of Anopheles franciscanus (Culicidae)
are correlated with female presence at swarms (Belkin et al. 1951). Mos-
quito swarm height changes induced by a clarinet note (Nielsen 1965) and
the human voice (Knab 1907) may have evolved as responses to the flight
tone of female mosquitoes (Roth 1948). Other cases are known of swarms
changing height as units (Arnaud 1952, Wharton 1953, Haddow and Corbet
1961, Corbet 1964, McAlpine and Munroe 1968, Savolainen 1978). It seems
likely that the causes of these movements were perceived or probabilistic
changes in the availability of females at various heights, but there is no
evidence for this.
The ant Lasius sitkaensis forms columnar swarms that occur in ag-
gregates of 10-15 and so constitute "composite swarms" of 6-12 m diameter
(Corbet and Ayre 1968). It is difficult to explain why aggregates with this
design would exist. Some swarms occurred over trees that might attract
females, but most were located over a highway. The female attractant in
these cases may have been the active masses of males constituting the com-
posite swarms, but it is unclear why the separate columns would be retained.
Since the individual swarms would compete for the females arriving at the
composite, one might predict that males would not remain in a swarm that
was shielded by other swarms, and a ring of swarms would result. Males
flying outside of these swarms yet still on the perimeter would face less
competition for the females being attracted by the composite, and the com-
posite with its heterogeneous distribution of male density should yield to a
homogeneous, single swarm. A similar case is recorded for an unidentified
"gnat"; several distinct columnar swarms occurred over each of a number of
trees (Knab 1906). Perhaps females at a distance respond to the entire tree
as a marker, while at a closer range they are attracted to particular


Dragonflies prey on swarming insects of so many species that "swarm
feeding" is a recognized category of behavior (Banks 1919, Corbet 1962,
Corbet and Haddow 1962, Nielsen 1965, van Someren 1975). Other swarm
predators include bats (Rao and Russell 1938, van Someren 1975), night-

Florida Entomologist 64 (1)

hawks (Warter et al., in Moser 1967), the digger wasp Bembix belfragei
(Blickle 1959), empidid flies (Downes 1970, Syrjamaki 1966) and ceratopo-
gonid midges (Downes 1978).
It is important to know the predation risk associated with each position
in a swarm. A male might be safest in the center of the group (Hamilton
1971), but this would probably conflict with the imperative to intercept
arriving females. From observations of stickleback fish attacking "swarms"
of Daphnia, Milinski (1977) found that most attacks were directed toward
stragglers, followed in order by the periphery and the center of the "swarm."
Corbet (1962) noted that dragonflies are able to feed selectively on the
members of a swarm, but swallows and bats feed by "seining" the swarm as
they swoop through repeatedly.

Swarms have not been found to occur during darkness, presumably be-
cause orientation to a marker is visually directed. The crepuscular and day-
light hours are all used for swarming. Most species swarm during only some
of these hours, typically at about the same times on successive days. Com-
mon swarming time patterns include (i) all the hours of strong light, (ii) a
few of these hours such as morning or afternoon, (iii) all crepuscular hours,
(iv) either dusk or dawn, (v) combinations of these times, such as dawn and
morning or afternoon and dusk (e.g. Nielsen and Nielsen 1958, Savolainen
The hours of male swarming must indicate the times when females are
normally active and inclined to fly to swarms (see Parker 1978). Female
activity may be generally influenced by predation and climatic factors.
According to Bellamy (1947) the diurnal disturbance flight of Pseudolimno-
phila inornata (Tipulidae) reflects intense predation by dragonflies, and
swarming begins at dusk as these predators go to roost. Danger from desic-
cation during the midday hours might keep some insects inactive. The
crepuscular hours in general are less windy than the daylight hours (Critch-
field 1960) and thus more favorable for swarming by small insects.
Crepuscular swarming at markers might evolve if the markers of certain
species are typically those objects that could be used by dispersing females
for navigation, to avoid flying in circles. Females would probably use these
markers at dusk and dawn when celestial objects are not generally available
to aid in holding a straight course. Males that aggregated about such
markers at these times would increase the frequency with which they en-
countered females. Selection for premating reproductive isolation might
cause divergence of the swarming times of sympatric, closely related species.
In two cases local climate has apparently affected the time of swarming.
The Enicocephalid bug studied by Schih (1970) swarms before noon while
other species in the family swarm at dusk. Schuh suggested that the shift is
due to torrential afternoon rains in Costa Rica. In Florida the chironomid
Glyptotendipes paripes swarms at dawn during the summer but in November
swarms occur at dusk (Nielsen 1962). Low temperatures at dawn during
November may induce the shift to dusk. Alternatively, the turbulence of the
lower atmosphere in Florida on summer afternoons may favor dawn swarm-
ing during this season.

March, 1981

Insect Behavioral Ecology-'80 Sullivan

The constitution of swarms changes continually as males enter and leave.
Where artificial markers are placed or swarms are removed by netting, full
strength swarms may form within a minute (Syrjamaki 1964, McAlpine and
Munroe 1968, Campbell and Kettle 1979). This flight between swarms is pre-
sumably due to males searching for their best mating chance and evaluating
various swarms with regard to the rate of female arrival and the degree of
male competition (see Parker 1978). An interesting exception to the general
pattern of male movement between swarms exists in Aedes taeniorhynchus
(Culicidae). Marked males were found to have a very low rate of movement
between swarms on succssive evenings (Nielsen and Nielsen 1953). A single
swarmer may leave a marker more rapidly than when other males are pres-
ent (Blickle 1959, Corbet 1964). The explanation may be that males in low
density situations are most successful if they visit a series of markers where
females may be waiting, while at higher densities flight between markers is
less likely to be productive since all markers are probably occupied by
The frequency of male combat in swarms, noted above, suggests that
competition for the arriving females may be a major influence on the flight
paths of individual males in swarms. Swarming males of many insect species
have a flight pattern that seems to provide for continuous sampling of vari-
ous areas on the swarm periphery for incoming females (see below)4. Male
flight might include a species-specific pattern so that females could distin-
guish swarms of their own species from those of other species that are
simultaneously active. In the absence of female choice, such a pattern would
not evolve if it were competitively inferior to other motions that a male
might employ. Therefore, if sympatric species show character displacement
in their flight patterns, the shift might only be among equally competitive
paths. Male flight may also include some indication of genetic quality if
female choice is effective (see above).
In many swarms the speed of flight is so great that the analysis of
motion pictures would be necessary to satisfactorily describe events. With
most swarms problems arise from the poor light in which they occur and the
3 dimensional nature of individual motions. The cecidomyiid Anarete prit-
chardi flies in small swarms in bright sunlight, and motion picture analyses
have provided detailed information about individual flight (Okubo et al.
1977, Chiang et al. 1978, and references therein)5. This system seems to
offer an unusual opportunity for the study of the mechanisms of male mating
competition in swarms. The action of female choice might be indicated by
studies of female motions.
A variety of motions by swarming males can be interpreted as adapta-
tions for mating competition. Horizontal looping presumably allows males
to systematically scan the swarm periphery for approaching females (Downes
1958, Syrjamaki 1966, Reisen et al. 1977). Occasional changes of level may
serve to test the rate of female arrival at various heights (Nielsen and
Greve 1950). A corkscrew course results if loops are combined wth continual
change of height (Knab 1907, Reisen et al. 1977). Horizontal zigzags or
figure 8's are frequent patterns in swarms (Wilson 1957, Nielsen and Nielsen
1962, Wright et al. 1966, McAlpine and Munroe 1968, Syrjamaki 1968, 1976,
Chiang et al. 1978, Eberhard 1978). Chironomus strenzkei has a complex

Florida Entomologist 64 (1)

March, 1981

flight pattern in which zigzags alternate with changes of height and oc-
casional vertical loops (Syrjamaki 1965). Culex pipiens combines zigzags
and horizontal loops with an oscillating forward motion (Chiba 1967).
In some Dipteran swarms the motions are mainly vertical (Corbet 1964,
Dahl 1965, Moorhouse and Colbo 1973). Swarming mayflies of most species
have a "pendular" flight, flying upward for a meter or so then drifting
passively down with the wings, legs and caudal filaments outstretched
(Spieth 1940, Brodsky 1973, Grandi 1973). It may be argued that energy is
conserved by this flight pattern, since the energetic savings of the long
descent might more than offset the cost of the ascent, relative to hovering
flight. But small mayflies in windy conditions achieve pendular flight by
beating the wings during both the ascent and descent, when a level hovering
would surely be less energetically expensive (Spieth 1940, Brodsky 1973).
Males may be sampling the air column for females at various heights, if the
range of heights at which females fly is greater than the distance from which
males can detect them. Some Nemopteridae (Neuroptera) form small
swarms, and resemble mayflies in having a pronounced vertical, flapping and
soaring component to the swarming flight (Balduf 1939, Tjeder 1967). It is
interesting that the hind wings of Nemopteridae are very thin and elongate
and trail in flight somewhat like the filaments of mayflies.
Many insects are apparently unable to conduct their usual maneuvers
when flying in a wind, and they change to hovering flight (e.g. Downes 1969).
Hovering in calm weather occurs among Simuliidae, Rhagionidae, and most
Tabanidae (see Table 1). The comparatively good lighting that exists during



Swarming Swarming







Savolainen 1978
Trost and Berner
1963, Edmunds
et al. 1976
Edmunds 1951
Berner 1950, Edmunds
et al. 1976, Savo-
lainen 1978
Berner 1950, Corbet
1961, Edmunds
et al. 1976, Savo-
lainen 1978
Hall et al. 1975
Savolainen 1978
Peters and Peters
Edmunds et al. 1976,
Savolainen 1978
Thew 1958

Neville 1959


Insect Behavioral Ecology-'80 Sullivan 53


Order/Family Dispersed Swarming Reference*

















Schuh 1970, Kritsky

Southwood 1957
Riek 1974
Balduf 1939, Tjeder
Balfour-Browne 1956
MacNeill 1962

Mori and Matutani
1953, Corbet 1961
Solem 1978
Mori and Matutani
Mori and Matutani
1953, Solem 1978
Solem 1978
Mori and Matutani
Mori and Matutani
Mori and Matutani

1959, Barton
Brown et al. 1969
Gruhl 1927

Arnaud 1952, Syr-
jamaki 1976, van
Achterberg 1977
Townes and Townes
Sudd 1967, Corbet and
Ayre 1968, Eber-
hard 1978
Michener 1974
Strang 1970

Alexander 1919,
Savolainen and
Syrjamaki 1971
Dahl 1965
Crampton 1929

Florida Entomologist 64(1)

March, 1981


Order/Family Swarming Swarming Reference*













Stuckenberg 1958
Nowell 1951
McGowan 1975
Horsfall 1955, Nielsen
and Haeger 1960,
Downes 1969
Downes 1955, 1958;
Campbell and
Kettle 1979
Gibson 1945, Young
1969, Oliver 1971,
Paasivirta 1972
Moorhouse and Colbo
1973, Hunter 1979
Edwards 1928
Thornhill 1980
McAlpine and Munroe
Chiang 1968
Rau 1937, Corbet and
Haddow 1962
Bailey 1948, Corbet
and Haddow 1962
Shemanchuk and
Weintraub 1961,
Turner 1974, Hoy
and Anderson 1978
Crane 1961, Downes
1970, Alcock 1973,
Alcock et al. 1979
Kessel and Kessel
Colyer 1954
Reid 1940, Miller and
McClanahan 1960
Huckett 1954,
Chillcott 1960
Gruhl 1955
McAlpine and Munroe
McAlpine and Munroe
McAlpine and Munroe




*I have preferentially listed reviews or other
their bibliographies.

references that include further references in

references that include

further references in

Insect Behavioral Ecology-'80 Sullivan

the diurnal swarming of these insects may allow males to conserve energy
by hovering, since they can see arriving females at a greater distance than
crepuscular swarmers, which are usually active in their swarm. However,
the Cecidomyiidae (Chiang 1968) and Lonchaeidae (McAlpine and Munroe
1968) fly in bright sunlight and yet are quite active in the swarm. Aggressive
interactions of hovering males have been described from Tabanidae and
Rhagionidae (Blickle 1959, Corbet and Haddow 1962, Hoy and Anderson
1978). Swarms of hovering insects have not been checked for position effects
on male mating success, but since females must enter a swarm from outside
it seems that males on the edge of the swarm might be most successful, and
large males should hold these positions. This would contrast with leks where
female choice is thought to operate and where centrally located males copu-
late most frequently (see Borgia 1979). This contrast provides for a test of
the function of swarms.
Increased competition due to a larger number of males in a swarm might
cause a greater flight speed as males increased the tempo of their peripheral
observations. This relationship between male numbers and flight speed has
been reported several times (Knab 1906, Gibson 1945, Reisen et al. 1977,
Syrjamaki 1965). Our only precise information about flight speeds in swarms
is provided by Chiang's (1968) analysis. He reported that the flight speed of
Anarete pritchardi is the same in small and large swarms, but the smallest
swarm studied had 50 males and flight speed might be maximized at even
lower levels of competition. Flight pattern as well as speed might change,
depending on the degree of male competition. I am not aware of evidence that
supports this possibility.
Male flight is sometimes altered by female arrival at swarms. From a
swarm of Blacus ruficornis, Southwood (1957) described a vertical pattern
of flight and a faster, horizontal style that accompanied condensation of the
swarm. The latter style was thought to be caused by the presence of a fe-
male in the swarm. Belkin et al. (1951) noted that increased activity and
disorganization in swarms of Anopheles franciscanus are apparently due to
female presence. Swarming males of Mansonia perturbans (Culicidae) in-
crease their flight speed in response to a clarinet tone, which may mimic the
sound produced by a flying female (Nielsen 1965).


Swarms consisting mostly or entirely of females have been reported.
Females of Serromyia femorata (Ceratopogonidae) prey on other midges
(Downes 1978), so the female swarms reported by Downes (1955) were
perhaps waiting for prey at the swarm markers of these other species. The
rare female swarms of Trichoceridae (Dahl 1965) may result from a low
male density and the temporary absence of males from a particular marker.
Female Simuliidae over a car (Service 1971) may have mistaken the reflec-
tive surface for water (Benham 1975, Last 1975) and gathered for oviposi-
tion. Female Simuliidae form oviposition "swarms" above streams (Peterson
1959, Moorhouse and Colbo 1973).
Female swarms reported by Eberhard (1978) for Acropygia sp. (Formi-
cidae) are of interest both for their number (89 were found in a pasture on
one occasion) and for the frequent occurrence of homosexual pairing by
females. Females carry symbiotic coccids on their mating and dispersal flight

Florida Entomologist 64 (1)

and perhaps pairing females try to steal from one another. Alternatively, it
is possible that these female swarms resulted from a deficiency of males, so
that swarming females competed for males and became less discriminating
and more aggressive in initiating pair formation.
Several equally puzzling cases involve sexually segregated swarms in
which both sexes fly in close proximity. Culex fatigans (Culicidae) is re-
ported to form both male and female swarms in the same room (Knab 1906).
Downes (1958) stated that females of several species of Ceratopogon fly a
few feet from the males; and that females of Aedes hexodontus (Culicidae)
swarm at 1 to 2 feet above a marker, and males at 3 to 12 feet.

