(ISSN 0015-4040)

FLORIDA ENTOMOLOGIST

(An International Journal for the Americas)

Volume 68, No. 1 March, 1985

TABLE OF CONTENTS

68th Annual Meeting-Announcement and Call for Papers ....--............. i

SYMPOSIUM: INSECT BEHAVIORAL ECOLOGY-'84
HUBBELL, T. H.-Unfinished Business and Beckoning Problems ............ 1
REISKIND, J.-That Fateful Friday: Darwin Between the Beagle and
the Origin .........--------..............-.---...... --.. --------..........----- ...... 11
HENRY, C. S.-The Proliferation of Cryptic Species in Chrysoperla
Green Lacewings Through Song Divergence ........---.---. ............ 18
BUSKIRK, R. E., AND K. J. SHERMAN-The Influence of Larval Ecology
on Oviposition and Mating Strategies in Dragonflies -...------......... 39
HUBER, F.-Approaches to Insect Behavior of Interest to both Neuro-
biologists and Behavioral Ecologists ....---........................................ 52
DEWSBURY, D. A.-From Flies to Mice-And Back Again ......--------........ 79
PUNZO, F.-Recent Advances in Behavioral Plasticity in Insects and
Decapod Crustaceans -------.-------- --.. .................... ................. 89
RIECHERT, S.-Why Do Some Spiders Cooperate? Agelena consociata,
A Case Study .--------.......---.... --------........ .............-...-... 105
HABECK, D. H.-Attracting Insects for Backyard Entomology .............. 117
O'MEARA, G.-Gonotropic Interactions in Mosquitoes: Kicking the Blood
Feeding Habit ....----- - -----------................ ..... ......... ... .............. 122
LLOYD, J. E.-On Watersheds and Peers, Publication, Pimps, and
Panache (An Editorial Abstract) .-..-............................. ............. 134

RAKHA, M. A., AND C. W. McCoY-Eupalopsellid Mites on Florida
Citrus, With a Description of Exothorhis caudata Summers De-
velopmental Stages --...........---..-.......... ..-- ..---...... ......- --........-.-- 141
GREANY, P. D., P. E. SHAW, P. L. DAVIS, AND T. T. HATTON-Senes-
cence-related Susceptibility of Marsh Grapefruit to Laboratory
Infestation by Anastrepha suspense (Diptera: Trephritidae) -._ 144
WooD, T. K., AND R. DOWELL-Reproductive Behavior and Dispersal in
Umbonia crassicornis (Homoptera: Membracidae) ...-----....--........ 151
Continued on Back Cover

Published by The Florida Entomological Society

FLORIDA ENTOMOLOGICAL SOCIETY

OFFICERS FOR 1984-85
President .......--..-......... ...-- -------.............. ----- M. L. Wright, Jr.
Preshident-Elect ..----..-...------ ..........------ ----------.. D. H. Habeck
Vice-President ..-..............- ....---..-........-- ---- D. J. Schuster
Secretary ...--...................----- ....-------....... -------- .D. F. Williams
Treasurer .............. -- ------------.. .. ................................ .....A. C. Knapp

R. G. Haines
C. W. McCoy
G. J. Wibmer
Other Members of the Executive Committee ..- R. W. Metz
R. C. Bullock
C. A. Morris
J. R. McLaughlin

PUBLICATIONS COMMITTEE

Editor .............--- ... ..... .......----------------- ------. J. R. McLaughlin
Associate Editors .--........-----......................----------------- W.C. Adlerz
A. Ali
J. B. Heppner
M. D. Hubbard
O. Sosa, Jr.
H. V. Weems, Jr.
W. W. Wirth
Business Manager ............... --------- ----------................ A. C. Knapp

FLORIDA ENTOMOLOGIST is issued quarterly-March, June, September,
and December. Subscription price to non-members is $20.00 per year in
advance, $5.00 per copy. Membership in the Florida Entomological Society,
including subscription to Florida Entomologist, is $15 per year for regular
membership and $5 per year for students. Inquires regarding membership
and subscriptions should be addressed to the Business Manager, P. O. Box
7326, Winter Haven, FL 33883-7326. 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 S 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,
JOHN R. MCLAUGHLIN, 4628 NW 40th Street, Gainesville, FL, 32606.

This issue mailed March 25, 1985

68th ANNUAL MEETING-FLORIDA ENTOMOLOGICAL SOCIETY
Third Announcement
1. DATE AND LOCALE: August 5-8, 1985; Ocho Rios Sheraton Hotel;
Ocho Rios, Jamaica, West Indies.
2. OFFICIAL TRAVEL AGENCY: EMMER TRAVEL, INC., 2801 S.W.
Archer Road, Gainesville, FL 32608; contact Ms. Lona Stein at (904)
377-1222 (Gainesville residents) or (800) 342-2223 (toll free for Florida
residents) or (800) 874-8487 (toll free for USA residents).
3. Refer to the December 1984 issue of FLORIDA ENTOMOLOGIST for
details. Contact Carl Barfield (904-392-7089) or Dave Schuster (813-
755-1568) if you have questions.

cdr
c I
C 0
Q co

S a
0 0a)

C11
a) -,

-4-)
000
r CM C
W~ w C/3

U2?F
11

P cB

o

CO

a,

~C)
a)

o

*r-4
M

0c

0 a)
a0

,.

E
-
u
M

-i-i

0a)
FI2
-1>
3
C

Sc,

^C O

oa,

o Fh
n -
ri'
k=
C +-
.5 u
S O
00 g
ouf
oQ F3

a|

m k

m1^
ra <

a u
*i-
TS?-
s a,
-is
+SI

<1 1

a)

0

-4-

Insect Behavioral Ecology-'84 Hubbell

UNFINISHED BUSINESS AND BECKONING PROBLEMS

THEODORE H. HUBBELL*

When I was inveigled into agreeing to take part in this symposium I
chose a title that left me plenty of room to decide later what to talk about
and one that might encourage speculation. I shall disappoint any of you who
may expect me to philosophize about the importance of behavioral studies
for systematics, or utter other words of wisdom. Actually, I don't belong
here. I do not myself work on insect behavior, but I am interested in it and
greatly impressed by the research you behaviorists are doing. And I must
tell you that although the insects I study neither sing nor flash lights, almost
every one of my projects has revealed problems or situations that can only
be explained in terms of behavior. Many other so-called old-fashioned
taxonomists, among whom I class myself, would doubtless say the same.
When I began my studies of the Orthoptera only a few men of insight
such as B. B. Fulton were using behavior as a systematic tool and an in-
dicator of relationship; the rest of us were quite oblivious of its possibilities.
We based our classifications on "characters" (morphological, of course), and
assessed degree of relationship according to closeness of resemblance in
those which we considered "good"-that is, distinctive and dependable. Dis-
tribution and ecology were also taken into account. Most of us still have to
operate under those rules.
Much of my early work on Orthoptera was done in Florida, and my
"unfinished business" relates mostly to Florida and adjoining states. As
background, I am going to describe briefly what it was like to be a biologist
at the University of Florida in the 20's and 30's-before most of you were
born. I had gone in 1922 to study under W. M. Wheeler at Harvard, whence
I was plucked the next year, degreeless, by the offer of a position as in-
structor in the Department of Biology and Geology at the University of
Florida. (Such things happened in those days). The offer came from
J. Speed Rogers, a professor at Grinnell College, Iowa, who had just been
appointed head of the department at Gainesville. We had first met at
Michigan, and had done field work together. It was his job to build a staff
from scratch, and with complete disregard for procedures now routine, he
picked the men he wanted from among his former fellow graduate Ptudents,
and thus assembled what came to be known as the University of Michigan
Department of Biology, Florida Branch. In spite of its faults this procedure
can, with judgement and good luck, make a small department strong in a
limited field at the cost of breadth and variety.
So it happened at Florida. Speed (craneflies), a mammalogist, a botanist
and I (Orthoptera) made up the original staff, plus an economic entomolo-
gist who soon went over to Agriculture. Each of us was interested in a group
of organisms and the problems it posed rather than in some biological
phenomenon as such. Only the botanist had a doctoral degree, but in those