The occurrence of swarming and dispersed swarming in the families of
insects is given in Table 1. Most listed families include non-swarming spe-
cies. The existence of only one reference to swarming for a number of
families suggests that the behavior may yet be recorded as an occasional
event in additional families.
Two taxa are particularly interesting because of their apparent phylo-
genetic isolation from other swarmers. Four species of Enicocephalidae are
the only Hemiptera known to swarm. Swarming is recorded from four
species of the braconid genus Blacus, but is otherwise unknown in parasitic
Hymenoptera with one apparent exception in the Ichneumonidae.


The behavioral tends and explanations proposed here can be tested by
specific observations and experiments in the field. The application of optimal-
ity and genetic selfishness concepts by swarm observers will speed the in-
crease of knowledge by providing for testable hypotheses concerning the
function and evolution of swarming behavior and the adaptive value of the
actions of swarming individuals. The potential for a rapid development of
understanding is great, because numerous workers worldwide are concerned
with the behavior of swarming insects that have economic and medical im-
pact on man".


Insect mating swarms are characterized by the simultaneous flight of
numerous males and the initiation of copulation while in flight. In dispersed
swarms males are distributed through an area and intercept flying females.
Swarms, on the other hand, consist of males flying at particular stations,
often at a conspicuous object (the swarm marker), and usually in aggrega-
tions. It is logical that females are attracted to these swarms, but accidental
interception has not been disproven. Swarms may result from intersexual
selection, and females approach swarms to choose and mate with males that
are of high genetic quality or that will invest in their progeny. Alternatively,
females may only fly to swarms as a last resort, if they require insemination
and the cost of waiting to be found by males exceeds the risk of searching for
swarms. The shape of swarms presumably results from marker size, the
number of participants, and the heights of approaching females. Changes of

March, 1981

Insect Behavioral Ecology-'80 Sullivan

swarm height and associated changes of shape are probably due to perceived
or probabilistic changes in the availability of females at various heights.
The time of swarming may be most strongly influenced by predation and
climatic factors. Male flight between swarms may maximize mating success
by providing for evaluation of competition and female availability at various
swarms. The major influence on flight patterns of swarming males is prob-
ably mating competition. Loops, zigzags, and figure 8's are common male
motions that seem suited to monitoring the swarm periphery for arriving
females. Mayflies and Nemopteridae have a "pendular," rising and falling
flight in swarms. Female arrival at swarms sometimes causes changes in
male flight. Swarms consisting mostly or entirely of females result from a
variety of factors.

1Lovebug "swarms" occur in pastures where adults emerge from turf.
Their classification as dispersed swarms is based on their large size (to 54
m2), the apparent absence of a marker, and the lack of horizontal stratifica-
tion. The largest males defend airspace nearest the ground and have the
greatest mating success because they have initial access to newly closed
females. Males on the lateral edges of the group would experience similar
advantages if females flew to the "swarm" from other parts of the pasture,
and large males should then hold the peripheral positions. The absence of
this pattern indicates that females are not flying to the group from outside,
and that males are flying over the entire area of significant female emer-
2At dusk, males of the mosquito Anopheles maculipennis atroparvus first
search individually for mates, then form a swarm that persists until dark
(Cambournac and Hill 1940). The mosquito Psorophora ciliata swarms at
dusk. During the day some males fly over pools of water, touching their legs
to the surface (Nielsen and Haeger 1960). The probable function of this
flight is to detect newly closed and ovipositing females. In synchronous
broods of the salt marsh mosquito, Aedes taeniorhynchus, most matings
occur on a single evening when flying males intercept females that rise from
perches. Swarming usually occurs 24 hours or more after this event (Nielsen
and Nielsen 1953, Nielsen and Haeger 1960). In contrast to these cases,
matings of the chironomid Allochironomus crassiforceps may simultaneously
occur at swarms and on substrate where males walk actively about, searching
for females (Syrjamaki 1964).
3Males might swarm at discrete auditory or olfactory sources that oc-
curred with appropriate frequency. The persistent calling songs of Ortho-
ptera might provide a suitable beacon. Smoke flies (Platypezidae) that
swarm at bonfires are able to locate smoke by olfactory means alone (Kessel
1960). Swarming male insects might produce sound or pheromones to attract
distant females. But signaling males could be parasitizedd" by non-signalers
that intercepted some of the attracted females without incurring the cost of
signaling (Lloyd 1973, Alexander 1975, Cade 1980). This parasitism would
militate against the production of signals unless females preferentially
mated with signalers.
4Female flight speed must be somewhat less than that of males for this
continual male motion to be competitively advantageous. The diagrams of
motion in Chiang et al. (1978) indicate that male speed in Cecidomyiid
swarms is greater than female speed. Male Tipulidae in swarms fly much
faster than females approaching the swarms (Sullivan, unpublished).
5M. H. Jansen of the Department of Entomology of Purdue University is

Florida Entomologist 64 (1)

March, 1981

studying the swarming flight of Chironomus riparius using computer-aided
reconstructions that are based on motion pictures.
6Virtually all information about swarms is worth recording. The value of
a field study depends on both the range of factors that are studied and the
amount of detailed information that is gathered for each factor. Swarm ob-
servers should consider the following factors for study: (i) swarm size,
shape, habitat, marker, sex ratio, time, and motions in swarms, and variation
in these factors among swarms and on different days; (ii) male arrival at
and departure from swarms, and the daily process of swarm development
and decline; (iii) the effect of size and other phenotypic traits of swarming
males on position, behavior, and mating success; (iv) the female approach
to swarms, the place and process of pair formation in the swarm, and the
fate of pairs; (v) ovarian development and previous mating experience of
females at swarms; (vi) larval habitat and its relation to the sites of swarm-
ing, the temporal pattern of adult eclosion, and adult density; (vii) other
components of the diurnal and seasonal cycle of adult activities, including
mating outside of swarms.


Tim Forrest, James Lloyd, and John Sivinski read the manuscript and
suggested many improvements. Ted Burke, Mike Greenfield, and John
Sivinski supplied references. Lewis Berner helped greatly by providing
access to reprints and discussion of mayfly swarming. J. G. Peters provided
unpublished information on a mayfly species. This paper was prepared during
the tenure of a National Science Foundation Pre-Doctoral Fellowship.

ALCOCK, J. 1973. The mating behaviour of Empis barbatoides Melander and
Empis poplitea Loew (Diptera: Empididae). J. Nat. Hist. 7: 411-20.
R. W. MATTHEWS. 1979. Observations on the mating behavior of the
empidid fly, Rhamphomyia pectoris (Diptera). J. Georgia Ent. Soc.
14: 351-4.
ALEXANDER, C. P. 1919. The crane-flies of New York. Part I. Distribution
and taxonomy of the adult flies. Cornell Univ. Agr. Exp. Stn. Mem. 25.
ALEXANDER, R. D. 1975. Natural selection and specialized chorusing behavior
in acoustical insects. Pages 35-77 in D. Pimentel, ed. Insects, science,
and society. Academic Press, New York.
ARNAUD, P. H. 1952. Swarming of Blacus sp. in California (Hymenoptera:
Braconidae). Ent. News 63: 149-50.
BAILEY, N. S. 1948. The hovering and mating of Tabanidae: A review of
the literature with some original observations. Ann. Ent. Soc. America
41: 403-12.
BALDUF, W. V. 1939. The bionomics of entomophagous insects. Part II. John
Swift, New York.
BALFOUR-BROWNE, J. 1956. [Exhibit notes]. Proc. R. Ent. Soc. London Ser. C.
Journal of Meetings. 20: 8, 51.
BANKS, C. 1919. The swarming of Anopheline mosquitoes. Philipp. J. Sci.
15: 283-8.
WELL. 1969. Mating flight behavior in three species of Oncopera moths
(Lepidoptera: Hepialidae). J. Australian Ent. Soc. 8: 168-72.
BELKIN, J. N., N. EHMANN, AND G. HEID. 1951. Preliminary field observa-
tions on the behavior of the adults of Anopheles franciscanus Mc-

Insect Behavioral Ecology-'80 Sullivan 59

Cracken in southern California. Mosq. News 11: 23-31.
BELLAMY, R. E. 1947. The crane-flies of the Welaka area. Unpubl. Ph.D.
Diss. Univ. Fla., Gainesville.
BENHAM, B. R. 1975. Swarming of Helophorus brevipalpis (Col., Hydro-
philidae) in North Devon. Ent. Mon. Mag. 111: 127-8.
BERNER, L. 1950. The mayflies of Florida. Univ. Fla. Press, Gainesville.
BLICKLE, R. L. 1959. Observations on the hovering and mating of Tabanus
bishoppi Stone (Diptera, Tabanidae). Ann. Ent. Soc. America 52:
BORGIA, G. 1979. Sexual selection and the evolution of mating systems. Pages
19-80 in M. S. Blum and N. A. Blum, eds. Sexual selection and repro-
ductive competition in insects. Academic Press, New York.
BRODSKY, A. K. 1973. The swarming behavior of mayflies (Ephemeroptera).
Ent. Rev. (Engl. Transl. Ent. Obozr.) 52: 33-9.
CADE, W. 1980. Alternative male reproductive behaviors. Florida Ent. 63:
CAMBOURNAC, F. J. C., AND R. B. HILL. 1940. Observation on the swarming
of Anopheles maculipennis, var. atroparvus. American J. Trop. Med.
20: 133-40.
CAMPBELL, M. M., AND D. S. KETTLE. 1979. Swarming of Culicoides brevitar-
sis Kieffer (Diptera: Ceratopogonidae) with reference to markers,
swarm size, proximity of cattle, and weather. Australian J. Zool. 27:
CAROLSFELD-KRAUSE, A. G. 1959. The mating of Hepialidae (Lep.). Ent. Rec.
J. Var. 71: 33-4.
CHIANG, H. C. 1961. Ecology of insect swarms. I. Exp. studies of the be-
haviour of Anarete near felti Pritchard in artificially induced swarms.
Anim. Behav. 9: 213-9.
1962. Ecology of insect swarms. III. Intraspecific interference in
the swarms of Anarete sp. near felti Pritchard (Cecidomyiidae,
Diptera). Proc. Minnesota Acad. Sci. 30: 31-3.
1968. Ecology of insect swarms. V. Movement of individual midges,
Anarete pritchardi, within a swarm. Ann. Ent. Soc. America 61:
---, B. J. METTLER, A. OKUBO, AND A. S. ROBBINS. 1978. Coupling of
midge individuals in a swarm, Anarete pritchardi (Diptera: Ceci-
domyiidae). Ann. Ent. Soc. America 71: 859-61.
CHIBA, Y. 1967. Observations on the swarming habit of the mosquito, Culex
pipiens pallens. Sci. Rep. Tohoku Univ. Fourth Ser. (Biol.) 33: 15-28.
CHILLCOTT, J. G. 1960. A revision of the Nearactic species of Fanniinae
(Diptera: Muscidae). Canadian Ent. Suppl. 14.
COLYER, C. N. 1954. "Swarming" of Phoridae (Diptera). J. Soc. British Ent.
5: 22-7.
CORBET, P. S. 1961. Entomological studies from a high tower in Mpanga
Forest, Uganda. XII. Observations on Ephemeroptera, Odonata and
some other orders. Trans. R. Ent. Soc. London 113: 356-61.
S1962. A biology of dragonflies. Quadrangle, Chicago.
S1964. Observations on the swarming and mating of mosquitoes in
Uganda. Proc. R. Ent. Soc. London Ser. A, Gen. Ent. 39: 15-22.
AND G. L. AYRE. 1968. Swarming and mating in the ant, Lasius
sitkaensis Pergande. Canadian Field-Nat. 82: 230-1.
--- AND A. J. HADDOW. 1962. Diptera swarming high above the forest
canopy in Uganda, with special reference to Tabanidae. Trans. R. Ent.
Soc. London 114: 267-84.
Cox, C. R., AND B. J. LEBOEUF. 1977. Female incitation of male competition:
A mechanism in sexual selection. American Nat. 111: 317-35.

Florida Entomologist 64(1)

March, 1981

CRAMPTON, G. C. 1929. A swarm of males of the rare and primitive crane-
fly Protoplasa fitchii observed near Chandler in the Gaspe Peninsula
(Diptera). Canadian Ent. 61: 70-2.
CRANE, A. E. 1961. A study of the habits of Rhamphomyia scutellaris
Coquillett (Diptera: Empididae). Wasmann J. Biol. 19: 247-63.
CRITCHFIELD, H. J. 1960. General climatology. Prentice Hall, New Jersey.
DAHL, C. 1965. Studies on swarming activity in Trichoceridae (Diptera) in
southern Sweden. Opusc. Ent. Suppl. 27.
DOWNES, J. A. 1955. Observations on the swarming flight and mating of
Culicoides (Diptera: Ceratopogonidae). Trans. R. Ent. Soc. Lond.
106: 213-36.
1958. Assembly and mating in the biting Nematocera. Proc. Tenth
Int. Congr. Ent. 2: 425-34.
1969. The swarming and mating flight of Diptera. Annu. Rev. Ent.
14: 271-98.
.1970. The feeding and mating behaviour of the specialized
Empidinae (Diptera); Observations on four species of Rhamphomyia
in the high arctic and a general discussion. Canadian Ent. 102: 769-91.
1978. Feeding and mating in the insectivorous Ceratopogoninae
(Diptera). Mem. Ent. Soc. Canada No. 104.
EBERHARD, W. G. 1978. Mating swarms of a South American Acropygia
(Hymenoptera: Formicidae). Ent. News 89: 14-6.
EDMUNDS, G. F., JR. 1951. New species of Utah mayflies. I. Oligoneuriidae
(Ephemeroptera). Proc. Ent. Soc. Washington 53: 327-31.
S. L. JENSEN, AND L. BERNER. 1976. The mayflies of North and
Central America. Univ. Minn. Press, Minneapolis.
EDWARDS, F. W. 1928. Diptera. Fam. Protorhyphidae, Anisopodidae, Pachy-
neuridae, Trichoceridae. Fasc. 190 in P. Wytsman, ed. Genera In-
sectorum, Bruxelles.
FROHNE, W. C., AND R. G. FROHNE. 1952. Mating swarms of males of the
mosquito, Aedes punctor (Kirby), in Alaska. Mosq. News 12: 248-51.
-- AND 1954. Diurnal swarms of Culex territans Walker, and
the crepuscular swarming of Aedes about a small glade in Alaska.
Mosq. News 14: 62-4.
GIBSON, N. H. E. 1945. On the mating swarms of certain Chironomidae
(Diptera). Trans. R. Ent. Soc. London 95: 263-94.
GRANDI, M. 1973. I comportamenti di volo nelle diverse fasi della vita aerea
degli Efemerotteri. Boll. Ist. Ent. Univ. Studi Bologna 31: 109-27.
GRUHL, K. 1927. Neue Beitrage zum Problem der Lufttanze. Z. Ent. 15:
1-13. (Not seen, in Downes 1969)
1955. Neue Beobachtungen an Schwarm-sind Tanzgesellschaften der
Dipteren (Dipt.). Dtsch. Ent. Z. 2: 332-53.
GUBLER, D. J., AND N. C. BHATTACHARYA. 1972. Swarming and mating of
Aedes albopictus in nature. Mosq. News 32: 219-23.
HADDOW, A. J., AND P. S. CORBET. 1961. Entomological studies from a high
tower in Mpanga Forest, Uganda. V. Swarming activity above the
forest. Trans. R. Ent. Soc. London 113: 284-300.
AND J. D. GILLETT. 1957. Observations on the oviposition-cycle of
Aedes (Stegomyia) aegypti (Linnaeus). Ann. Trop. Med. Parasitol.
51: 159-69.
-- AND 1958. Laboratory observations on the oviposition-cycle
in the mosquito Taeniorhynchus (Coquillettidia) fuscopennatus Theo-
bald. Ann. Trop. Med. Parasitol. 52: 320-5.
HAEGER, J. S. 1960. Behavior preceding migration in the saltmarsh mos-
quito, Aedes taeniorhynchus (Wiedemann). Mosq. News 20: 136-47.
HALL, R. J., L. BERNER, AND E. F. COOK. 1975. Observations on the biology

Insect Behavioral Ecology-'80 Sullivan

of Tricorythodes atratus McDunnough (Ephemeroptera: Tricory-
thidae). Proc. Ent. Soc. Washington 77: 34-49.
HAMILTON, W. D. 1971. Geometry for the selfish herd. J. Theor. Biol. 31:
HILSENHOFF, W. L. 1966. The biology of Chironomus plumosus (Diptera:
Chironomidae) in Lake Winnebago, Wisconsin. Ann. Ent. Soc. Amer-
ica 59: 465-73.
HINTON, H. E. 1974. Symposium on reproduction of arthropods of medical
and veterinary importance. III. Accessory functions of seminal fluid.
J. Med. Ent. 11: 19-25.
HORSFALL, W. R. 1955. Mosquitoes: Their bionomics and relation to disease.
Ronald Press Co., New York.
HOWARD, L. 0., H. G. DYAR, AND F. KNAB. 1915. The mosquitoes of North
and Central America and the West Indies. Vol. Three: Systematic
Description. Part I. Carnegie Inst. Wash.
HOY, J. B., AND J. R. ANDERSON. 1978. Behavior and reproductive physiology
of bloodsucking snipe flies (Diptera: Rhagionidae: Symphoromyia)
attacking deer in northern California. Hilgardia 46: 113-68.
HUBBARD, M. D., AND B. NAGELL. 1976. Notes on an extraordinarily high
mating swarm of the ant Myrmica laevinodis (Hymenoptera:
Formicidae, Myrmicinae). Ent. News 87: 86.
HUCKETT, H. C. 1954. A review of the North American species belonging to
the genus Hydrotaea Robineau-Desvoidy. Ann. Ent. Soc. America 47:
HUNTER, D. M. 1979. Swarming, mating and resting behaviour of three
species of black fly (Diptera: Simuliidae). J. Australian Ent. Soc.
18: 1-6.
KESSEL, E. L. 1960. The response of Microsania and Hormopeza to smoke
(Diptera: Platypezidae and Empididae). Pan-Pacific Ent. 36: 67-8.
AND B. B. KESSEL. 1961. Observations on the mating behavior of
Platypezina pacifica Kessel (Diptera: Platypezidae). Wasmann J.
Biol. 19: 295-9.
KLASSEN, W. H., AND B. HOCKING. 1964. The influence of a deep river valley
system on the dispersal of Aedes mosquitoes. Bull. Ent. Res. 55: 289-
KNAB, F. 1906. The swarming of Culex pipiens. Psyche 13: 123-33.
-- 1907. The swarming of Anopheles punctipennis Say. Psyche 14: 1-4.
KRITSKY, G. 1977. Observations on the morphology and behavior of the
Enicocephalidae (Hemiptera). Ent. News 88: 105-10.
LAST, H. 1975. Swarming of Helophorus brevipalpis Bedel (Col., Hydro-
philidae). Ent. Mon. Mag. 111: 127-8.
LLOYD, J. E. 1973. Fireflies of Melanesia: Bioluminescence, mating behavior,
and synchronous flashing (Coleoptera: Lampyridae). Environ. Ent.
2: 991-1008.
--- 1979. Mating behavior and natural selection. Florida Ent. 62: 17-34.
MACNEILL, C. D. 1962. Observations on the flight behavior of an ascalaphid
of the genus Ululodes (Neuroptera: Ascalaphidae). Pan-Pacific Ent.
38: 186-8.
MAYNARD SMITH, J. 1978. The evolution of sex. Cambridge University
Press, Cambridge, England.
MCALPINE, J. F., AND D. D. MUNROE. 1968. Swarming of Lonchaeid flies
and other insects, with descriptions of four new species of Lonchaeidae
(Diptera). Canadian Ent. 100: 1154-78.
McGOWAN, L. M. 1975. The occurrence and behaviour of adult Chaoborus
and Procladius (Diptera: Nematocera) from Lake George, Uganda.
Zool. J. Linn. Soc. 57: 321-34.

Florida, Entomologist 64 (1)

March, 1981

MICHENER, C. D. 1974. The social behavior of the bees: A comparative study.
Belknap Press, Cambridge, MA.
MILINSKI, M. 1977. Do all members of a swarm suffer the same predation?
Z. Tierpsychol. 45: 373-88.
MILLER, L. A., AND R. J. MCCLANAHAN. 1960. Life-history of the seed-corn
maggot, Hylemya cilicrura (Rond.) and of H. liturata (Mg.)
(Diptera: Anthomyiidae) in southwest Ontario. Canadian Ent. 92:
MOORHOUSE, D. E., AND M. H. COLBO. 1973. On the swarming of Austro-
simulium pestilens MacKerras and MacKerras (Diptera: Simuliidae).
J. Australian Ent. Soc. 12: 127-30.
MORI, S., AND K. MATUTANI. 1953. Daily swarming of some caddisfly adults
and their habitat segregations. Zool. Mag. (Tokyo) 62: 191-8.
MOSER, J. C. 1967. Mating activities of Atta texana (Hymenoptera,
Formicidae). Insectes Sociaux 14: 295-312.
NEEDHAM, J., J. TRAVER, AND Y. C. Hsu. 1935. The biology of mayflies with
a systematic account of North American species. Comstock, Ithaca.
NEVILLE, A. C. 1959. Mating swarm in Enallagma cyathigerum (Charp.)
(Odon., Coenagriidae). Ent. Mon. Mag. 95: 181.
NIELSEN, E. T. 1962. Contributions to the ethology of Glyptotendipes
(Phytotendipes) paripes Edwards. Oikos 13: 48-75.
AND H. GREVE. 1950. Studies on the swarming habits of mosquitoes
and other Nematocera. Bull. Ent. Res. 41: 227-58.
AND J. S. HAEGER. 1960. Swarming and mating in mosquitoes. Misc.
Publ. Ent. Soc. Am. 1: 72-95.
AND A. T. NIELSEN. 1953. Field observations on the habits of Aedes
taeniorhynchus. Ecology 34: 141-56.
AND H. T. NIELSEN. 1958. Observations on mosquitoes in Irag. Ent.
Medd. 28: 282-321.
AND 1963. The swarming habits of some Danish mosquitoes.
Ent. Medd. 32: 99-170.
NIELSEN, H. T. 1965. Swarming and some other habits of Mansonia pertur-
bans and Psorophora ferox (Diptera: Culicidae). Behaviour 24: 67-89.
AND E. T. NIELSEN. 1962. Swarming of mosquitoes: Laboratory ex-
periments under controlled conditions. Ent. Exp. Appl. 5: 14-32.
NOWELL, W. R. 1951. The dipterous family Dixidae in western North Amer-
ica (Insecta: Diptera). Microentomology 16: 187-270.
OKUBO, A., H. C. CHIANG, AND C. C. EBBESMEYER. 1977. Acceleration field of
individual midges, Anarete pritchardi (Diptera: Cecidomyiidae),
within a swarm. Canadian Ent. 109: 149-56.
OLIVER, D. R. 1971. Life history of the Chironomidae. Annu. Rev. Ent. 16:
OTTE, D. 1974. Effects and functions in the evolution of signaling systems.
Annu. Rev. Ecol. Syst. 5: 385-417.
PAASIVIRTA, L. 1972. Taxonomy, ecology and swarming behaviour of Tany-
tarsus gracilentus Holmgr. (Diptera, Chironomidae) in Valassaaret,
Gulf of Bothnia, Finland. Ann. Zool. Fenn. 9: 255-64.
PARKER, G. A. 1978. Evolution of competitive mate searching. Annu. Rev.
Ent. 23: 173-96.
PETERS, W. L., AND J. G. PETERS. 1977. Adult life and emergence of Dolania
americana in northwestern Florida (Ephemeroptera: Behningiidae).
Int. Rev. Gesamten Hydrobiol. 62: 409-38.
PETERSON, B. V. 1959. Observations on the mating, feeding and oviposition
of some Utah species of black flies (Diptera: Simuliidae). Canadian
Ent. 91: 147-55.
QURAISHI, M. S. 1965. Swarming, mating and density in nature of Anopheles

Insect Behavioral Ecology-'80 Sullivan

stephensi mysorensis. J. Econ. Ent. 55: 821-4.
RAO, T. R., AND P. F. RUSSELL. 1938. Some field observations on the swarm-
ing and pairing of mosquitoes, particularly A. annularis, in south
India. J. Malaria Inst. India 1: 395-403.
RAU, P. 1937. Notes on the courtship and mating of the fly, Ptecticus tri-
vittatus Say. Psyche 44: 141-2.
REID, W. J., JR. 1940. Biology of the seed-corn maggot in the coastal plain
of the south Atlantic states. United States Dept. Agr. Tech. Bull 723.
REISEN, W. K., Y. ASLAM, AND T. F. SIDDIQUI. 1977. Observations on the
swarming and mating of some Pakistan mosquitoes in nature. Ann.
Ent. Soc. America 70: 988-95.
RIEK, E. F. 1974. The Australian moth-lacewings (Neuroptera: Ithonidae).
J. Australian Ent. Soc. 13: 37-54.
ROTH, L. M. 1948. A study of mosquito behavior: An experimental labora-
tory study of the sexual behavior of Aedes aegypti (Linnaeus). Am.
Midl. Nat. 40: 265-352.
SAVOLAINEN, E. 1978. Swarming in Ephemeroptera: The mechanism of
swarming and the effects of illumination and weather. Ann. Zool.
Fenn. 15: 17-52.
AND J. SYRJAMAKI. 1971. Swarming and mating of Erioptera gemina
Tjeder (Dipt., Limoniidae). Ann. Ent. Fenn. 37: 79-85.
SCHUH, T. 1970. Swarming in an undescribed Enicocephalid (Hemiptera)
from Costa Rica. Ent. News. 81: 66-8.
SCOTT, J. A. 1968. Hilltopping as a mating mechanism to aid the survival of
low density species. J. Res. Lepid. 7: 191-204.
SERVICE, M. W. 1971. Observations on swarming of the adults of Simulium
(Simulium) austeni Edwards (Dipt., Simuliidae). Ent. Mon. Mag.
107: 167-8.
SHEMANCHUK, J. A., AND J. WEINTRAUB. 1961. Observations on the biting
and swarming of snipe flies (Diptera: Symphoromyia) in the foot-
hills of southern Alberta. Mosq. News 21: 238-43.
SHIELDS, O. 1967. Hilltopping. An ecological study of summit congregation
behavior of butterflies on a southern California hill. J. Res. Lepid. 6:
SIVINSKI, J. 1980. Sexual selection and insect sperm. Florida Ent. 63: 99-
SOLEM, J. 0. 1978. Swarming and habitat segregation in the family
Leptoceridae (Trichoptera). Norweigan J. Ent. 25: 145-8.
SOUTHWOOD, T. R. E. 1957. Observations on swarming in Braconidae
(Hymenoptera) and Coniopterygidae (Neuroptera). Proc. R. Ent.
Soc. London Ser. A, Gen. Ent. 32: 80-2.
SPIETH, H. T. 1940. Studies on the biology of the Ephemeroptera. II. The
nuptial flight. J. New York Ent. Soc. 48: 379-90.
STEYSKAL, G. 1941. A curious habit of an empidid fly. Bull. Brooklyn Ent.
Soc. 36: 117.
STRANG, G. E. 1970. A study of honey bee drone attraction in the mating
response. J. Econ. Ent. 63: 641-5.
STUCKENBERG, B. R. 1958. Taxonomic and morphological studies on the
genus Paulianina Alexander (Diptera: Blepharoceridae). Mem. Inst.
Scient. Madagascar (Ser. E) 10: 97-198. (Not seen, in McAlpine and
Munroe 1968)
SUDD, J. H. 1967. An introduction to the behavior of ants. St. Martins Press,
New York.
SYRJAMAKI, J. 1964. Swarming and mating behavior of Allochironomus
crassiforceps Kieff. (Dipt., Chironomidae). Ann. Zool. Fenn. 1: 125-45.
-- 1965. Laboratory studies on the swarming behaviour of Chironomus

Florida Entomologist 64 (1)

March, 1981

strenzkei Fittkau in litt. (Dipt., Chironomidae). I. Mechanism of
swarming and mating. Ann. Zool. Fenn. 2: 145-52.
.1966. Dusk swarming of Chironomus pseudothummi Strenzke
(Dipt., Chironomidae). Ann. Zool. Fenn. 3: 20-8.
1967. Laboratory studies on the swarming behaviour of Chironomus
strenzkei Fittkau in litt. (Dipt., Chironomidae). II. Daily rhythm of
swarming during artificially changed light intensities. Ann. Zool.
Fenn. 4: 19-28.
S1968. A peculiar swarming mechanism of an Arctic chironomid
(Diptera) at Spitsbergen. Ann. Zool. Fenn. 5: 151-2.
S1976. The mystery of the missing females in connexion with male
swarming of Blacus ruficornis Nees (Hym., Braconidae). Ann. Ent.
Fenn. 42: 66-8.
THEW, T. B. 1958. Studies on the mating flight of the Ephemeroptera. I.
The mating flights of Ephoron album (Say) and Stenonema canadense
(Walker). Florida Ent. 41: 9-12.
THOMPSON, P. H. 1967. Swarming behavior of Mansonia perturbans
(Walker). American Midi. Nat. 77: 533-5.
THORNHILL, R. 1977. The comparative predatory and sexual behavior of
hanging-flies (Mecoptera: Bittacidae). Misc. Publ. Mus. Zool. Univ.
Michigan No. 69.
1979. Adaptive female-mimicking behavior in a scorpion-fly. Science
205: 412-4.
.1980a. Sexual selection within mating swarms of the lovebug,
Plecia nearctica (Diptera: Bibionidae). Anim. Behav. 28: 405-12.
1980b. Competitive, charming males and choosy females: Was Dar-
win correct? Florida Ent. 63: 5-30.
TITMUS, G. 1980. Distribution and behaviour of adult chironomidae (Dipt.).
Ent. Mon. Mag. 115: 145-8.
TJEDER, B. 1967. Neuroptera-Planipennia, The lace-wings of southern
Africa. 6. Family Nemopteridae. Ch. 13 in B. Hanstrom, P. Brinck,
and G. Rudebeck, eds. South African Animal Life, Vol. 13. Berlingska
Boktryckeriet, Lund, Sweden.
TOWNES, H., AND M. TOWNES. 1959. Ichneumon-flies of America north of
Mexico: 1. Subfamily Metopinae. United States Nat. Mus. Bull. No.
TROST, L. M. W., AND L. BERNER. 1963. The biology of Callibaetis floridanus
Banks (Ephemeroptera: Baetidae). Florida Ent. 46: 285-99.
TURNER, W. J. 1974. A revision of the genus Symphoromyia Frauenfeld
(Diptera: Rhagionidae). I. Introduction. Subgenera and species-
groups. Review of biology. Canadian Ent. 106: 851-68.
VAN ACHTERBERG, C. 1977. The function of swarming in Blacus species
(Hymenoptera, Braconidae, Helconinae). Ent. Ber. (Amst.) 37: 151-2.
VAN SOMERSEN, G. R. CUNNINGHAM. 1975. A further note on swarming of
male mosquitoes and other Nematocera in Kenya. Ent. Mon. Mag.
111: 147-60.
WEBER, N. A. 1963. Ten kilometers of swarms of an ant. Proc. Ent. Soc.
Washington 65: 109.
WHARTON, R. H. 1953. The habits of adult mosquitoes in Malaya. IV.
Swarming of Anophelines in nature. Ann. Trop. Med. Parasitol. 47:
WILLIAMS, G. C. 1975. Sex and evolution. Princton Univ. Press, NJ.
WILSON, E. 0. 1957. The organization of a nuptial flight of the ant Pheidole
sitarches Wheeler. Psyche 64: 46-50.
WRIGHT, J. E., K. D. KAPPUS, AND C. E. VENARD. 1966. Swarming and mat-
ing behavior in laboratory colonies of Aedes triseriatus (Diptera:

Insect Behavioral Ecology-'80 Sullivan 65

Culicidae). Ann. Ent. Soc. America 59: 1110-2.
YOUNG, R. M. 1969. Flight observations on a midwinter breeding flight of
Diamesa arctica (Diptera: Chironomidae). Ann. Ent. Soc. America
62: 1204.