*Theodore H. Hubbell is Emeritus Professor of Zoology at the University of Michigan and
former Director of the Museum of Zoology of that institution. He was on the faculty of the
University of Florida from 1923 to 1946, and was President of the Florida Entomological
Society in 1942-43. He is a student of the Orthoptera and in particular of the camel-crickets
(Rhaph'dophoridae). Address: Museum of Zoology, University of Michigan, Ann Arbor, Mich-
igan 48109.

Florida Entomologist 68 (1)

March, 1985

Fig. 1. Field trip to Florida panhandle, 4 April 1927: Rock Bluff Landing
(near "Old Camp Torreya" ravine). Left to right, standing: Edward T.
Boardman, Cyrus Crosby, unknown local youth, Mortimer D. Leonard, James
G. Needham, Fred W. Walker, T. H. Hubbell, William Everts; sitting: H. K.
Wallace, Kenneth Haggert. (Identifications by T. H. Hubbell and H. K.
Wallace; photo by J. Speed Rogers).
days a degree was not too important. The administration was satisfied if
we gave our courses acceptably. I taught Geology, Physical Geography,
general Zoology, and occasionally Entomology. The M.S. was then the
highest degree given in biology at Florida, and we tended to make it a little
Ph.D., complete with committee, research problem and thesis. When, some
years later, the doctoral program was added we professors took turns in
going off and getting properly annointed. Research by the teaching staff was
regarded as commendable and not in any way discouraged. For example,
the dean would overlook an occasional few days' absence from campus of a
professor and group of students, ostensibly for a course exercise but often
little more than a camouflaged collecting trip. The 5000 students were all
male, and the faculty was so small that most of us knew one another. All
in all it was an easy-going place. One great difference from now was the
lack of pressure to publish, which fostered a relaxed attitude towards one's
own research. In fact, almost the only incentives for doing research were
curiosity about nature and the desire for recognition by one's colleagues
elsewhere.
Actually we worked hard. We were young and enthusiastic, in love with
Florida, and passionately fond of field work. We were fired by the belief
that we had a unique opportunity and mission-to make known the fauna
of Florida by our own efforts and those of our students. Each student was
encouraged to work on a different group, or, as an alternative, to study the
biota and ecology of some characteristic Floridian environment. With this
as the department's program we hoped that its faculty and students would
become a cooperative, self-renewing body of field-oriented biologists with
interests focused on Florida. And do you know, after we got going this

-a5

Insect Behavioral Ecology-'84 Hubbell

Fig. 2. James G. Needham, Rock Bluff Landing, 4 April 1927 (photo by
J. Speed Rogers).

plan worked pretty well for quite a while. It produced a crop of students
each of whom soon knew far more about his chosen group than did his
major professor, a series of theses on various Florida taxa and environ-
ments, and eventually a number of distinguished biologists, some of whose
names are certainly familiar to you. It also fostered a relationship among
faculty and students that in my estimate was well-nigh unique in the respect
and high regard we had for one another. How much this depended on one
man-James Speed Rogers-is hard to say; everyone contributed. But it is
perhaps no coincidence that departmental solidarity and elan suffered after
his eventual departure to Michigan.
It was about the time we were assembling that Thomas Barbour, Di-
rector of the Museum of Comparative Zoology of Harvard University, pub-

Florida Entomologist 68(1)

March, 1985

lished his book, "That Vanishing Eden," in which he deplored the rape and
devastation of the state he had known for so long. But for us late-comers
Florida was virgin country-wonderful in its diversity of life and environ-
ment-fresh and new and strange. We threw ourselves into the study of our
chosen groups, spending every possible moment in the field. Exploring and
collecting. We wanted to know what occurred, and where. Accomplishment
was to find the right place, locate the species, collect specimens and bring
them home to study-dead, but accompanied by meticulous field notes. We
did not, of course, spend all our time and energy on Florida. Most summers
we dispersed, to do field work in more northerly climes, study in museums,
or even, rarely, to take vacations. (And please note: we ourselves paid for
our field trips and expeditions, sometimes with a little help from a museum
for which we collected. There was no such thing as a grant-in-aid. Those
expenses were not great in terms of dollars, but they were not small in
relation to our resources; my salary as instructor at Florida in 1923 was
$1800, and 13 years later, as professor, I was making $3200).

Fig. 3. T. H. Hubbell, Rock Bluff Landing, 4 April 1927 (photo by
J. Speed Rogers).

Insect Behavioral Ecology-'84 Hubbell

Fig. 4. Left to right: Albert P. Morse, William T. Davis, Willis S.
Blatchley, T. H. Hubbell, on another west Florida trip.

Most of our Florida field work was done in the vicinity of Gainesville.
Only once in a long time did we get as far south as Royal Palm Park or the
Keys, but forays to central peninsular Florida or the panhandle were more
frequent. It took us two or even three days to get to the Apalachicola ravines
over the unpaved roads, hand-pushing the cars when they got stuck in the
deep sand and sleeping in the woods. We generally camped beside a ravine
south of the present park, a site which became known as Old Camp Torreya
and is the type locality of many insect species. On some of our visits we
were accompanied by out-of-state visitors, among them, on one such oc-
casion, Cy Crosby and James G. Needham from Cornell and W. M. Barrows
from Ohio State, and on another A. P. Morse, W. T. Davis and W. S.
Blatchley. The accompanying photos were taken on such trips. (Figs. 1-4).
To a biologist one of the fascinating things about Florida is the variety
and distinctness of its habitats, and the often sharp boundaries between
them. Detailed county soil maps were available even in the '20s, and having
been produced largely by mapping vegetation types they proved useful in
our field work. Sand scrub is one of the most distinctive Floridian environ-
ments, developed on the little altered dune and beach sands that constitute
the St. Lucie and Lakewood soils (Fig. 5). It is very dry, and supports a
thin forest of sand pine and scrub oaks, with shrubs such as Xolisma and
Ceratiola, sparse grass and herbage, and much exposed sandy soil. The
scrub is widespread in central and northern Florida, occurring as a few
large and a multitude of small island-like patches isolated sometimes by
streams but mostly by intervening swamps, flatwoods, hammocks, or 'high
pine" ridges-that is, pine-oak forests on loamy uplands. As is now well
known, prolific speciation has occurred in this scrub archipelago among
animals that share the following characteristics: they are generally small
but not minute, flightless or at least sedentary, not normally subject to
passive dispersal, and for one reason or another closely associated with the
scrub environment. Among such animals we observe (1) the existence of