Florida Entomologist 64 (1)

March, 1981



How are the movement patterns of animals adapted for finding the targets
of their search? This question has been central to several recent studies
dealing with the movement patterns shown by animals as they search for
various resources, usually either food or mates. Examples of such work are
the studies of Smith (1974) on food searching thrushes and Waddington
(1980) on the flight patterns of nectar seeking bees.
The same principal question was also asked in the present study of mate-
searching fireflies. However, the flash communication system of fireflies
makes them unique among the animals studied thus far. This signaling sys-
tem is not only an integral part of a firefly's search behavior, but also greatly
facilitates the study of their search movement. The present study deals
specifically with the fireflies of the Photinus collustrans complex (Coleoptera:
Lampyridae), P. collustrans and the closely related P. tanytoxus.
In both species, males fly in search of flightless females that become re-
sponsive to males only during a brief period each night, from May to Novem-
ber in north-central Florida. Both species are found in relatively open
habitats such as fields and lawns, and are occasionally found inhabiting the
same site, as in this study. However, the two are temporally isolated by
having non-overlapping activity periods.
Photinus collustrans begins its activity abruptly between 11 and 26 min.
after sunset. Searching males fly over the habitat emitting a 0.3 sec. flash
every 2.2 sec. at 220C. with very low variance in interval for any one male
(at constant temperature). Females wait on the ground near their burrows
and upon seeing a male flash, wait about 0.5 sec. ("female delay time") and
then emit a single response flash. Upon seeing a response, the male quickly
flies down to the female and mates with her. Males actively search for females
for about 15 min. and then abruptly end their activity (Lloyd 1966).
After a brief hiatus, males of the sibling species P. tanytoxus become
active. The signaling and mating system of P. tanytoxus is nearly identical
to that of P. collustrans, with the primary difference being a longer delay
time (1.5 sec.) for P. tanytoxus females. Male activity lasts for 35 to 70 min.,
apparently depending upon the size of the deme (Lloyd 1966).
This description of the mating behavior of these fireflies brings to light
several features that make this group an ideal system for the study of
searching behavior. The open habitats in which these species search allow
for easy tracking of individuals in flight. Also, the relative immobility of
females and their "ground-hugging" distribution permits us to assume that
these targets for which males search are both non-moving and distributed
in just two dimensions. Thus, the most important component of male move-
ment is parallel to the ground where females are found, and the vertical

*Greg Adams is currently a graduate student and research assistant at the University of
Florida where he is completing his Master's degree in Entomology. His research has focused
on the mate-searching behavior of Photinus fireflies and the use of these insects as model
systems for testing the concepts and theory of searching behavior in general.
Current address: Dept. of Entomology and Nematology, University of Florida, Gainesville,
FL 32611. Florida Agricultural Experiment Station Journal Series No. 2818.

Insect Behavioral Ecology-'80 Adams

dimension of their flight can be ignored. This greatly simplifies the measure-
ment and analysis of male flight paths.
In addition, the constant flash rate of individual males gives them the
appearance of moving "beacons" with flashes temporally far enough apart
to allow marking each one, but spatially close enough that a string of suc-
cessive flashes gives a good approximation to the insect's complete movement
path. A male's flashes facilitate the actual analysis of their search paths by
naturally dividing up the path into approximately linear segments, each
completed in the same amount of time. This division process is the usual first
step in the analysis of any continuous path of movement, but, as Pyke (1978)
notes, is a rather arbitrary process with most animals. With these two fire-
flies the division is both automatic and non-arbitrary.
Another important feature of the system studied here is the discrete,
brief activity periods of both species. In P. collustrans males are essentially
"trapped" into searching for females within a 15 min. period each night,
corresponding to the time when ambient light is low enough to allow effective
signaling, but before the sibling species P. tanytoxus becomes active (Lloyd
1979). Within this brief period, male competition for females is intense, and
the chance of an average P. collustrans male finding a female within any one
night has been calculated to be about 0.139 or less (Lloyd 1979). Because of
this intense competition and the obviously close correlation that ability to
find females must have with long term reproductive success, search move-
ment in these fireflies may be one of the most clear-cut examples of search
specific adaptation thus far studied.
A final distinguishing feature of fireflies is the process by which they
search. With the majority of animals, the discovery of food or mates de-
pends mainly on the one-way process of encountering one of these items
within the animal's perceptual field. With fireflies, a two-way process is
involved. Males can normally only detect females by their response flashes,
but a female will only respond if she sees a male flash. Thus, the primary
task of the male firefly is to "distribute" his flashes in a way that maximizes
the chances that females will detect him (see also Otte and Smiley 1977).
Consequently, it is primarily the spatial distribution of a male's flashes over
the habitat that determines his mating success. Marking these flash distribu-
tions thus became the initial goal in this study.
In order to mark flash distributions of fireflies, individual males were
first singled out and at the instant the male emitted the first flash to be
marked, a stopwatch was started. At this time two-inch, red styrofoam balls
were dropped beneath each flash, and at a slight delay relative to the flash,
so the searching firefly would not be disturbed. Flashes were marked until
the male was lost or the observer ran out of markers, at which time the
stopwatch was stopped and the exact time of night noted. Observations were
made on the flight characteristics (height, erratic maneuvers, etc.) of the
male, environmental conditions in which 'he flew (calm, breeze, etc.) and a
qualitative judgement of male density along his flight path. Males were fol-
lowed throughout the entire activity period in order to sample male search
under the entire range of nightly conditions to which they were exposed.
The next day each trail was quantified by measuring the distances be-
tween consecutive markers and the angles formed by consecutive, overlapping
triplets of markers. This gave a complete description of the 2-dimensional

Florida Entomologist 64 (1)

flash distribution for a male's flight path. Occurrence times for flight path
segments were designated as the time when the midpoint of a trail was
reached. Because the onset of male activity seems to be triggered at a certain
level of ambient light (Lloyd 1966, and in prep.), these midpoint times were
expressed as "minutes after the onset of P. collustrans activity" to eliminate
some of the between night variance in light caused by cloud cover or moon-
light. A total of 70 flight paths (P. collustrans and P. tanytoxus combined)
were compared with respect to their directionality, a measure of the correla-
tion between directions of successive turns in a movement path1.
The relevance of directionality to search movement is that an animal with
a highly directional path has a lower probability of overlapping areas al-
ready searched than a similar animal with lower directionality (assuming
comparable move and memory lengths for both). Thus, it will most often
have a greater search efficiency than the latter. In fact, for an animal in an
unbounded search space, the optimal search path is a straight line, and, thus,
a directionality of 1.0 (Pyke 1978). In reality, the directionality which an
animal should use to maximize its search efficiency depends on numerous
factors, as stressed by Pyke (1978).
Using computer simulation of animal movement Pyke predicted the
optimal directionalities expected for model organisms he termed "harvesters."
Harvesters were defined as those animals whose search movement at any
point depends only on previous movement and not upon sensory detection of
food or mates at a distance. He predicted that if animals search areas of
large size relative to their mean move length, and have realistic turning be-
havior at the boundaries of their search areas, then most of them should
have search directionalities between 0.8 and 1.0.
Upon reviewing the directionalities of all animals for which data were
available (goldfish, thrushes, finch flocks, bumble bees, etc.), Pyke found
that the majority had significantly lower directionalities than predicted. The
known values ranged from 0.03 to 0.8 with most being significantly lower
than the maximum. He concluded that searching animals usually do not act
like harvesters, but actually move in large part with respect to sensory de-
tection of food or mates at a distance. This results in low directionality paths
which reflect the distribution of resources, instead of in high directionality
trails which minimize overlapping ground already searched. Since a male
firefly can usually only detect females by their responses to the male's own
flash, they do not move with respect to individual females over the great
majority of their search paths. Fireflies thus closely approach a harvesting
strategy in their search behavior and would be expected to have highly
directional search paths.
In this study, the range in directionality for P. collustrans was from
'0.560 to 0.865 with a mean of 0.7593, while the P. tanytoxus range was from
0.641 to 0.899, with a mean of 0.786. These fireflies thus have some of the
highest directionalities known, with matly being in the range predicted for
harvesting animals. But why is there such a range in directionality for males
that are searching for the same set of females in the same locality?
When "directionality" was plotted against the occurrence times for male
flight paths, a wide scatter of points was obtained. However, when lines are
drawn connecting the points of males I followed within the same night, in
almost every case there is a continuous increase in directionality from the

March, 1981

Insect Behavioral Ecology-'80 Adams

first to the last male. This is indicated in Fig. 1 by the positively sloping
lines, each indicating a separate night. Combining these results with other
data yielded a total of 14 of 17 comparisons in which there was an increase
in directionality from the early to the late-followed male. This trend is sig-
nificantly different from random at the 0.01 level, and held for 11 of 14 P.
collustrans and all (N=3) P. tanytoxus comparisons. In cases where direc-
tionality did not increase between consecutive males, the decreases were small
(0.790 to 0.770, 0.816 to 0.803, 0.794 to 0.777).
Examples of flight paths for males followed at different times on the same
night are illustrated in Fig. 2. Notice that the change in directionality from
male 1 to male 2 consists of a loss of erratic turns and an overall increase in
the correlation between directions of consecutive inter-flash movements. This
pattern of change is typical for both species and results in smoother, more
regular flight paths in later males.
Why don't males search with a single optimal value of directionality
during the entire activity period? Although the fireflies in this study cannot
directly detect female location in advance of receiving a response flash, they
should use any detectable environmental "cues" to female location, and move
preferentially toward these spots. Even in the homogenous-appearing field
habitat of these species, there is much subtle variation due to grass clumps,



-5 0 5 10 15 20 25 30
Fig. 1. Within-night trends in directionality between successively followed
P. collustrans males as they searched for females. Each point represents the
directionality of an individual male plotted against the time at which the
male was tracked. Lines connect points for males followed on the same night.

Florida Entomologist 64 (1)

March, 1981



TIME= 20.65min. D= .897
TIME= 32.38min.

Fig. 2. Flight paths for an "early" and a "late" P. tanytoxus male. Trails
begin at the left, with each dot representing a flash. Flight durations: Male
1, 41.9 sec.; Male 2, 74.6 sec. "D" = Directionality (see text). "Time" =
minutes after onset of P. collustrans activity. Bar = 5m.

bare patches, and bushes, which might be positively associated with female
location. Preferential male movement toward these scattered spots would
result in a larger amount of turning and thus lower directionality in male
search paths. Similar behavior was noted in P. pyralis by Lloyd (1966). This
would explain the relatively low directionality of early-followed males, but
not the nightly tendency of males to gradually straighten their search paths.
However, P. collustrans males begin their search activity at dusk, when
ambient light is decreasing at a very rapid rate. Males thus search under a
range of conditions from relatively light, when substrate variation could be
easily detected, to near darkness when low ambient light would preclude
males from detecting and moving with respect to these terrestrial cues to
females. The increasing directionality of these mate-searching fireflies may
thus be due, in part, to the continually decreasing visibility of terrestrial
cues to females.
Early in the activity period males may act as true "searchers," in Pyke's
(1978) terminology, detecting and moving toward scattered points in the
field that appear most likely to harbor females. In fact, early males of both
species do look as if they move and direct their flashes toward certain spots
on the ground, and often hover momentarily above these spots. Gradually as
the field darkens, males will become less able to distinguish the field's subtle
physical features and behave more like true harvesters, moving with the
increasingly directional trails predicted by theory.
The picture becomes more complex when the species are compared directly.
Comparisons of P. collustrans and P. tanytoxus males that fly at about the
same time of night indicate that something more than ambient light affects
male directionality. Given the nearly identical flash patterns and search
areas of both species, early P. tanytoxus males would be expected to begin
searching with directionalities typical of late P. collustrans males that search
under essentially the same conditions of ambient light. Instead, there are 9

Insect Behavioral Ecology-'80 Adams

cases of a drop in directionality on moving from the last P. collustrans to
the first P. tanytoxus followed in any one night. Assuming that both species
have roughly the same visual capabilities2, there must be some other factor
affecting their searching behavior.
One condition that is probably not the same for late P. collustrans and
early P. tanytoxus is female availability. Since responding females are con-
tinually being found and mated throughout any one activity period, female
density may vary from a peak near the beginning of male activity to a low
near the end of the activity period. If so, late P. collustrans are searching
for more thinly distributed females than P. tanytoxus males searching at the
same time.
There is some evidence that these fireflies search according to some rough
"expectation" of female density in the search area, based on the time of
night or the numbers of conspecific males searching an area. In peripheral
areas where there are few other searching males, or late in the activity
period when most females have been mated or are no longer responding,
males appear to search with highly directional trails that would maximize
the chances of finding widely scattered, isolated females.
Bond (1980) has given evidence that prey searching lacewing larvae
move according to "inferred differences" in prey distribution, based on their
level of hunger. Larvae deprived of food for longer periods of time moved
more quickly and had a lower overall turning rate than low deprivation
larvae, which resulted in more directional search paths. It was suggested that
these larvae moved according to their expectation of a low density of prey
patches based on their high level of hunger. Evidence that inferred female
density may affect the search paths of these fireflies was obtained on nights
late in the season when few males were flying.
A good example of a male searching under conditions of low probable
female density is the male followed on 20 October 1979. This was the last
night of the season on which trails were marked, and very near the end of
the season for this firefly. Male 1 of this date, though searching during the
first 4 min. of male activity, had a directionality of 0.823, which was higher
than any other P. collustrans followed this early on other nights.
The two P. tanytoxus followed on 7 September 1980 had the highest direc-
tionalities for any two fireflies in the same night, being 0.879 and 0.899 for
males 1 and 2, respectively. Again, both males were flying in areas with no
other males nearby, and on a night of very low overall male density, perhaps
indicative of a low female density.
The males of 20 October 1979 are indicated in Fig. 1 by the line of flat
positive slope, indicative of the small increment in directionality that oc-
curred between the early and late male on that night. However, in spite of
the evidence that these fireflies can modify their search characteristics in
response to expected female density, there are no data for female distribu-
tion and the adaptive significance of the, lower directionalities for early P.
tanytoxus is still unclear.

Mate searching males of both P. collustrans and P. tanytoxus show a
common pattern of change in their search paths throughout an activity
period. This change entails a progressive decrease in the frequency and size

Florida Entomologist 64 (1)

of turns along the flight path, and results in males having progressively
higher directionalities. This straightening out of male flight paths through
the night probably results from selection on males to maintain a high search-
ing efficiency under a wide range of ambient light conditions. In addition,
the directionality of a male's search path may also be affected by his expecta-
tion of female density in an area.

Sincere thanks go to Dr. James E. Lloyd for the information, criticism,
and general support provided during this research. I would also like to
thank Dr. Jane Brockmann, John Sivinski, and Tim Forrest for their helpful
comments on the manuscript. Thanks also to Vicki Turner for help with com-
puter programs and Thelma Carlysle for photographic assistance. This work
is based upon work supported in part by the National Science Foundation
under Grant No. DEB-7921744 to J. E. Lloyd. Other financial support was
obtained from Grant No. 572 of the Alexander Bache Fund, National Acad-
emy of Sciences.

1Pyke (1977) has referred to this value as the "average vector component
of successive movements in the straight ahead direction." The value varies
from 0.0 to 1.0, where zero denotes no correlation between successive turns,
i.e. random movement, and a value of one denotes perfect correlation between
all successive turns, i.e. a straight line. Directionality itself does not give a
measure of the overall linearity of a movement path (except when = 1.0), as
this depends on the sequential patterning of turns within a path. It does give
a measure of both the amount and degree of turning in a movement path.
Values for directionality in this study were obtained by means of a computer
program based on Cody's (1971) description of its calculation.
2That this assumption is not completely warranted is suggested by the
recent finding that the spectral sensitivity (and flash color) of early-active
fireflies differs from that of late active species-the eyes of the latter being
like those of dark-adapted (nonfirefly) insects (Lall et al. 1980).

BOND, A. B. 1980. Optimal foraging in a uniform habitat: the search
mechanism of the green lacewing. Anim. Behav. 28: 10-9.
CODY, M. L. 1971. Finch flocks in the Mojave Desert. Theoret. Pop. Biol. 2:
LALL, A. B., H. H. SELIGER, W. H. BIGGLELY, AND J. E. LLOYD. 1980. Ecology
of colors of firefly bioluminescence. Science 210: 560-2.
LLOYD, J. E. 1966. Studies on the flash communication system in Photinus
fireflies. Misc. Pub. Mus. Zool., Univ. of Michigan No. 130. 95 p.
.1979. Sexual selection in luminescent beetles. Pages 293-342 in
Sexual selection and reproductive competition in insects, M. S. Blum,
and N. A. Blum, eds. Academic Press, New York, N.Y.
OTTE, D., AND J. SMILEY. 1977. Synchrony in Texas fireflies with a considera-
tion of male interaction models. Biol. Behav. 2: 143-58.
PYKE, G. H. 1977. Optimal foraging. Quart. Rev. Biol. 52: 138-54.
1978. Are animals efficient harvesters? Anim. Behav. 26: 241-50.
SMITH, J. M. N. 1974. The food searching behavior of two European

March, 1981

Insect Behavioral Ecology-'80 Adams 73

thrushes. II. The adaptiveness of the search patterns. Behaviour 49:
WADDINGTON, K. D. 1980. Flight patterns of foraging bees relative to the
density of artificial flowers and distribution of nectar. Oecologia 44:

Florida Entomologist 64 (1)

March, 1981



Polistes wasps are ideal subjects for studies on the evolution of social
organization because all females beginning nests in the spring are mated and
capable of functioning as egg-laying queens on their own nests, but some do
not. Why they do not is the topic of this paper, and my research investigates
the evolution of social behavior.
Evolutionary theory predicts that individuals behave in ways that result
in their passing as many copies of their genes on to the next generation as
possible. Classical natural selection models used number of offspring as the
measure of reproductive success. But Hamilton (1964 a,b) recognized that
under certain circumstances it was possible to pass on more genes by raising
relatives that were not offspring. For potential mothers to pass on more
genes by raising the progeny of sisters, aunts or other relatives, it is neces-
sary to raise greater numbers of them, to make up for the loss in relatedness.
In Polistes wasps abdicating mothers-potential queens-do this.
In the spring Polistes annularis females form nesting groups in which
some females lay more eggs than others. Marking studies have shown that
the females in these groups were born on the same nest the preceding
autumn (Strassmann 1979). Therefore, they may be sisters, sharing 3/4 of
their genes (because Hymenoptera have haploid males'), if their mother (the
egg layer in the autumn nest) was mated to only one male (Metcalf and
Whitt 1977, Metcalf, 1980)2. On a spring nest a female that is helping her
sister is raising nieces (3/8 shared genes) who are more distantly related
than offspring (1/2 shared genes3. According to Hamilton's theory of kin
selection, for an aunt to pass on more genes this way, she must raise at least
1/3 more nieces than offspring, had she exercised her "mother option." I have
found that a female alone on a nest raises about three offspring, while each
female on a nest with 2-7 females, raises about five Strassman, in prep.).
This is enough more nieces to account for females joining groups and raising
nieces, instead of becoming mothers and raising their own offspring.4

The female that lays eggs in the nest gains most, but how far is the
(dominant) queen willing to escalate a fight to maintain absolute control
and lay all of the eggs? The queen has the most to lose if subordinates leave
(see appendix note 5 for the concept of offspring-equivalents, the measure
central to these considerations and comparisons), and she is the one that
controls subordinate reproduction. Evolutionary theory predicts that she will
allow some egg-laying by subordinates.

*Joan E. Strassmann is an Assistant Professor of Biology at Rice University. She studies
the evolutionary biology of Polistes wasps. Current address: Biology Department, Rice Uni-
versity, Houston, TX 77001.
'Superscripts refer to notes in the appendix.

Insect Behavioral Ecology-'80 Strassmann

What proportion of the eggs should a queen allow the subordinates to lay?
An equilibrium will be reached when the queen and the subordinates lose
equal amounts by leaving the group (Parker 1974). If all females are able to
raise identical numbers of offspring, were they to form individual nests, then
all are predicted to lay equal numbers of eggs in the communal nest, unless
the advantage of being in a group (measured in numbers of offspring raised)
is reduced when all females lay because the efficiency that results from a
division of labor is lost. However, it is likely that the queen is somewhat
more physically fit than her subordinates, (since she won the fights for
queenship), and therefore could probably raise more young were she alone
than could her subordinates were they alone.
In this paper, I report and discuss the development of the interactions of
females in a group, from nest initiation (in March) through the maturation
into adult workers of the first eggs laid, to determine exactly how foundresses
divide up egg laying and nest chores and what changes occur as spring pro-
gresses. The question asked is: on nests with more than one foundress do
subordinates have opportunities for direct reproduction, as well as the op-
portunity to raise relatives? Of concern are the females born and mated in
the previous autumn, that hibernated, and then in spring began nests.
Worker females born in spring are not considered and their behavior is not
included. Males do not participate to any extent, and there are few of them
in the spring, so they are not part of this story either.


The study site was a limestone cliff overlooking a reservoir 48 km (30
miles) west of Austin, Texas (Fig. 1). The nests occurred right on the lime-
stone cliffs that face west where they are protected from sun until 1400 hrs.
I marked females on the nests they emerged from in the autumns of
1976-1979 (marks were individually-coded enamel spots). Wasps were ob-
served for 2-26 days and for 6-20 hours each day.6 Behaviors were recorded
as classified below. Two nests contained only eggs, seven had eggs and larvae,
five had eggs, larvae and pupae, and four had eggs, larvae, pupae and work-
ers. During observations on two nests containing eggs and larvae, the queens
were dethroned-one by a subordinate and one by a queen from a neighboring
nest that had been destroyed (Table 1).
Ideally in a study of the behavior of animals in groups, all behaviors are
recorded-what each individual does at any given time, the duration of each
behavior, and where it occurred-for 24 hours a day. This was not possible
in this study, so I used a number of conventions: (1) It was assumed that a
daytime sample of behavior that lasted 6 to 20 hours was representative. (2)
All behaviors not directly involving another individual were recorded with-
out noting where the others were at the time. Interactive behaviors were
recorded as pairs, one female the subject, the other the object. If a fe-
male arrived at the nest and three others took some caterpillar meat from
her, these were recorded as three separate exchanges. (3) Behavior durations
were not timed, though total observation time was noted. Most behaviors took
little time and did not vary much in duration. Therefore, counting the times
an individual performed a given behavior allowed a comparison of nestmates
and females on different nests.

Florida Entomologist 64(1)

%~-~.s- 8. 9
-- ?
I )

-rl~b ~r
I- "~r%


Fig. 1. Limestone cliffs where Polistes annularis nests were studied.

Interactions classified as aggressive ranged from touching with antennae,
to chewing or climbing on another while chewing on her and attempting to
sting her.7 Chewing was by far the most common aggressive behavior. Since
there were no systematic differences in which of these aggressive behaviors

March, 1981











srIeuL.ulo( uo
aouuipaoqnSg q
uoissaa2.t8v %

uaanZ) Aq
nuon Xq

uoIssoa2V %0

s').ov ahssajS"ll

*noH Jaod

S13V aAisseJS2V

Aq patuToajd
LuLatawS Jo %

'H IIad

q p uaan

s. a %

JO -oN


adoc pooaa
pa0ZUApv IsoN




Insect Behavioral Ecology-'80 Strassmann

C) Cl 11 w0 0 0 LO It 4 <= uz <= 0 l 00 000 w 0 C1:

1- 4,- i L 'O


N NCoS ciM.Ot C0 lioOC=0 000000
COCO 00 N Z.0 M D M 0 CO C M000 L 004-1CoLM Cc) tVD 00 0 -0-100 I 00I M
00 1 L-VCi00 00=_

i0(0M000 tCA0 C CO 00 CCO01OC000

0L -00 MMC) 00 O = t O- 0o0 OC,01 1>00CO 0000
r-i I-f I-1
CO L_ IO 1i l Cq N l Vz O ~ 't'^ ~ t C

Cu:1 C CCIZ COq C> a) 00 N 0000 LO 0000 01 Cl 00

1- M M000 MC,0 000 L M 0 V00

M 1-- 00C0 0 0001 000 0LC c0000
i l -1 ;l C L-l 00 ir t 4 Cl rl0 C 0

0 0 0 -0 1H -1 1-1 10

MOOi1 0 O t CcO F
0 U0 000 0 000 F-I
0)0C) - -~ CP qpq : ~mm m :

00 00 C, C0 CO 0O Co Q o C.0 Co t 41 N
-l -l 4rl i N 0 0 00 00 C 0 Cl

CO q Cl (=oo
^-'^''.'-''-'^ ^*^ '' ^.^ a Oa T-T- B "! 3 3 3
-0 -0001 -1 C M 11 -1
-- -_ -1 C5 -.-.00

O000000l0"0, ICI 00COl

>-4 >> o >CL| > - >- >-
1-i r-i M -x _.

0400000 ~
'-04> >' -

M LoM -4 M -L 1L 0 1
LO :)-q O "TL-g o 1 ( V Lt) )

hh$;~ o~h E o 1~dX


' w


C )

1, 4
o dB

44 a) 5 >)

Q 0) (U 0 v3

0 0 C)

* ;^* ^-*

Florida Entomologist 64 (1)

March, 1981

occurred among females, all were combined for this analysis and their total
for each female was used to establish a linear dominance hierarchy.8
Data on 12 behaviors are considered here. Other behaviors occurred but
these 12 cover the important activities including dominance interactions, egg
laying and brood care.
(1) Tailwagging is vibrating the abdomen from side to side across the
nest. Since only dominant females do this it may spread a "dominance
pheromone." (See Lloyd, this Symposium). This behavior is usually per-
formed repeatedly as the female walks across the nest antennating cells
containing eggs. Each tailwag takes less than a second. (2) Check cells is
often performed in conjunction with tailwagging. It is also performed by a
female with a load of pulp in her mandibles as she searches for a cell to en-
large, or for a place to begin a new cell. Cells that are antennated usually
contain eggs and not larvae. The female may be checking for empty cells in
which to lay an egg. (3) To lay an egg, a female inserts her abdomen into
a cell, extrudes her sting, and holds it out of the way. Laying an egg takes
about 2 min. (4) Smearing is a slow-motion version of tailwagging and is
directed to the nest pedicel about 70 percent of the time. Smearing deposits
an ant-repellent chemical on the pedicel (Jeanne 1970). (5) Pedicel nibbling
consists of depositing a secretion on the pedicel, and the back of the nest and
substrate near the pedicel, by moving the mandibles over the surface for 2-3
minutes. When this secretion dries, it is hard and transparent like lacquer.
(6) Building is either adding onto an existing cell, or building a new cell. A
ball of pulp is tamped into place with the mandibles while the female touches
both sides of the wall under construction with her antennae. Building
typically takes about 4 min. Here beginning a new cell is not distinguished
from adding onto a cell. However, all eggs were laid in cells that had just
been built by that female except the listed cases of egg eating, so egg laying
and building a new cell occur at the same frequencies. (7) Arriving with
pulp. Common sources of pulp are weathered wood, and weed stems. The
average length of a trip for pulp is 11 (S.D.=10, N=29) min. and the
median trip length is 9 min. (8) Arriving with caterpillar. Caterpillars are
the principal food of the larvae, and hunting for them takes 30 (S.D.=21,
N=18) min.; the median trip length is 27 min. This is probably the most
difficult and risky task for wasps (since they are most vulnerable to preda-
tors while off the nest walking on vegetation), and involves the longest time
away from the nest. (9) Arrive with nothing visible includes returns when
females bring back nectar or water (which are usually carried in the crop
and therefore not visible), or nothing. Such trips average 19 (S.D.=19,
N=23) min., and the median trip length is 8 min. The difference between
the mean and the median is probably the result of mixing different types
of trips since trips for water which are included here because the water is
not usually visible, usually take very little time. (10) Leave is always by
flight and is usually preceded by a quick walk across the nest face. (11)
Exchange with larvae usually means an adult wasp is feeding a larva, but
larvae will sometimes spit up a drop of liquid for the adult, and it is often
not possible to tell who is feeding whom, unless the wasp just returned to the
nest with caterpillar or just exchanged with someone who did. Therefore,
both behaviors are grouped here. Each exchange takes several seconds. (12)
Fan nest is done by buzzing the wings when standing on the nest. This cools

Insect Behavioral Ecology-'80 Strassmann

the nest by evaporation when water has been deposited on the cell walls.
Fanning is usually done on very hot days when the sun is on the nest, and
lasts from 1-7 min.
Because behavioral data are not normally distributed, the statistical
analyses presented here are nonparametric. Means between two categories
are compared using the Mann-Whitney U test, and Kendall Tau correlations
are used (Siegel 1956). For the sake of space values of Mann-Whitney U's
are not given. Whenever two means are stated to be different, they are sig-
nificant < 0.05 level. All correlations are also significant < 0.05 level.