Florida Entomologist 68 (1)

March, 1985

Fig. 5. The archipelago of "scrub islands" in central Florida (black);
the largest is that located in Ocala National Forest.

numerous morphologically distinct, allopatric or parapatric populations,
often with very small ranges separated by what appear to be trivial bar-
riers; (2) the absence from many isolated scrub patches of any member of
the group; and (3) a tendency for the taxa of these microspeciating groups
to share a common distributional pattern.
When we began our studies none of this was known, and most of the
taxa involved were undescribed. I became particularly interested in the
flightless scrub grasshoppers of the genera Melanoplus and Aptenopedes
and the burrowing scarabs of the genus Mycotrupes. In 1932 I published a
revision of the Puer Group of Melanoplus, in which the scrub species were
distinguished largely by their concealed male genitalia. Surprisingly, this
was the first time those protean and easily observable structures had been

Insect Behavioral Ecology-'84 Hubbell

used in the taxonomy of New World grasshoppers. I concluded that most of
the species do not intergrade even when their territories are only narrowly
separated. A single one, however, Melanoplus puer, has an extensive north-
south range in the sandy areas along the east coast and their inland exten-
sion in central Florida north of Lake Okeechobee. Its populations show broad
regional or subspecific differences and also many localized morphotypes. It
seems probable that in the central part of its range a complex zone of
hybridization exists, with differences in the amount and content of gene
exchange between populations.
In my revision I suggested that the scrub areas where the microspecies
now occur might once have been real islands at times of high sea level during
the Pleistocene (shades of vicariance biogeography!). This hypothesis I
later elaborated in a study of Mycotrupes (1954). It needs reexamination,
particularly with regard to the timing of postulated events, in the light of
more recent interpretations of the terraces of the southeastern coastal plain.
The Puer study also demonstrated that much more field work was needed,
and in the summer of 1938 a graduate student and I investigated numerous
previously unvisited areas. I even persuaded my dentist, an amateur pilot,
to fly me over much of the territory north of the Oklawaha River in search
of any scrubs we might have missed. We turned up three more species of
the Puer group, one in west and one in northeast Florida and another con-
fined to the Ocala Forest (all still undescribed). The Ocala species proved
to be separated from its nearest neighbor to the south only by Alexander
Spring Creek and its narrow bordering swamp.
And there this particular problem sits. Events conspired to prevent me
from pursuing it, and there has been no followup. It would be interesting
to determine how far individuals of the parapatric species wander in a life
time and whether they can and do penetrate the insignificant-seeming
barriers between colonies, and to try cross-mating the scrub microspecies
and also the local forms of puer among themselves to see if any behavioral
barriers exist and whether the genitalic differences cause mechanical isola-
tion. Chromosomal and electrophoretic studies might also provide data
indicative of degrees of relationship and relative ages of the scrub forms.
Just as there is a pattern for xerophiles in this region, so is there a
reverse one for hygrophiles, among which belong the grasshoppers of the
Furcatus Group of Melanoplus. The three species of this group, clypeatus,
furcatus and symmetricus, inhabit the shrub zone on seepage slopes border-
ing the edges of swampland. The first two face each other across the Al-
tamaha River along a large part of its course, as shown by many paired
collections. In west Florida furcatus and symmetricus are separated only by
a two-mile wide sand ridge. The middle species, furcatus, includes several
populations that differ in male cereal form and genitalia; all but one of
these are known to intergrade with their neighbor or neighbors, and that
one is suspected of doing so. Two of these subspecies, if one may call them
so, are separated by a narrow ridge at the east edge of the Okefinokee
Swamp but intergrade around its south end. How or why these parapatric
colonies are kept separate across such an insignificant distance (insignifi-
cant to a medium-sized grasshopper, that is) remains a tantalizing question-
just one more instance of unfinished business.
I shall end by describing two problems among the camel-crickets (the

Florida Entomologist 68 (1)

March, 1985

a

b
Fig. 6. Pronotal form in males
of Ceuthophilus seclusus: (a) un-
modified pronotal form of minor
males inhabiting peripheral part of
range of species; (b) modified
pronotal form of major males in-
habiting core part of range cen-
tered on the Ozarks (the arrow
points to the adventitious cleft
separating the fore and hind parts
of the protuberance).

group Theodore Cohn and I are currently studying)-problems that can
only be solved by behavioral studies. Both of them involve the presence in
single species of two kinds of males-major and minor-that differ strik-
ingly in bodily conformation but not at all in genitalic structure. The first
case, that of Ceuthophilus seclusus, I described in 1936. In every collection
from a sizeable area centering on the Ozarks every male is a major, with a
pronotum that is strongly elevated and deformed (Fig. 6); such a modifica-
tion occurs in no other species of the genus. Outside that area, forming a
complete ring around it and extending into Arkansas, westernmost Indiana,
Illinois, Iowa, eastern Nebraska, Kansas and Oklahoma, is a zone in which
the male pronotum is entirely ordinary. Only in a few specimens from the
northern part of that zone does a hint of the modification occur. Further-
more the constricted neck of the pronotal elevation has been traumatically
destroyed in a high percentage of the major males seen-almost surely by
the jaws of the female during or prior to mating. Can it be that there is an
abrupt behavioral change at the edge of the area where only major males
occur? How could there be a behavioral dine when the structure involved
is either present or absent? Is the major morphotype so successful that
where it is present only majors can secure mates, and is it consequently
spreading into peripheral territory at the expense of minors? And how do
provincial females acquire their taste for the pronotal juices or pheromones
or whatever it is that the sophisticated major males provide? Questions such
as these can only be answered by observation, but I find behavioral observa-
tions frustrating. Last year I watched a female and major male seclusus in

Insect Behavioral Ecology-'84 Hubbell

I

t'4:

r

Fig. 7. Differences in the abdomens (upper, dorsal views; lower, lateral
views) of three supposed species of Pristoceuthophilus: cercalis (Sc.) (left;
minor), gaigei Hub. (center; major), and sargentae Gurney (right; major);
see text.
dim light for several hours, and then suddenly it was all over; they were
in copula, apparently with no more preliminaries than a nudge or two by
the male, and the female did not even give him a nibble.
The other beckoning problem that I shall briefly mention is that posed
by three "species" of Pristoceuthophilus that occur in the northwestern
United States and adjacent Canada-cercalis, gaigei (one that I described),
and sargentae. They differ enormously in the ornamental nodules, processes
or bosses on the dorsum of the abdomen of the males: cercalis has either a

10 Florida Entomologist 68(1) March, 1985

smooth back or a large number of tiny low tubercles; gaigei has a very
peculiar enlarged structure in the middle of the abdomen accompanied by
long finger-like lateral projections, and sargentae has, in addition to "fing-
ers," an enormous median organ as tall and as wide as or wider than the
abdomen (Fig. 7).
For a long time we were certain that we were dealing with at least two
species. It was not until we plotted the distributions of cercalis and gaigei
that we noticed that their ranges are absolutely coextensive, that in prac-
tically every large collection made anywhere from Idaho to British Colum-
bia to Oregon both are present, and that they are synchronic. Later we
found that the distinctive male genitalia are completely alike in all three
forms. We have been forced to conclude that we are dealing with a re-
markable instance of male dimorphism, with cercalis based on the minor
male phase and gaigei and sargentae representing extremes of the major
phase. Majors and minors occur in more or less equal numbers, at least in
many places, and are of the same general size; we therefore presume that
they are equally successful in obtaining mates, from which it seems logical
to suppose that this is accomplished by different tactics. The central organ
of the majors may disperse pheromones; but what can be the role of the
"fingers," unless to enable the female to discriminate among the males by
tactile stimuli? Having exposed this problem by old-fashioned means, we
would give a great deal to learn the answers to some of these questions of
function and behavior.
The "natural history" approach to systematics that has pervaded my
talk may seem naive and old-fashioned to you, who are applying new tools
and new concepts to the study of systematic and evolutionary problems. But
are the concepts really so new? Is it not true that, by whatever means, we
are all simply engaged in filling in the details of Darwin's great vision in
ways he never dreamed of? Since by some magic he is expected to attend
this meeting, I wonder what he will think of the studies to be described here
today. I think he will find them fascinating; perhaps he may even ask some
penetrating questions.