Females return to their parental nest in the spring, just as they do on
warm days all winter (Strassmann 1979). They then leave the nest and walk
about on the cliff usually within 3 m of the nest and antennate the substrate.
Before new nests are begun, small groups form on the cliff face. Females in
these groups are always from the same parental nest and interact ex-
tensively, chewing and antennating each other, and remaining close together.
When a flying wasp approaches a group all group members spread their
wings. They chase away non-nestmates, but allow nestmates to land. Anten-
nal battles take place within these groups, and I have observed them at no
other time. Females facing each other clash antennae, first those of one
female contact the other's from above, then vice-versa. These confrontations
may go on for several minutes before the females separate. Sometimes they
escalate into grappling and both females rear up on their hind legs, grab at
each other with the first two pairs of legs, and slap antennae. These grouped
females return to hibernacula (protected cracks in the cliff where they
hibernated) at night. After a few days one female in the group begins a nest.
A load of pulp is used for the nest base, another load or two for the pedicel
and then the first cell is built. This females begins as the principal egg layer
in the new nest. In general, females do not switch from one new nest to an-
other, and do not join other nests begun by females born on the same
parental nest (Strassmann, in prep.).
Table 1 presents data on the nests observed in this study. Some of the
nest variables in Table 1 were correlated. Number of females on the nest was
positively correlated with number of eggs laid per hour (r=0.28), number of
tailwags per hour (7=0.37) and number of aggressive acts per hour
(T=0.36). Number of females was negatively correlated with percent eggs
laid by the queen (r=-0.31), percent tailwags by queen (r=-0.47), percent
smears by queen (T=-0.51) and percent aggression by queen (-=-0.53).
Queens control larger nests less absolutely than they control smaller nests.

I ranked females on each nest in "dominance hierarchies" according to
who chewed on (or otherwise attacked) whom.8 It was easy to rank the top
five individuals but toward the lower end of the hierarchy (on nests with
more than five females) individuals are less clearly differentiated, and
neither initiate nor receive much aggression. Some of the time females of
ranks above (>) 5 could not be ranked among themselves according to ag-
gression. A female was assigned a lower numbered rank if she had laid an

Florida Entomologist 64 (1)

egg or tailwagged more. This only involved a few of the 38 percent of all
females of ranks above five. Individuals of ranks 1 through 5 were compared
pairwise to see if females of different ranks also performed different fre-
quencies of other behaviors. Comparisons were done among nests with the
same brood types so changes in behavior over time would not add confusion.
The top ranking female (rank 1), the queen, differed in many behaviors
in pairwise Mann-Whitney U comparisons with her subordinates (ranks
2-5), and will be discussed in more detail below. Females of ranks 2-5 varied
in aggressive rank, but not consistently in frequency of performance of other
behaviors as determined by Mann-Whitney U comparisons; therefore in the
comparisons of subordinate ranks with queens, ranks 2 and higher were com-
bined (Tables 2-3).9 Mean frequency of each behavior is given for queens
(Table 2) and non-queens (Table 3) on nests with different brood stages.
Queens laid more eggs and tailwagged more often than their subordinates
on nests of all stages. Queens checked more cells and were more aggressive
than subordinates on nests with larvae, pupae or workers. Queens smeared
more often on nests with larvae or pupae. No queen ever foraged for cater-
pillars. On nests with workers, subordinate foundresses did not forage for
caterpillars. On nests with larvae and nests with pupae, all caterpillar
foraging was done by subordinate foundresses. On nest 2, four of the eggs
laid were eaten by others. On this nest, all eggs eaten were laid by sub-
ordinates and all egg eating was done by subordinates. On nest 55, the queen
ate one egg laid by a subordinate and replaced it with her own. This is the
only egg eating I observed in this study.




A. B. C. D.
Eggs, Larvae, Pupae, Workers,
N=2 N=6 N=5 N=4
Behavior R S. D. R S. D. R S. D. x S. D.

Tailwags 3.40 0.71 8.36 7.39 8.65 3.98 7.42 3.16
Check cells 4.63 4.77 4.82 4.53 2.88 2.50 2.94 2.16
Lay eggs 0.18 0.04 0.16 0.13 0.14 0.14 0.13 0.16
Smear 3.48 1.87 1.18 0.94 0.59 0.63 0.00 0.00
Nibble pedicel 0.75 0.00 0.19 0.35 0.13 0.22 0.00 0.00
Build 0.78 0.18 0.63 0.32 0.52 0.56 0.10 0.20
Arrive with
pulp 0.53 0.18 0.07 0.13 0.17 0.24 0.00 0.00
Arrive with
caterpillar 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Arrive 0.30 0.07 0.15 0.18 0.14 0.16 0.00 0.00
Leave 0.80 0.21 0.11 0.21 0.28 0.34 0.04 0.08
Feedlarvae 0.00 0.00 2.31 2.27 3.00 3.50 0.38 0.29
Fan 0.30 0.42 0.67 1.59 0.03 0.08 0.03 0.05
Attack wasp 4.00 1.70 5.32 5.29 3.02 1.26 0.57 0.72
Be attacked
by a wasp 0.00 0.00 0.12 0.21 0.06 0.09 0.00 0.00

March, 1981

Insect Behavioral Ecology-'80 Strassmann


The behavior of queens and subordinate foundresses changed as the
season progressed and brood developed from eggs to workers (Tables 2, 3).
Queens were more active on younger nests. They smeared more, pedicel-
nibbled more, built more, arrived with pulp more frequently, arrived and
left more often, and were more aggressive (Table 2). Behaviors that most
distinguished queens from subordinates, such as tailwagging, cell checking
and egg laying did not change significantly over the season. In general,
queens on younger nests were less specialized and did more work than queens
on older nests.
Subordinates laid more eggs on nests with only eggs than they lay on
later nests (Table 3). Subordinates also smeared more, built more, arrived
with pulp more often, and left more frequently on nests with eggs than they
did on nests with other stages of brood. Subordinates worked more on nests
with pupae, and arrived with caterpillars and pulp more often, than when
on nests with larvae. Then, when workers emerged, subordinate foundresses
dramatically decreased many behaviors and did less smearing, pedicel nib-
bling, building, arriving (behaviors 7-9), leaving; and feeding of larvae,
and were involved in less aggression than subordinates on earlier nests.
Kendall T correlations among all behaviors were calculated for females




A. B. C. D.
Eggs, Larvae, Pupae, Workers,
N =5 N=43 N=38 N=16
Behavior X S. D. x S. D. X S. D. R S. D.

Tailwags 0.05 0.09 0.15 0.33 0.61 1.42 0.49 0.66
Check cells 2.28 2.76 0.82 1.16 0.32 0.48 0.87 1.14
Lay eggs 0.04 0.04 0.01 0.03 0.01 0.04 0.00 0.00
Smear 0.85 0.52 0.15 0.20 0.08 0.12 0.02 0.05
Nibble pedicel 0.15 0.14 0.04 0.11 0.07 0.11 0.00 0.00
Build 0.54 0.23 0.22 0.22 0.34 0.30 0.07 0.17
Arrive with
pulp 0.46 0.25 0.06 0.10 0.22 0.16 0.06 0.13
Arrive with
caterpillar 0.00 0.00 0.08 0.11 0.26 0.31 0.06 0.15
Arrive 0.33 0.21 0.31 0.28 0.36 0.39 0.08 0.15
Leave 0.79 0.39 0.36 0.29 0.44 0.44 0.20 0.29
Feed larvae 0.00 0.00 0.70 1.10 2.20 2.36 0.17 0.14
Fan 0.70 1.48 0.44 1.09 0.10 0.30 0.10 0.38
Attack wasp 0.14 0.19 0.35 0.66 0.37 1.11 0.07 0.15
Be attacked
by a wasp 1.74 1.21 1.08 1.66 0.75 0.89 0.21 0.40

Florida Entomologist 64 (1)

on nests of each of the four brood stages, for convenience here named for the
most advanced stage only. Nests with larvae also have eggs, etc. A sig-
nificantly correlated pair of behaviors means a female performing one be-
havior frequently is also likely to perform the other behavior frequently.
Higher values of the variable "rank" indicate a more subordinate female.
Only nests without takeovers were used. Correlations are given in paren-
theses in this order: (1) nests with eggs, (2) nests with larvae, (3) nests
with pupae, (4) nests with workers. Among females on nests of each brood
stage, rank was negatively correlated with tailwagging (-0.79, -0.53, -0.70,
-0.74) ; egg laying (-0.70, -0.46, -0.39, -0.41); aggression (-0.90, -0.65, -0.59,
-0.39); cell checking (-0.41,n.s., -0.59, -0.67, -0.51); smearing (-0.62, -0.38,
-0.37, -0.03,n.s.); and building (-0.58, -0.32, -0.36, -0.22,n.s.). On nests with
larvae and pupae, more subordinate females foraged more for caterpillars
(0.35, 0.60). The conclusion from these results is that highly ranked females
(ranks 1-3) behave more like queens than do subordinate females (ranks
>4). This is not a circular argument since it is who chewed on whom that is
used to rank almost all females. Who chewed on whom is not the same as how
many times any one female chewed on other females. The correlations indi-
cate whether other behaviors vary predictably with rank.
The Mann-Whitney U comparisons between pairs of ranks (other than
1 i.e. queen) disclosed few differences in contrast to these results. Probably
the differences were not large enough to be clear in 2-category comparisons
even though the trends across ranks were significant as seen in the correla-

Reproductive competition is clearly intense among these females, and they
establish a linear dominance hierarchy on the nests. The queen and dominant
females laid more eggs, tailwagged more, checked more cells, and were more
aggressive, while subordinate females foraged more for caterpillars and were
the victims of aggression more often. Queens differed more from all sub-
ordinates than subordinates of different ranks varied among themselves.
Subordinates behaved more like queens on nests with eggs and nests with
workers, and differed most from queens on nests with larvae and pupae. This
is probably because subordinate work was most vital on nests with larvae
and pupae, where subordinate foundresses alone were responsible for raising
the brood.
Queens laid a mean of 80 percent of the 47 eggs laid in the 16 nests with-
out takeovers that were observed 190 hours. On the two nests with takeovers,
the new queens laid only two of the six eggs, so new queens do not immedi-
ately attain control (Table 1). The queen was overthrown on two of 18
nests, once by a subordinate and once, by a group of intruders from a
neighboring nest. Thus subordinates lay eggs when they can, sometimes
overthrow the queen and they perform dangerous work such as caterpillar
foraging only when it is essential to the success of the nest, i.e. before any
workers have emerged. This is consistent with a view of subordinates that
considers direct reproduction very important. Additional chances to take
over the nest also occur later in the season, beyond the period of this study.
When queens die in the presence of workers and subordinate foundresses, a

March, 1981

Insect Behavioral Ecology-'80 Strassmann

subordinate foundress and not a worker always takes over (Strassmann in


Using the data obtained here, calculations of numbers of offspring equiv-
alents can be made for females of different ranks on nests of different sizes.
It is necessary to make several estimates from the data that may not always
hold. First, we assume that foundresses are all related to each other as full
sisters. Each foundress on a multiple foundress nest produced a mean of five
young; on a one-female nest, three young were produced. The queen laid a
mean of 80 percent of eggs. In this study one of the seven nests with larvae
was taken over by a subordinate, and the queen fell then to bottom rank. In
addition, at least 20 percent of all nests lost their queens after the period of
this study (Strassmann, in prep.), so we will assume the second-ranked
female takes over the queenship on a total of 35 percent of the nests, the
queen then dropping to the bottom and everyone else moving up a notch.
Since rank is negatively correlated with egg laying, I will assume that the
second ranked female lays 15 percent of the eggs and the third ranked
female lays 5 percent of the eggs, and more subordinate females lay no eggs.
Using methods outlined in the introduction and detailed in the appendix,
Table 4 gives numbers of offspring equivalents produced on nests with 1, 2, 4,
7 and 15 founding females.10 All females produce more offspring equivalents
than the three they would have produced on a lone nest, but the queen and
the second ranked female do much better than more subordinate females. The
top two females also are attributed with more offspring equivalents, the
more females there are on the nest. Under this model it is easy to see why a
female joins a queen since the joiner will have 4.77 offspring equivalents,
1.77 more than she could have produced alone and only 0.46 less than the
queen produces, supporting the theory of kin selection. Both females will
encourage additional females to join because their inclusive witnesses rise, but
the females ranked = 3 do not produce many more offspring than they would
have produced alone, thus, they have less "incentive" to join large groups.
All of these calculations assume all females on a nest are full sisters.
However, a queen laying 80 percent of the eggs, using 90 percent sperm


Number of
females on Female rank
nest 1 2 3 4 5 6 7

1 3.00 -
2 5.23 4.77 -
4 6.68 5.57 3.96 3.79 -
7 9.11 6.88 3.96 3.79 (Females 5-7 3.75 each)
15 15.61 10.38 3.96 3.79 (Females 5-17 3.75 each)

Florida Entomologist 64 (1)

March, 1981

from one male means 72 percent of all females likely to nest together are
actually full sisters. Direct reproduction becomes much more important in
calculating the offspring equivalents raised by subordinates on nests where
the queen is their half sister (r = 1/4), cousins (r = 3/8), or half cousin
(r = 1/8). Calculations of offspring equivalents produced under these cir-
cumstances are not included in this paper.
A number of testable adjustments in the model may explain the occur-
rence of large groups of foundresses, although testing them is beyond the


Pro- spring
No. off- portion equiv-
% Eggs spring shared alency
laid raised genes factor

A= direct reproductive success
B= indirect reproductive success
C= credit for nieces rank 2 fe-
male raised
D = credit for nieces rank 3 fe-
male raised
E = credit for nieces rank 4 fe-
males & up raised
(where x = total number of

Total offspring equivalents
raised (nest with at least
3 foundresses) :
Rank 2 Female
A= direct reproductive success
B = indirect reproductive success
C= credit for nieces raised by
Total offspring equivalents
Rank 3 Female
A= direct reproductive success
B = indirect reproductive success
C = credit for nieces raised by
Total offspring equivalents
Females of Ranks 4 and Up
B = indirect reproductive success
Total offspring equivalents

= 0.8 (5)
= 0.2 (5)

(1/2) (2)
(3/8) (2)

= 0.75

= 1.06


(x-3) (1.25)

= 7.0 + 1.25 (x-3)

= 0.15 (5) (1/2) (2) =0.75
= 0.85 (5) (3/8) (2) =3.19

= 0.19


= 0.05 (5) (1/2) (2) =0.25
= 0.95 (5) (3/8) (2) =3.56



= 1.00 (5) (3/8) (2) =3.75

= 3.75

Insect Behavioral Ecology-'80 Strcssmann

scope of this paper. (1) Perhaps all parameters of the model are correct, and
low ranking subordinates join and lay no eggs because they are in fact of
much lower physical fitness than the top two females. This could be tested
by comparing size, weight, fat and nitrogen content of females of different
ranks, and by determining whether numbers of offspring produced by lone
foundresses varies according to the lone foundress' size, dry weight, etc. (2)
In this model it is assumed that all nests with more than one foundress pro-
duce five offspring per foundress. Perhaps larger groups of foundresses ac-
tually produce more reproductive offspring, so this number should be in-
creased for larger groups of females. This could be determined more precisely
by counting offspring produced at the end of the season, and not just at the
end of the foundress period. This change may increase the subordinates'
number of offspring equivalents raised. (3) Perhaps takeovers by sub-
ordinates and egg laying by subordinates increases on nests of larger size.
In this study it was only possible to assign an overall level of egg laying by
subordinates because of the small sample sizes. However, queens have less
control on nests with more females, since number of females is negatively
correlated with percent of eggs laid by queen (-0.31), percent tailwags by
queen (-0.47), percent smears by queen (-0.51), and percent aggression by
queen (-0.53), and number of females is positively correlated with number
of eggs laid (0.28). Perhaps increased egg laying by subordinates on larger
nests increases the numbers of offspring produced by subordinates and de-
creases the number attributed to the queen.
Polistes wasps have been extensively studied and inclusive fitness has been
tested using other species, (West Eberhard 1969, Metcalf and Whitt 1977,
Gibo 1978, Noonan 1981, Strassmann 1981). Pardi (1948) first described
dominance hierarchies in Polistes and related high rank to egg laying, a
finding supported by West (1967). Egg laying by subordinates is common in
other Polistes species. Noonan (1981) found subordinates lay enough eggs
so a second-ranked female on a two-female nest raised more offspring equiv-
alents than she could have raised alone. Noonan also found aggression in-
creased when eggs that would be reproductive instead of workers were laid.
Gamboa (1978) found that subordinates helped deter takeovers by non nest-
mates, but that subordinate foundresses were chased off the nest by workers
when workers emerged, a marked contrast to the situation in P. annularis.

Polistes annularis foundresses can be ranked in a dominance hierarchy

Queen [7.00 + 1.25 (x-3)] 0.65 + (3.75) 0.35 = 5.86 + 0.81 (x-3)
Rank 2 female (4.13) 0.65 + [7.0 + 1.25 (x-3)] (0.35) = 5.13 + 0.44 (x-3)
Rank 3 female (3.87) (0.65) + (4.13) 0.35 = 3.96
Rank 4 female (3.75) 0.65 + (3.87) 0.35 = 3.79
Rank 5 female (3.75) 1.00 = 3.75
and up

Florida Entomologist 64(1)

according to aggressive behaviors. Females at the top of the hierarchy lay
more eggs, tailwag more, and smear more while females at the bottom of the
hierarchy forage for caterpillars more. The top-ranked females or queens
differ more from all their subordinates than the subordinates differ among
themselves. Subordinates are most different from queens on nests with larvae
and pupae as the most advanced brood stages. On nests with only eggs,
queens forage and differ little from subordinates. When workers emerge from
the nest, subordinate foundresses stop working. On average queens lay 80
percent of all eggs, and subordinates overthrow queens on about 35 percent
of all nests: thus, highly ranked subordinates have opportunities to lay
eggs. When reproductive success is calculated using the concept of offspring
equivalents, it is found that subordinates produce more offspring equivalents
as subordinates than they would have by nesting alone. The larger the
group of foundresses, who are assumed to be sisters, the greater the number
of offspring equivalents attributed to the queen and the second-ranked fe-
male. These findings support the theory of kin selection.

For help with nest observations, I thank Christine Becker, Paul Bruton,
Doug Crosby, Diana Crowell, Madeline Daigle, Juan Ibarra, Tony Jones, Bob
Matlock, Dana Meyer, Tony Mitchell, Diane Stallings, Christi Stienbarger,
and Richard Thomas. I thank Bill Mueller and Jim Lloyd for comments on
the manuscript. I was supported by NIH training grant 5 T32 GM 07126,
NSF National Needs Postdoctoral Fellowship #SPI-7914902 and NSF
#PCM-8003766. Rice University kindly provided computer funds. I thank
Eve Marsden for outstanding eleventh hour typing.

1Since males are haploid, all the sperm they produce is identical, and
sisters have 100 percent of their father's genes in common. Mothers are
diploid, and so sisters share half of their mother's genes, just as in all other
diploid organisms. Half of a female's genes come from her mother and half
come from her father, so sisters are related by 1 x 1/2 (father) plus 1/2 x
1/2 (mother) equals 3/4 shared genes.
2If sisters share the same father (as they must do to be full sisters), they
share 3/4 of their genes. If they have different fathers, they will be related
as half sisters and will share only 1/4 of their genes, all through the mother.
However, in two species of Polistes, Metcalf has shown that queens usually
mate twice but use one male's sperm to father 90 percent of their daughters,
(the other male fathers only 10 percent), so most sisters are full sisters and
share 3/4 of their genes.
3In most cases a niece will be the daugther of her aunt's full sister, and
will therefore be related to her aunt by 3/8, which is exactly half the amount
sisters are related to each other (3/4). A mother contributes only half the
genes to an offspring, and the father contributes the other half. An aunt is
related to her niece by 3/8, which is 1/8 less than the relatedness of a
mother to her own daughter, which is 1/2.
4Numbers of offspring a female raises is calculated at the end of the
foundress period in May. Subsequently, young are raised by both workers
and foundresses. Most but not all of the females and males who mate and
become queens are raised by workers at the end of the season, in August.

Mqxch, 1981

Insect Behavioral Ecology-'80 Strassmann

Numbers of young raised at the end of the foundress period are correlated
with the numbers of reproductive progeny that eventually emerge in August.
A female alone raises 3 offspring (3 x 1/2 x 2) while a female raising nieces
is raising 3.75 offspring equivalents, or 5 x 3/8 x 2, where 3/8 is her related-
ness to nieces, 5 is the number of nieces she raises, and 2 is the constant all
terms are multiplied by to get offspring equivalents. (see next note).
sOffspring equivalents equal the product of the number of young raised
twice the fraction of shared genes, so an offspring is 1/2 x 2, (West Eberhard
1975). In the simple case of a 2-female versus a lone-female nest, the lone
female raises 3 offspring-equivalents (3 x 1/2 x 2), a subordinate raising
nieces 3.75 (5 x 3/8 x 2), while the egg-laying queen raises 6.25 (5 x 1/2 x 2)
+ 1.25, where the 1.25 is the queen's credit for the extra needed for the work-
er's 3.75 offspring equivalents to make up the 5 females raised by that worker.
Thus if the association breaks up, the worker loses 0.75 offspring equivalents
and the former queens loses 3.25 offspring equivalents. If there are more
than 2 females on the nest, the offspring equivalents for any one worker
stays the same while the offspring equivalents attributed to the queen goes up
by 1.25 for every additional subordinate assuming the queen is the only egg
layer. So on a nest with 10 females-a queen and 9 subordinates-3.75 off-
spring equivalents are produced by each subordinate and 11.25 offspring
equivalents are attributed to the queen.
6Eighteen nests were observed for a total of 212 hours. Nest 55 was ob-
served when it contained only eggs and again with workers. Nest 102 was
observed before and after a subordinate overthrew the queen. Nest 24 was
observed just after a group of females took over the nest when their own
nest had been destroyed.
7One wasp touching another with her antennae is included with aggres-
sive behavior because this behavior occurs in the same manner as more ag-
gressive behaviors. Females who were chewed on were also antennated.
sAll subordinates ranked as such by this method chewed on their superiors
an average per nest for all subordinates combined of only 8 S.D. 11 per-
cent of the time. The rarity of subodinates chewing on dominants indicates
that the dominance hierarchies in these wasps are much stricter than those
found in primates.
9Significant differences between pairs of ranks 2-5 include that on nests
with larvae, ranks 2 and 3 did more cell checking than rank 5. On nests with
pupae, rank 2 females did more tailwagging than ranks 4 or 5. Rank 2 also
did more building than rank 4 while rank 3 arrived more often than rank 4.
On nests with workers, rank 2 females checked more cells than rank 4 or 5
toNumbers of offspring equivalents produced by females of different ranks
under the constraints of the model given in the discussion were calculated as
shown in Tables 5 and 6. Since the queen has a 35 percent chance of being
overthrown the value assigned the queen in Table 4 is calculated as shown in
Table 6.

GAMBOA, G. J. 1978. Intraspecific Defense: Advantage of social cooperation
among paper wasp foundresses. Science 199: 1463-5.
GIBO, D. L. 1978. The selective advantage of foundress associations in
Polistes fuscatus (Hymenoptera: Vespidae): A field study of the
effects of predation on productivity Canadian Ent. 110: 519-40.
HAMILTON, W. D. 1964a. The genetical evolution of social behavior, I. J.
Theoret. Biol. 7: 1-16.
HAMILTON, W. D. 1964b. The genetical evolution of social behavior, II.
J. Theoret. Biol. 7: 17-52.

Florida Entomologist 64 (1)

March, 1981

JEANNE, R. L. 1970. Chemical defense of brood by a social wasp. Science
168: 1465-6.
METCALF, R. A., AND G. S. WHITT. 1977. Relative inclusive fitness in the
social wasp, Polistes metricus. Behav. Ecol. Sociobiol. 2: 353-60.
METCALF, R. A. 1980. Sex ratios, parent offspring conflict, and local competi-
tion for mates in the social wasps Polistes metricus and Polistes
variatus. American Natur. 116: 642-54.
NOONAN, K. M. 1981. Individual strategies of inclusive-fitness-maximizing in
Polistes fuscatus foundresses. Pages 000-000 In Natural selection and
social behavior: recent research and new theory. Eds. R. D. Alexander
and D. W. Tinkle, Chiron Press.
PARDI, L. 1948. Dominance order in Polistes wasps. Physiol. Zool. 21: 1-13.
PARKER, G. A. 1974. Assessment Strategy and the evolution of fighting be-
havior. J. Theor. Biol.: 47: 223-43.
SIEGEL, S. 1956. Nonparametric statistics for the behavioral sciences. Mc-
Graw Hill, New York, 312 p.
STRASSMANN, J. E. 1979. Honey caches help female paper wasps survive
Texas winters. Science 204: 207-9.
1981. Kin selection and satellite nests in Polistes exclamans. Pages
000-000 In Natural selection and social behavior. Recent research and
new theory. Eds. R. D. Alexander and D. W. Tinkle, Chiron Press.
WEST EBERHARD, M. J. 1969. The social biology of polistine wasps. Misc.
Publ. Museum of Zoology, University of Michigan, 140: 1-101.
1975. The evolution of social behavior by kin selection. Quart. Rev.
Biol. 50: 1-34.
1967. Foundress associations in polistine wasps: Dominance hierar-
chies and the evolution of social behavior. Science 157: 1584-5.

Insect Behavioral Ecology-'80 Lloyd



"... even a bad solution read, and its faults discovered,
has often given rise to a good one in the mind of an
ingenious reader." (Benjamin Franklin)

Sexual rapprochement and contact in bumblebees falls into three general
patterns: (1) males gather at a nest entrance from which virgin queens
(gynes) emerge; (2) males wait individually at prominent objects in bee
habitats, stand on or hover over them, and dart at passing gynes (and ob-
jects that resemble them); and (3) males establish flight circuits with scent
marks, and fly these routes awaiting the arrival of females. A fourth system
is sometimes recognized by splitting (3) into one system that includes the
nest in its route and another that does not (Wilson 1971, Michener 1974,
Alford 1975, and Svensson 1980, and refs.). Few observations have been
made on the first system and Alford (1975) quoted Smith's (1858) observa-
tions of more than a century ago to illustrate it. I observed this system in
Bombus fervidus in late summer (10-20 Sept. 1980) on a farm lawn ad-
jacent to agricultural fields and hedgerows in upstate New York, near Oneida
(Madison County).
Prelude. Drones of perhaps two or more nests, judging from setal
(pelage) color variation, gathered around the nest entrance from early
morning until late afternoon. As many as 20 males were present, especially
in midmorning, but usually 5-10 males (occasionally none) were. The focal
point of drone activity was the entrance hole itself, and seemingly prime
positions at its margin and over it on low, overhanging leaves (Fig. 1 and 2).
Grooming was common in drones perched in the vicinty (< 1 m) of the
entrance, and drones standing at the entrance often fanned their wings (Fig.
Throughout the day workers entered or left the nest and returning work-
ers often carried pollen. They were not protective of the nest, and in several
hours of observation within inches of the entrance I was stung only once (by
a bee trapped between camera and face). During morning, workers oc-
casionally appeared briefly at the entrance and returned below, especially
before the emergence of gyne. In one instance a worker made four appear-
ances in rapid succession, bumping heads of males that were outside the nest
entrance; their flying and buzzing greatly increased, although a queen did
not subsequently appear. Workers took positions just inside the hole and
sometimes further down the shaft at about midday, and chased drones from
the entrance, but they returned immediately.

*J. E. Lloyd is a Professor in the Department of Entomology and Nematology, University
of Florida. His primary research interests are in insect adaptation and mating behavior, and
especially the communication and systematics of fireflies.
Current address: Dept. of Entomology and Nematology, University of Florida, Gainesville,
FL 32611. Florida Agricultural Experiment Station Journal Series No. 2887.

Florida Entomologist 64(1)

March, 1981

Mating. When a gyne emerged from the entrance drones seized her im-
mediately. A grappling "ball" of 2-6+ drones and gyne (Fig. 5-7, 9) con-
tinued up to 3 minutes, tumbling and gradually moving away from the
entrance. Eventually a drone achieved and held onto the prime position,
astride the gyne, holding on with fore and hind legs (Fig. 8), sometimes
fending rivals with his middle ones. Action was rapid and difficult to follow.
A Super-8 movie film of one "balling" (Fleming 1972) reveals that at times
3 males were stacked atop the gyne. Probably, though it is not clear in the
film, males attempted by pushing, pulling (Fig. 8), and biting, to dislodge
males that had achieved the prime position directly on the queen. Males also
hovered over balls (Fig. 9), darted in and joined them. In the filmed ball,
males gradually left the competition until one persistent haggler was left
facing the queen and rider. He made a single jerking motion forward and
fanned his wings, paused, then left. At the conclusion of another balling, the
persistent competitor was confronted by the gyne, which, leaning away from
him, curled her abdomen toward him and raised her middle leg. He withdrew
immediately, and gyne with rider flew off. During balling, gynes "struggled
to escape," running and fighting to get airborne. One gyne left the entrance,
acquired no pursuing males and made 6+ short (< 1 m) flights away from
the entrance and back. Finally she got a rider and departed.
When gynes and consorts left the nest site, 4 of 5 noted flew up the slope
(east). One flew to a building 60 m away, landed on it for 15 sec, then flew
around the building and after 50 m landed on the lawn. She dug headfirst into
the soft grass, drone still riding, until only the tip of her abdomen was visi-
ble (Fig. 10). The pair remained together for 2.5 hrs. The female somer-
saulted in the grass, in repeated bouts of several rotations, in both vertical
and horizontal axes. The riding (not copulating) male sometimes made
sharp body twitches in the anterior-posterior axis. These were spaced at
intervals of 2-4 sec in intermittent bouts of 2 or 3 twitches. While riding and
twitching, the genitalia of the male were not near those of the female; he was
grasping her at the thorax and the tip of his abdomen was several mm short
of reaching hers.
Somersaulting at the first nuptial site created a tight little grass cham-
ber, and in it male twitching was (apparently) not accompanied by the wing-
fanning that occurred in the second and larger chamber. The pair remained
at the first site 100 mins and joined genitalia 3 times, each for ca 5 mins
and at intervals of ca 20 mins. The queen then flew 30 m to a hedge of large
trees, wheeled and flew back over the lawn and landed 20 m beyond the first
mating site. The second site was a 4 cm earthen pocket. Somersaulting con-
tinued, and when the male twitched in this larger chamber, he fanned and
buzzed his wings. They joined genitalia for 30 see, uncoupled, then rejoined.
This last coupling, perhaps the time of actual sperm transfer, lasted 25 mins.
After 10 mins the male was swept from her back during somersaulting, by a
straw that caught between his thorax and abdomen. They remained coupled
for 15 additional mins. During this time she occasionally pushed him with
the apical ends of her hind tibiae, and finally they separated. He faced her
at 4 cm, she addressed him with the defensive posture previously described,
and after a 30 sec pause he flew off toward the nest site. She waited 1 min
and flew in a different direction. Two min later I found him (identified by
paint marking) competing for a prime position at the nest entrance.