LITERATURE CITED
HUBBELL, T. H. 1925. A new species of Pristoceuthophilus from the Olympic
Mountains, Washington. (Orthoptera, Tettigoniidae). Pan-Pacific
Ent., (2) 1: 39-42.
-- 1925. A revision of the Puer Group of the North American genus
Melanoplus, with remarks on the taxonomic value of the concealed
male genitalia in the Cyrtacanthacrinae. Misc. Pub. U. Michigan Mus.
Zool, No. 23, 64 pp.
1936. A monographic revision of the genus Ceuthophilus
(Orthoptera, Gryllacrididae, Rhaphidophorinae). Univ. Florida Pub.,
Biol. Ser. 2 (1): 551 pp.
OLSON, A. L., HUBBELL, T. H., AND H. F. HOWDEN. 1954. The burrowing
beetles of the genus Mycotrupes (Coleoptera, Scarabaeidae, Geo-
trupinae). Misc. Pub. U. Michigan Mus. Zool., No. 84: 59 pp.

Insect Behavioral Ecology-'84 Reiskind 11

THAT FATEFUL FRIDAY:
DARWIN BETWEEN THE BEAGLE AND THE ORIGIN

JONATHAN REISKIND*

SYNOPSIS
On Friday, 28 September 1838, Charles Darwin read for amusement
Malthus' Essay on Population, and "the rest is history." This emphasis on
the importance of Malthus' essay reflects Darwin's own emphasis in his
autobiography (written many years later), though some present-day his-
torians conclude otherwise. It was a moment of great pivotal importance to
Darwin and his theory. In the monologue here, a Darwin from 1844, un-
aware of the revolution he will bring, believes he has been asked to talk
about his entomological interests and his well-known voyage on HMS Beagle.
But he digresses, and, aware of the privacy afforded by this meeting beyond
his future, he confides in the audience and outlines his "recent" views about
transformism (evolution) and the mechanism of natural selection. It would
be another 14 years before he would present his theory in public (in a joint
publication with A. R. Wallace).
Several sources were used in ascertaining Darwin's views and attitudes
at that time: mainly Kohn (1980), but also Darwin (1962, 1969, 1980), Gale
(1982), Mayr (1977), Bowler (1984) and Ruse (1979). The last two refer-
ences have especially valuable bibliographies.

Setting:
The annual meeting of the Florida Entomological Society in Orlando,
Florida (July 24-27, 1984) offered an ideal opportunity to bring Charles
Darwin to speak to a group of assembled scientists. The symposium moderator
has just introduced Professor Reiskind, who strides to the podium to intro-
duce Mr. Darwin...
It is with great pleasure that I announce the first success of a remark-
able interdisciplinary program at the University of Florida. The discoveries
at the Center for Temporal and Spatial Translocation have recently per-
mitted the judicious first-hand examination of events up to 150 years into
the past, even permitting the transport of individuals from the past to the
present. The first public exhibition of this remarkable breakthrough is our
next speaker. I was priviledged to be asked to choose both the year and the
person to be translocated, since the History Department got hopelessly en-
tangled in intradepartmental bickering.
I chose Charles Darwin in 1844, exactly 140 years ago to the day. We
were able to pick a quiet, late, summer afternoon at his home in Down to
"spirit" him to our meeting. Mr. Darwin was 35 years old at the time and
involved in many projects, although still recognized publicly mainly for his
popular book on the voyage of the Beagle.
Now, without further ado, I am pleased to present to you-Mr. Charles
Darwin, 26th of July 1844.

*Jonathan Reiskind is Associate Professor in the Department of Zoology at the University
of Florida. His interests are principally in the systematics and ecology of spiders but he finds
historical perspectives essential to the appreciation of science and life. Current address: De-
partment of Zoology, University of Florida, Gainesville, FL 32611.

12 Florida Entomologist 68(1) March, 1985

(Mr. Darwin (played by David Crabtree*) enters rapidly from the side
and jumps onto the low platform. He is dressed appropriately for a young
man of reasonable wealth in the early Victorian period-plaid waistcoat and
handsome coat. He is wearing a moustache and long sideburns (Figure 1).')
My dear friends and fellow entomologists, who would have thought it
possible to travel 140 years into the future and almost 4000 miles from my
study in Down? I certainly regret that I will not be permitted to get into
the field in this area so familiar to many of us in England as a result of
William Bartram's fascinating "Travels". It is a pleasure to be with you
today, especially with you entomologists, for my first love is entomology, and
especially the beetles which, I trust, still make up the plurality of insects,
even after all the explorations.
No pursuit at Cambridge was followed with nearly so much eagerness or
gave me so much pleasure as collecting beetles. It was the mere passion for
collecting, which I know many of you will understand. I will give proof of
my zeal: one day, on tearing off some old bark, I saw two rare beetles and
seized one in each hand; then I saw a third and new kind, which I could
not bear to lose, so that I popped the one which I held in my right hand into
my mouth. Alas it ejected some intensely acrid fluid, which burnt my tongue
so that I was forced to spit the beetle out, which was lost, as well as the
third one.2 I'm sure each of you have a pet anecdote about the one that got
away. But I know you didn't invite me here to speak about my haphazard
university days collecting beetles while contemplating my future as, first, a
clergyman and then a physician.
Nowadays my interests include coral reefs and barnacles .... and yet
my mind is on the voyage, the diversity of life and how it all got to be that
way, and I suppose the basis of this kind and generous invitation was the
popularity of my book about that voyage-"The Voyage of the Beagle".
Indeed, I spent an exciting 57 months around the world and the insects and
spiders were fascinating both on land and at sea. The true magnitude of
insects was brought home to me in Bahia, Brazil, where my attention was
caught one day by many spiders, cockroaches, and other insects, as well as
some lizards, rushing in the greatest agitation across a bare piece of ground.
A little way behind, every stalk and leaf was blackened by small ants. The
swarm having crossed the bare space, divided itself,, and descended an old
wall. By this means many insects were fairly enclosed; and the efforts
which the poor little creatures made to extricate themselves from such a
death were wonderful. When the ants came to the road they changed their
course, and in narrow files re-ascended the wall. Having placed a small
stone so as to intercept one of the lines, the whole body attacked it, and
then immediately retired. Shortly afterwards another body came to the
charge and again having failed to make any impression, this line of march
was entirely given up. By going an inch round, the file might have avoided
the stone, and this doubtless would have happened, if it had been originally
there: but having been attacked, the lion-hearted little warriors scorned
the idea of yielding.3 And I could go on and on with these anecdotes.
But ever since returning to England in '36 I have been grappling with

*David M. Crabtree is working towards a Masters of F'ne Arts in Theater at the Uni-
versity of Florida. He has appeared in productions of the Florida Players and is a founding
member of a professional theater company in Jacksonville, Florida-A Company of Players.

Insect Behavioral Ecology-'84 Reiskind 13

Fig. 1. ". and perhaps such reservations that I now have about my
past musings are just hindsight . well, ... finally we come to that fateful
Friday . ." (See text; David Crabtree as Darwin; photo by Frank Mead,
FADCS-DPI).

14 Florida Entomologist 68(1) March, 1985

the multitude of biological and geological data I have collected, and the new
directions it has lead me in have been somewhat unsettling. Although I have
confided in only a few of my closest colleagues, the separation afforded me
in both time and space here will allow me, I think without fear, to share
some of my recent ideas which I would not have the audacity to share with
my contemporaries for fear of ostracism.
For soon after the conclusion of my circumnavigation of the earth, I was
to come to the inescapable fact of the transmutation of species. Not only
that but I have come to an even more remarkable realization-a reasonable
explanation. But I get ahead of myself. Let me go back to my most preg-
nant voyage. I went with high enthusiasm, which continued throughout the
trip although I was soon knowledgeable of my perpetual seasickness-which
never did ameliorate! Ah, the pleasure of the landfall. It was at one of these
landfalls in South America that Charles Lyell's latest book (the second
volume of his Principles of Geology) finally caught up with me. What a
wondrous and illuminating work it was, elucidating Hutton's uniformi-
tarianism for all to appreciate. Geology has always been a distinct pleasure
for me (in fact I was assisting Sedgewick in Wales the summer before my
departure) and my latest papers have been geological, involving the work
of ancient glaciers. But I digress. Lyell spoke on species as well and I be-
lieved, at the time, that he spoke with a great deal of sense. Surely species,
being so perfectly fitted to their existence, were immutable-and yet many
species have become locally extinct (as I observed in the fossil beds of the
Argentine) and others have no doubt arisen locally, thus maintaining a
balance. This gradual birth and death of species accounted for the many
layers of fossils seen in Patagonia.
I continued to accept Lyell's views as reasonable through the rest of the
voyage and it was not until the Spring of 1837 that the preponderance of
my observations led me to a quite different view of life's diversity. I had had
the opportunity to read Lamarck's Philosophie Zoologique that winter and
had been working on my bird collections, necessitating many fruitful dis-
cussions with John Gould. And then there was the Galapagos Archipelago.
These young, volcanic islands were a wealth of amazing life offering more
evidence. The finches there-all so similar yet each island having its own
assemblage of species-were so convincing of transmutation of a single
species into many. Their progenitor clearly arrived in the recent past from
the nearest mainland-South America. This, with the additional evidence
of remarkable anatomical resemblances between certain species (to wit, the
skeletal resemblances between vertebrates) and the similarities between
fossil species and modern forms, could best be explained by the assumption
that species were mutable; old species being transformed into new species.
Once I had accepted transformism in the Spring of '37, I was faced with
a much tougher problem-How? How could forms change though time? That
mechanism of urge and need proposed by Lamarck made little sense. Could
I come up with a better mechanism?
I wrestled with several schemes.
At first I contemplated the existence of two llamas in South America,
one extinct and one presently extant (both in the same type of environs).
Why two species in the same environment? From this I was tempted to
believe that the species, like an individual, was created for a definite time

Insect Behavioral Ecology-'84 Reiskind

and then became senescent and naturally died out, to be replaced by a new
species produced at one blow from the old species quite independent of the
physical environment. But while this was a way of looking at the actual
progression of life it surely was not a mechanism to explain how things
changed. That would require something with respect to passing on traits
from parent to child. And so I came, by that summer, to a mechanism allow-
ing change to occur. I assumed that the physical environment both changed
gradually [Darwin pauses reflectively and then continues] (Lyell would like
this) and constantly and that it induced variability in the organisms in that
environment, and that such variability was adaptive and could be passed
on, and finally, that only in sexual reproduction do progeny show variability
(asexual reproduction allowing none) and thus it can spread to others in
the species. The species then, to survive, must be transformed. While my
theory of generation appeared logical, it surely left much to be desired, since
little is known of the bases of inheritance. In any case, those species that
had not had sufficient variability would die out, a result of non-adaptation
of circumstances and would be replaced by other species. Those with suf-
ficient variability would have been transformed by the environment induc-
ing the appropriate variation.
I continued in this vein but gradually became aware of weaknesses in
my thinking; perhaps my analogy of species and individuals was not quite
accurate. Also from the animal breeders I began to recognize the uniqueness
of the individual, and perhaps my assumptions of the impact of the environ-
ment on variability as the basis of directional change ascribed too much
influence to the malleability of the inheritable material . and perhaps such
reservations that I now have about my past musings are just hindsight . .
well, ... finally we come to that fateful Friday in September of 18384 when
I perchance, for amusement, read Malthus' Essay on Population. Suddenly
the "struggle for existence" he spoke of in the human condition, and which
I appreciated from my long-continued observation of the habits of animals
and plants5, fitted in with the variation observed in each species. That
variation being not the result of environmental induction, but rather an
integral part of the individuals making up each and every species. Thus,
with the limitation of resources available to all, the potential geometrical
increase of populations [turning to a member of the audience and addressing
him] (seen in the progeny in every species I have observed outnumbering
their parents) and the observed stability of all populations in nature (previ-
ously thought to be the result of some divine balance) it at once struck me
that under these circumstances favorable variations would tend to be pre-
served, and unfavorable ones destroyed6. No longer was it necessary to
speculate on mysterious mechanisms (such as environmentally induced vari-
ation or species senescence), now I had at last got a theory by which to
work7. I celebrated within myself that weekend and began the task of sub-
stantiating this novel idea. It was not for almost four years that I allowed
myself the satisfaction of writing a brief abstract of 35 pages in pencils
and only now I am enlarging it into a more substantial work. But it remains
only known to my closest confidants (and now to you). What will come of
this I do not know, but I believe it will ultimately bring reason and science
to bear on a subject too long the purview of metaphysics and wishful think-

16 Florida Entomologist 68(1) March, 1985

ing. However, I have instructed by beloved Emma, that, if anything should
happen to me, this summer's work be published posthumously.
Suddenly I now have a theory that requires only those things that
natural organisms have in abundance-heritable variation and offspring
(that in turn have offspring)--and a world in which populations are stable
and which has been here for quite a while (as well documented by Lyell,
who thinks it's been here perhaps forever!). The selection that must in-
evitably follow each individual's struggle for existence naturally culls the
less capable from the population. Hence change and the origin of new
species.
But I regret I have gone far afield of the subject I had been asked to
speak to-that is, my entomological observations while on the voyage of
HMS Beagle. Please accept my apologies for letting my personal enthusiasm
for these novel speculations monopolize your time. I pray I have not bored
you.
Well, now I am out of time and my perpetual dyspepsia precludes my
continuing in any case9. Thank you for your kind attention and forbearance.

APPENDIX

1His moustache is quite a surprise. Portraits (e.g. the watercolour by
George Richmond in 1840 and the chalk drawing by Samuel Lawrence in
1853) show him cleanshaven. Undoubtedly he was experimenting, perhaps
influenced by the hirsute new Prince Consort, Albert.
2"No pursuit at ... passion for collecting" and "I will give proof ... the
third one." from Darwin (1969:62).
3"one day my attention . the idea of yielding" from Darwin (1962:34).
428 September 1838. Referred to in Notebook D of Charles Darwin. This
date is the basis of the title of this talk. That title perhaps suggests a
greater importance of Malthus' essay in Darwin's theory of natural selection
than some contemporary historians are willing to give it. However it re-
flects Darwin's own emphasis in his autobiography (written many years
later). Whatever the case it nonetheless was a moment of great pivotal
importance both to Darwin and (presumably) to his theory as well.
5"struggle for existence" and "long-continued . .and plants" from
Darwin (1969:120).
6"at once struck ... to be destroyed" from Darwin (1969:120).
7"I had at last... to work." from Darwin (1969:120).
8"allowed myself ... in pencil" from Darwin (1969:120).
sFor more information about Darwin's illness you can consult "To Be an
Invalid" by Ralph Colp, Jr. (1977), The University of Chicago Press, Chi-
cago.

LITERATURE CITED

BOWLER, P. J. 1984. Evolution, the history of an idea. University of Cali-
fornia Press, Berkeley.
DARWIN, C. R. 1962. The voyage of the Beagle. Natural History Library
Edition, Doubleday & Co., Garden City. [First published 1839]
1969. The autobiography of Charles Darwin. Edited by Nora
Barlow. W. W. Norton & Co., Inc., New York. [Written by CRD in
1876]
1980. The red notebook of Charles Darwin. Edited by Sandra
Herbert. Cornell University Press, Ithaca.

Insect Behavioral Ecology-'84 Reiskind 17

GALE, B. G. 1982. Evolution without evidence: Charles Darwin and the
origin of species. University of New Mexico Press, Albuquerque.
KOHN, D. 1980. Theories to work by: rejected theories, reproduction, and
Darwin's path to natural selection. Studies in the History of Biology
4: 67-170.
MAYR, E. 1977. Darwin and natural selection: how Darwin may have dis-
covered his highly unconventional theory. American Scient. 65:
321-27.
RUSE, M. 1979. The Darwinian revolution: science red in tooth and claw.
University of Chicago Press, Chicago.

18 Florida Entomologist 68(1) March, 1985

THE PROLIFERATION OF CRYPTIC SPECIES IN
CHRYSOPERLA GREEN LACEWINGS THROUGH
SONG DIVERGENCE

CHARLES S. HENRY*

SYNOPSIS
Green lacewings of the economically important genus Chrysoperla pro-
duce vibratory signals with their abdomens during courtship and mating.
These signals or "songs" are complex and species-specific, and must be ex-
changed in a reciprocal manner between the male and female of a courting
pair before copulation will take place. Such a system of communication could
have evolved in response to sexual selection, selection from species isolation,
or both. Sexual selection requires reasonably high variance among indi-
viduals in reproductive success, manifested as a polygamous mating system.
The sympatric sibling species Ch. plorabunda and Ch. downesi, taken as
representative of the genus, exhibit relatively low polygamy of females but
extreme polygamy of males. However, because of irreversible sperm deple-
tion in males, effective lifetime ability to produce offspring is approximately
the same in both males and females, resulting in relatively low variance in
reproductive success regardless of sex. Consequently, sexual selection alone
is probably insufficient to account for the evolution and differentiation of the
observed songs. On the other hand, experiments testing choice by females of
conspecific versus heterospecific songs and courtship partners demonstrate
clearly that calls are of the utmost importance in the reproductive isolation
of closely related, potentially interfertile species. Other laboratory studies
reveal that the basis for call differences between species like Ch. plorabunda
and Ch. downesi is polygenic yet still simple, indicating that complete re-
productive isolation and hence speciation in the genus need not involve much
genetic change. Thus it is not entirely unexpected that in certain regions,
such as the mountains of western North America and central Europe, exist
several physically indistinguishable but acoustically unique "song morphs"
of the carnea-plorabunda-downesi complex that seem to be reproductively
isolated from their often sympatric relatives only by their calling behavior.
I suggest here that small, chance alterations in the genes controlling call
patterns can catalyze speciation in such lacewings by preventing normal
individuals from dueting and copulating with mutailts. Sexual selection
probably facilitates the process of divergence by assuring that individuals
prefer partners with extreme and therefore more attractive calls. Thus,
species proliferation within the complex can occur without obvious adaptive
or historical cause.

INTRODUCTION

Courtship in animals has always fascinated human beings. Because we,
too, are animals, many remarkable parallels exist between our own sexual
behavior and that of other mammals, reptiles, birds, and even insects. With

Charles S. Henry is an Associate Professor of Biology within the Systematics and Evolu-
tion Section of the University of Connecticut's large biology group. He is addicted to the
Neuroptera, that peculiar order of insects nobody knows much about; he blames that addic-
tion on Frank M. Carpenter, under whose supportive supervision he completed his Ph.D.
work. Past and present research ranges from the morphology, systematics and life histories
of Ascalaphidae, including their unique ant barriers for their eggs, to the acoustic behavior
and reproductive biology of green lacewings. Current address: Box U-43, The Biological
Sciences Group, The University of Connecticut, Storrs, CT 06268.

Insect Behavioral Ecology-'84 Henry 19

not a small dose of voyeurism, we observe, describe, categorize and interpret
the intricate interactions between the sexes that often characterize even the
most primitive organisms. Yet the same parallels that spark our interest in
animal courtship also color or badly distort our interpretations of its ulti-
mate functional significance, and we are too apt even as naturalists and
scientists to anthropomorphize what we observe. The struggle against this
human weakness has led many students of animal behavior to direct their
work away from "dangerous" areas requiring subjective interpretation or
speculation. For example, a reductionist approach can avoid such problems,
and the vast literature on the neurophysiological bases of various behavioral
responses and fixed action patterns attests to the popularity and validity of
that solution (see Punzo, this volume). Alternatively, one can make the
assumption that human and animal behavior are fundamentally different,
cannot be compared, and therefore exist for different reasons. Such a view
underlies most of the work on animal mating systems done prior to about
1975-not coincidentally, the date of publication of Wilson's Sociobiology
and Alexander's contribution to Insects, Science and Society. Although the
reproductive behavior of both humans and "other animals" is complex and
often rather stereotyped, it was felt by these earlier workers that in animals
such complexity prevented the individuals of two closely related species from
interbreeding and thereby producing inviable or unfit hybrid offspring,
while in humans it served a social or cultural function. Thus, most courtship
features in animals were thought to exist as "intrinsic, premating reproduc-
tive isolating mechanisms" (Mayr 1963 and refs.). In fact, it was felt that
courtship differences between two closely related species would be enhanced
by the action of natural selection against poorly adapted hybrids, and that
such a process could be construed as "character displacement" in zones of
overlap (sympatry) between the close relatives (Wilson and Brown 1956).
Biologists spent a great deal of time looking for and describing these species-
specific, ethological reproductive isolating mechanisms in the acoustical
signals of cicadas (Alexander 1956), crickets (Walker 1964), frogs (Blair
1974), and birds (Thielcke 1969); in the flash-patterns of fireflies (Lloyd
1969); in the chemical signals of fishes (Bardach and Todd 1970), moths
(Jacobson 1972), and mammals (Eisenberg and Kleiman 1972); and in the
visual displays of butterflies (Silberglied 1977), birds. (Murray 1971) and
lizards (Ferguson 1977).
A reaction against such dogma was bound to set in, given the human
proclivity simultaneously to reject the establishment and embrace new fads.
To be fair, it must also be mentioned that the accumulating data were not
always consistent with the reproductive barrier hypothesis. Not only were
many bizarre patterns of courtship difficult to explain as species isolating
mechanisms, but also the whole concept of character displacement was found
to be questionable as repeated attempts to find real examples of the phe-
nomenon met with failure (Grant 1972, Walker 1974). Consequently, today
our understanding of mating systems in all animals, including humans, has
been revised completely, so that the evolution of such systems is now viewed
more from the perspective of the individual and its personal fitness than
from that of the species and its mode of origin (Thornhill and Alcock 1983,
West-Eberhard 1983). Courtship can be thought of as an examination or
test; each individual must correctly answer a series of questions proposed

Florida Entomologist 68 (1)

March, 1985

by its potential mate, and will not be accepted as a mate until the whole
test has been completed with a passing grade. Thus, more basically, courtship
serves to reveal to each individual vital information about its partner: its
presence at a particular location, what species he or she is, and how good a
mate he or she will make. In this way, mistakes that are costly to the indi-
vidual in terms of wasted reproductive effort or inferior progeny can be
avoided (Halliday 1978). Species recognition, then, is only one relatively
small part of why an individual chooses a particular mate. Of greater im-
mediate importance in many species is the ability of an individual to con-
vince potential sexual partners, by honest or deceitful means, of his or her
desirability as a mate, through the advertisement of morphological and be-
havioral characteristics and through competition by members of one sex for
access to the other sex. Since the cost per copulation is generally lower in
sperm-donating males than it is in egg-producing females, it behooves (in
an evolutionary sense) "less valuable" males to advertise or compete more
vigorously for females than it does for females to do so for males. Thus a
special form of natural selection, called sexual selection (Darwin 1871), acts
(usually) on males in particular to accentuate the sexually important char-
acteristics of many animal species. In theory, the complexity and form of
courtship behavior should be especially sensitive to the action of sexual
selection (West-Eberhard 1983).
As numerous contributions over the last few years to this symposium
series attest (e.g. Thornhill 1980, Lloyd 1981, Rutowski 1982, Waage 1983,
Wing et al. 1983, Sivinski 1984), this total restructuring of our thinking
about mating systems has had a most stimulating and positive effect on
studies of animal-and especially insect-behavioral ecology and evolution.
However, the glitter of this "new biology" has sometimes blinded us to the
truths laboriously documented in earlier works (compare the earlier with
the later papers of R. D. Alexander, for example). What if complex be-
havioral patterns associated with courtship in at least some animal species
really do function principally as premating barriers to reproduction between
closely related taxa? We are presently so predisposed to think of everything
in the context of sexual selection that such a possibility might not even be
considered. Yet in the green lacewings that I have studied over the last
seven years, it is looking more and more likely that just such an antiquated
view may best describe the evolution of the complex courtship songs char-
acteristic of the genus Chrysoperla Steinmann.
Identification of bona fide reproductive isolating mechanisms in the court-
ship displays of closely related insect species is an exciting prospect, because
it pertains directly to the question of how species originate and proliferate.
The formation of new species-cladogenesis-is, to be sanguine, poorly
understood (Gould 1980, Lande 1980, Bush 1982, Ayala 1982). At least in
animals, the biological species concept (Mayr 1963) suggests that the
process is complete when successful hybridization between two sympatric
populations is no longer possible in nature because of premating or post-
mating barriers to reproduction. Therefore, analysis of reproductive isolat-
ing mechanisms and their relationship to the geographic distribution and
ecology of closely related species is a prerequisite to understanding specia-
tion. One important aspect of speciation is the degree of genetic divergence
accompanying the event; estimates in the literature, based on electrophoresis

Insect Behavioral Ecology-'84 Henry

(Avise 1974, Ayala 1975), DNA hybridization (King and Wilson 1975), and
mendelian analysis (Danforth 1950, Tauber and Tauber 1977a, b), range
from negligible to prodigious, supporting radically different theoretical
models of speciation (Futuyma and Mayer 1980, Bush 1982). It is seldom
possible, however, to assess the genetic basis of reproductive isolation itself,
since (a) the precise nature of the barriers is unknown or (b) the two
species cannot be forced to mate or produce offspring. As I will describe
here, both problems have been overcome in my studies of a song-producing
species-complex of green lacewings. Several morphologically very similar
species of Chrysoperla interbreed freely in the laboratory, yet remain re-
productively isolated from one another under natural conditions in areas
of sympatry. In two of these species, Ch. plorabunda (Fitch) and Ch.
downesi (Banks), I have been able to establish that different call patterns
serve as effective reproductive barriers between the taxa. Mendelian analysis
in hybrids and backcrosses of these species has permitted assessment of the
genetic basis of reproductive isolation. Finally and most significantly, field
studies have identified several distinctive "song morphs" of both Ch. plora-
bunda and Ch. downesi in the mountains of western North America and
central Europe, suggesting that proliferation of cryptic species can occur
through song divergence alone.

THE PROTAGONISTS: CHRYSOPIDAE
One of the most remarkable things about the acoustical behavior of green
lacewings (family Chrysopidae within the archaic holometabolous order
Neuroptera) is that it was not noticed earlier. The green lacewings are
common, widespread, easily collected insects extensively used for the bio-
logical control of crop pests; their predatory habits as larvae and (some-
times) adults and the ease with which they can be reared in the laboratory
suit them well for artificially regulating populations of aphids and other
plant lice on grapes, corn, and cotton (New 1975). Consequently, many
aspects of their reproductive biology and life histories have been well
studied, since efficient mass-rearing of adults has commercial application
(Smith 1922, Tauber 1974, Tauber and Tauber 1982b and refs., Hagen et al.
1970). However, only within the last few years have entomologists learned
of the richness of song repertory within certain species-groups of lacewings
or of the great importance of these signals for successful courtship and
copulation of individuals (Henry 1979 and later papers).
Green lacewings produce their "acoustical" signals and songs by oscillat-
ing their abdomens up and down in such a way that the low-frequency
vibrations are transferred to lightweight substrates like leaves, blades of
grass, or conifer needles. The abdomen does not strike the substrate (Henry
1980b), so the insect cannot be said to drum; rather, it "tremulates," as do
many acoustically silent ensiferan grasshoppers (Morris 1980) and spiders
(Rovner and Barth 1981), shaking its perch with the motions of its body.
One can record this behavior quite easily, by placing a high-output ceramic
piezoelectricc) monaural children's record-player cartridge in contact with
a plastic sarann wrap) membrane stretched over the top of a paper coffee
cup: the insect is placed inside the cup, and when he or she tremulates on
the saran wrap the signal can be recorded on cassette tape (Fig. 1). To get

Florida Entomologist 68 (1)

March, 1985

Fig. 1. System for recording the vibratory songs of green lacewings. The
playback recorder is used to drive the loudspeaker with a previously taped
signal in order to induce singing in the caged insect. For best results, a
monaural, ceramic cartridge with shielded cables should be used.

things started, it is often useful to include a receptive member of the op-
posite sex with the original insect, since both sexes generally produce iden-
tical songs (Henry 1984b) and will duet for some time with one another
prior to mating. Alternatively, a previously recorded tape of a singing
conspecific can be played to the insect in the cup through a small low-
frequency loudspeaker suspended above the chamber (Fig. 1) : the lacewing
will then synchronize its calls with the recording and'its "answers" can be
taped on another recorder.
Not all species of lacewings call prior to mating. For example, tremula-
tion during courtship in the chrysopid genera Chrysopa Leach (sensu
stricto, Mallada Navas, Ceraeochrysa Adams, and Leucochrysa Banks
seems optional and perfunctory and confined to the male sex (Henry 1979
for review, Henry 1982b). However, acoustical behavior is especially well
developed in Chrysoperla, a genus containing many of our common and
agriculturally important "stinkless" green lacewings such as Ch. carnea
(Stephens), Ch. plorabunda, and Ch. rufilabris (Burmeister) that over-
winter as adults. In these taxa, both sexes of a given species always ex-
change identical (usually) signals in a reciprocal, "polite" manner before
copulating. The signal exchanged between courting individuals is defined as
the shortest repeated unit (SRU) of the insect's call. Depending upon the
species, this unit may be very brief and relatively simple, consisting of just
one volley of abdominal vibration, or it may be long and complex, lasting

Insect Behavioral Ecology-'84 Henry

many seconds and made up of numerous volleys of one or more different
types (Henry 1980a). Lacewing songs, then, can vary in many different
ways, including not only the length, oscillation frequency, spacing, and
amplitude (intensity) characteristics of individual volleys but also the
length and volley structure of the shortest repeated units (Henry 1984a and
Fig. 2).
Two of the best studied species of Chrysoperla with respect to calling
pattern are the sympatric siblings Ch. plorabunda (formerly synonymous
with Ch. carnea) and Ch. downesi. These widely distributed North American
taxa differ in their ecological characteristics and color intensity but other-
wise cannot be told apart by conventional morphological criteria, including
wing structure and male genitalia (P. A. Adams and E. G. MacLeod, per-
sonal communications). Ch. plorabunda is a meadow species whose light
green pigmentation matches the overall color of its habitat; it also has sev-
eral generations each year at north temperate latitudes. In contrast, the
darker-green Ch. downesi inhabits pine and spruce forests and blends well
with the more somber hues of conifer needles; adults remain in reproductive
diapause from their emergence in June until they breed the following spring,
resulting in a univoltine (one generation per year) life cycle. The geo-
graphical ranges of the two species overlap extensively through much of
cold-temperate North America, but Ch. downesi seems to occur further
north and Ch. plorabunda much further south than their counterparts
(Bickley and MacLeod 1956). Although the two taxa interbreed fairly
readily in the laboratory, hybrid individuals have not been shown to exist
in nature (Tauber and Tauber 1981, Henry in press). Most striking are
the differences in their songs: Ch. plorabunda individuals rapidly exchange
(i.e., alternate) single, short volleys of abdominal vibration with one an-
other (Fig. 2 and 3, PI), while each courting Ch. downesi (Fig. 2 and 4,
D1) produces a long, distinctive sequence of volleys that is then answered
in kind by its partner (Henry 1980a). These sibling species also differ in
all the finer details of their calls, and show very little variation in their

SRU
D D---Ds
AE ,E C

DOWNESI

BL
PLORABUNDA C
-E.SRU

Fig. 2. Comparison of the calls of Chrysoperla downesi and Ch. plora-
bunda, redrawn from oscilloscope tracings, showing the principal measurable
features of the signals. SRU = shortest repeated unit; other letters repre-
sent homologous features of the two calls discussed in other papers (e. g.,
Henry 1984a).

Florida Entomologist 68 (1)

March, 1985

1 *,. u. Uscillographs of abdominal vibration patterns produced by sex-
ually receptive individuals in the plorabunda complex of Chrysoperla in
North America, at around 280 C. Pictures at right are X10 details of whole
calls shown at left (writing speed = 2.00 sec/div. left, 0.20 sec/div. right).
P1 is standard Ch. plorabunda, while P2 and P3 are song morphs from the
Sierra Nevada of western North America.
songs from individual to individual across most of their respective ranges
(Henry in press). In the mountains of western North America, however,
things get confusing. Here, as I will describe in detail later, exist at least
two additional, morphologically indistinguishable forms of Ch. plorabunda,
and two more of Ch. downesi, that are characterized by unique songs and
that probably constitute additional, "hidden" species of the plorabunda-
downesi complex (Henry 1984b and in press). Oscillographs of their calls
are shown in Figures 3 and 4, next to those of "real" Ch. plorabunda (P1)
and Ch. downesi (Dl) for comparison. In the plorabunda group (Fig. 3),
the major obvious differences between the songs of 'the newly discovered
morphs and that of the familiar species are longer volleys and greater
volley spacing in the morphs; additionally, morph P3 differs from the others
by the fact that its shortest repeated unit includes three to five of these long
volleys rather than just a single one. Similarly, morph D2 in the pair of
downesi relatives (Fig. 4) produces distinctly longer and more spaced out
volley-, than the familiar form, although D3 shows the opposite tendency
with -horter, more closely spaced volleys and no sign of the long introduc-
tory pulses seen in D1 and D2. Strangely, the ecological differences so
typical of Ch. plorabunda and Ch. downesi in eastern North America seem
to disappear in the mountains of the western part of the continent, since all
song morphs were collected from the same low branches of douglasfir, in-
cense cedar, and giant Sequoiadendron. It is no wonder that lacewing
taxonomists have despaired of naming and describing the western repre-
sentatives of the "carnea species-complex" (Tauber and Tauber 1982b,
Adams pers. comm.).

Insect Behavioral Ecology--84 Henry

Fig. 4. Oscillographs as in Figure 3, of Ch. downesi (D1) and its song
morphs (D2 and D3) in western North America.

Not so long ago it was believed, for good morphological reasons, that
North American Ch. plorabunda was really just part of a single, vast cos-
mopolitan species called Ch. carnea, so the former was quietly synonymized
with the latter (Tjeder 1960). As a consequence, scientists throughout the
world in such divergent fields as physiology (Miller and Olesen 1979), be-
havior( Duelli 1980, Principi and Canard 1984), taxonomy (Holzel 1970,
Bullini and Cianchi 1984), and biocontrol (New 1975) have assumed that
they were working on one and the same insect species. That this is not the
case became clear from studies of calling behavior and hybridization po-
tential in North American and European populations of "Ch. carnea," so
the plorabunda name has been reinstated for nearctic representatives of the
taxon (Henry 1983, Duelli unpublished ms.). Now it seems that just as we
must re-evaluate the New World taxonomic situation because of the dis-
covery of new, cryptic song morphs of Ch. plorabunda and Ch. downesi, so