Insect Behavioral Ecology-'80 Lloyd

Male Competition. Males competed for perches at the nest entrance, and
certain positions seemed to be favored over others. Males standing on the
lower rim, or on a leaf that extended from the rim over the hole, sparred
(Fig. 1), and could usually remain there only briefly because other males
buzzed ("dive bombed") them in apparent and occasionally successful at-
tempts to knock them into the hole. I saw no evidence of males daubing
honey on rivals, as fervidus workers do in certain attacks on other bees
(Plath 1934). Defending males raised their middle leg(s) high to the side
or over their back (Fig. 4). Males doing this sometimes incurred temporary
or permanent damage, by having the leg jammed in this position. Sometimes
it appeared caught behind the wing, but at others, locked at the joint.
Of quite a different and more intense nature were individual combats in
which two males carried their fight away from the presumptive prime mate-
grabbing stand at the entrance, and grappled and tumbled venter to venter
for minutes. Males often hovered briefly over grappling males, perhaps con-
fusing fighting with gyne balling (Fig. 11) (they even closely examined a
crippled drone clinging to a straw I held over the nest area; Fig. 13). Six of
the fights I observed continued into what may be the most highly escalated
phase that occurs in drone combat. In each case one male achieved a position
atop or beside his opponent and seized his wings (on one side), near their
base, in his mandibles (Fig. 12). In one fight after a male broke away and
ran, his opponent chased and pounced on him, and began wing pulling.
Wings often bent at right-angles during the pulling and may sometimes be
permanently damaged or torn off, as demonstrated by a male found walking
near the nest entrance (Fig. 14). While pulling, which sometimes lasted up
to 5 or more minutes, winning males pushed against opponents with their
tibiae. In a cinematic film clip of this behavior the winner appears to pull
and jerk his head from side to side (like a dog pulling on a pant leg), and to
push and paw at the apical portion of the wing (see cover of Symposium
reprint), in a seeming attempt to sever the wings at the point the mandibles
are holding them. Losers appeared to completely relax and cease struggling,
perhaps minimizing damage. In one fight a male seized his opponents wings,
stood atop him and turned around and around, pirouetting, six times in close
succession, in a manner "obviously" evolved because it dealates and perma-
nently incapacitates a rival. In this case the wings withstood the twisting
and eventually the victim escaped. In species that use this fighting technique,
male wings must be under strong selection to be tough and resilient, and in-
spection of cabinet specimens may indicate something of a species' mating
system, and an individual's fights.

In a straightforward discussion of these observations one could give
parsimonious, simple explanations for: male wing fanning (thermoregula-
tion permitting fast take-off), grooming (cleaning repellant chemicals rivals
had smeared on them), prolonged, highly-escalated fights that move away
from the source of females (behavior belonging to some other situation re-
leased or over-extended inappropriately, or the male determines that no
female is likely to appear and takes time out to dispose of a leading con-
tender), male mating success (ability to be at the entrance at the moment a

Florida Entomologist 64 (1)

gyne emerged, get on her back and not be dislodged by rivals or the subse-
quent somersaulting of the queen), and so forth ....
Likewise, one could also develop a model for the evolution of mating sys-
tems 3a and 3b from system 1: the short scanning flights of males at the
nest-hole, during which males show interest in fights and balls (Figs. 9, 11)
and airborne bee-like objects (Fig. 13), evolved into patrols of greater
length (during which they sometimes detected flying pairs, or pairs at a
nuptial site), that at first included and then excluded the nest-hole itself.
Instead, I prefer to discuss (and then extend) a complex explanation for
mating system 1 that will incorporate several observed behaviors into a
single scheme. The likelihood that it is entirely correct is slim. However, if
it is partially correct and to some extent its key element is paralleled in
other insects, it means that in our present considerations of insect behavior,
and especially mating biology, we are overlooking the single most-important
piece of the puzzle.
The cost, if it is absurdly wrong, is far outweighed by the benefit, if it is
just a little right.
To simplify and save space I shall not use the subjuntive mood but state
the bee model as fact.

Sexual Selection In The Fervid Bumblebee. Drones compete at the en-
trance for a position from which they can fan their individual-specific
pheromone down into the nest to monitoring gynes. (Bumblebee gynes in some
species remain in the nests for several days after eclosion, Alford 1975;
labial-gland pheromones of some bumble bees are known to be individually
specific, Bergstrom and Svensson 1973). A male that is able to send his
"personalized" signal down the burrow over an extended period must be able
to maintain a prime position at the entrance, and is vigorous, a good com-
petitor, etc., etc. (inter-sexual selection). More than being signal-prints,
(individual-distinctive signals), drone emissions are (or carry) signatures-
they have evolved and are used in the context of individual recognition, and
they have historically and do presently promote the emitting individual's
reproductive success.1
When gynes leave the nest for mating they are predisposed to accept
certain males, known by their signatures, over others, and to manipulate the
rivalry and victory of the drones present (Fig. 24a). Even if a "wrong" or
"inferior" male achieves and holds the riding position during competition
and at the nuptial site, and ultimately inseminates the gyne, she may still
choose not to use his sperm (see Lloyd 1979a, Sivinski 1980, W. Walker
1980). The fanning of the mounted queen by "unsuccessful" rivals facing
her, and by the male with her at the nuptial site, enhances the propagation
of his signature, and her opportunity to recognize him then, or later, if she
returns to the nest for additional matings. The extent of gyne somersaulting
at the nuptial site is an indication of her dissatisfaction with her consort.
When competing at the nest-hole not only do males themselves send their
signatures down the shaft, but by dive-bombing they smear their signatures
onto others-then, until they groom, smeared males advertise for their rivals.
Escalated fights away from the entrance, with rival destruction, occur be-

March, 1981

Insect Behavioral Ecology-'80 Lloyd

tween first-rate opponents of long standing; each must dispose of the other
in order to achieve unchallenged priority (favor) with gynes at subsequent
Tracking and Recognition. Two sorts or levels of individual identification
are at first distinguished. A drone may initiate a fight with another because
he has remained in constant unbroken contact with him, and thus knows him
to be a single (the same) troublesome individual. On the other hand, drone
individual identification could be of a second and more complex kind. Drones
could store information about rivals, could remember them and recognize
them, perhaps by their signal-prints, on later occasions of contact even
though separated between times. This usage of recognition is similar to that
of Barrows et al. (1975).
Individual identification of drones by gynes is explicitly required in the
model to be of the recognition kind, i.e. the signature of a male must be
learned and used in later judgment. Continuous tracking is not possible, for
drones are away from the entrance from time to time, and even all night.
Recognition, the reknowing of an individual after a separation, requires dis-
crimination and memory. Tracking does not; it merely requires discernment
for "locking-on." However, tracking and recognition are two modes on a
spectrum of individual identification.
The "intellectual" complexity required for recognition is accepted for
many Hymenoptera. Dominance hierarchies in Polistes provide an example
(Strassmann, this symposium). Barrows et al. (1975) found that male
halictine bees of Lasioglossium zephyrum and Augochlora pura, in controlled
laboratory experiments, discriminated among odors of individual females;
and Bombus pratorum males habituate to the individual odors of unrespon-
sive gynes (J. B. Free, cited in Barrows et al. 1975). Comparable neural
complexity would seem to be demonstrated in the homing of others, such as
the non-social digger and mud-dauber wasps. For other examples of hy-
menopteran learning see Tinbergen 1968, Evans and West-Eberhard 1970,
Menzel 1979, and Brockmann 1980.
Though females of the roach Nauphoeta cinerea are known to prefer-
entially mate with the dominant male in a group, which they recognize on the
basis of odor (Breed et al. 1980), I am not aware of other nonhymenopteran
examples of this sort of learning (see Alloway 1972).
Reasons that so little is known about the existence of individual recogni-
tion in non-hymenopteran insects may include: (a) it is, in fact, uncommon
or lacking; (b) the dominating tendency for population thinking among
entomologists; (c) the difficulties in our knowing insects as individuals (see
also T. Walker, this Symposium); (d) the belief of most researchers that
individual recognition is too complex for run-of-the-mill insects, so they
haven't expected it, or incorporated it into their models and tested for it; (e)
the theorists' arguments that a eugenically-significant choice of a mate is
inconceivable, have reduced the attention naturalists give to animal indi-
viduality when they watch the most common and basic of social interactions,
mating behavior, and when they reflect upon sexual selection; and (f) in-
dividual recognition usually involves signal channels we don't easily monitor
-we observe chemical signals only with difficulty, and often by literally
blending the signals (signal-prints, signatures?) of several individuals; and

Florida, Entomologist 64(1)

March, 1981

the individuality of the best known non-social insects, the singing Orthoptera,
may occur largely in the ultrasonic range and be overlooked.
Faced with the exquisite detail of adaptive fine-tuning one finds in the
morphology of insects-in leaf-like wings with fake midribs, veins, fungal
blemishes, leaf-mines, bird-dropping splashes, etc.; in the flashing-type light
organ of fireflies (Ghiradella 1978, Case and Strouse 1978); in ommatidia
of insect eyes; and in the intracellular ultrastructure that facilitates com-
parable biochemical niceties-it is remarkable that insect biologists haven't
expected the same adaptive fine-tuning and complexity in insect behavior.
Certainly the genetic information required for ordering memory circuits
(and/or compounds) cannot be that much more complex; and theorists have
also argued that sexual reproduction and bumblebee flight are not feasible.
Intuitively, and theoretically too, at least with certain given assumptions,
females with some memory capability are expected to enjoy higher fitness
(Janetos 1980).
Nevertheless, the prevailing notion among most students of insect com-
munication is that individual identification is merely a simple unbroken
tracking. A female cricket or katydid listening to and moving toward a
stationary singing male; a stationary female Photinus firefly flash-answering
and attracting a flying male; a butterfly female using reflected-light signals
to attract a male that will render her sexually captive by "aphrodisiac"
pheromones; a flying female Pteroptyx malaccae firefly approaching and land-
ing near a perched, synchronously-flashing male; a periodical cicada female
perched in the midst of multitudes of males of her kind-all of them, what-
ever information they are supposed to accumulate about their prospective
mating partners as individuals (species, vigor, age, mated state, status in a
lek, etc.), are presumed to do it by means of continuous monitoring, i.e.
simple tracking.
But, when one tolerates the view that individual recognition may occur in
insects, the new perspective suggests a number of experiments and testable
hypotheses for the same old animals, and explanations for the previously in-
explicable begin to take shape; and some fundamental old questions in
natural history become fundamentally altered.
Individuality of Signaling Insects. I have observed individual differences
in the signals of competing katydid neighbors on several occasions. In New
Guinea, males of the "Quanker" katydid (Hexacentrus mundus) emit a long
buzz followed by a number of short, nasal quanks (Fig. 17). Neighbors emit
tarts (raspberries) of quite different acoustical construction (Fig. 18) dur-
ing the quanking, the tactic being to synchronize with the beginning of the
rival's quank, and the counter-tactic to dodge but keep the rhythm (Lloyd
1979a). In New Guinea I also recorded the dueting of a pair of "Flivver"
katydids (Eumecopoda sp. ?) perched within a few decimeters of each other,
that sounded like interacting 1928 Ford cars. Their signal-prints are distinc-
tive (Fig. 19). Large "Piper" katydids ('Sexava nubila) living in banana
trees in the Banda Islands (Indonesia) interact with loud, piping songs of
1-7 regular or syncopated pulses. Nearest-neighbor Pipers are tonally dis-
tinct to the unaided human ear, and have distinctive signal-prints (Fig.
20-23). In North American forests the calls of individual true-katydids,
Pterophylla camellifolia, in treetop choruses can be distinguished by human

Insect Behavioral Ecology-'80 Lloyd

More than 20 years ago Frings and Frings (1957) noted "individual
differences in the songs of different [Neoconocephalus ensiger individuals that
were] great enough to be detectible to the human ear without physical
analysis. Analysis of visible patterns of the sounds [confirmed] the auditory
diagnoses of differences. The differences [were] temporal-missing chirps or
irregularities in sequences-or in the nature of the sound-degree of
'raspiness or sonority'." Greenfield (manuscript) found individual differences
in the songs of the Panamanian conehead Neoconocephalus affinis.
In New Guinea, Diamondback fireflies (Luciola obsolete) perch in loose
groups of up to 100 individuals, in low vegetation. Males emit long-continued
soliloquys of highly variable flashes, off and on throughout the night. When
glowing females take flight males chase them for several meters. Mating
takes place after landing, but is preceded by a courtship "dance" that in-
cludes further luminescent display. I previously suggested that in the seem-
ingly variable emissions of males (Fig. 15 and 16) there is individuality that
females remember and use to certify, by comparison with the luminescent
display given during the dance, that the successful chaser is one she has ob-
served and evaluated (Lloyd 1972, 1973a, Fig. 24b). (This explanation
seemed extreme at the time, but individuality and mate choice go hand in
hand. I am going to suggest here, that this sort of thing may occur in other
insects as well.)
In these examples, if females are discriminating among males and using
signatures for reference, then it is more than merely a discernment of an
entire target-class (e.g. recognizing a member of the same species), and there-
fore more than simple tracking. The Quanker rivals were near each other,
and so were the Flivver males, and their singing overlapped and probably
they moved around. They could not be locked-onto acoustically and followed
individually for evaluation unless they were discriminated and remembered
as individuals. The likelihood of loss of contact and target confusion during
tracking must be variable among insects and depend upon densities and be-
haviors. Tracking grades into recognition. Likewise, in the Diamondback
firefly, for a female to monitor and store information simultaneously on two
or more males, and sort them out after a chase in which they further mix
themselves in aerial dogfights, requires memory.
However, if they had not moved around, but remained at the same discrete
(isolated) signaling station during evaluation, females could simply have
associated their qualities with their address (zip-code) and gone to them.
In insect individuality, three distinct modes now appear: (1) Tracking.
Insects are held in sensory contact by another individual and continuously
tracked. Information about them may be accumulated (stored; perhaps
simply checked off), but their individual identity is not learned. (2) Site-
learning. Individual insects associate themselves with a location-for ex-
ample a banana tree or burrow. They do not broadcast continuously, nor are
their sites necessarily monitored continuously, but after their monitor leaves
and then returns, it continues to know them, and store information about
them as the same individual because they are on the same site. This is "hom-
ing" to another's home. (3) Recognition. The individual is known (learned)
by some characteristic (s) of his phenotype. Sensory contact with him is not
continuous. Information about his individual identity is stored, and he can
be picked out of a crowd or known when he broadcasts from a different

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs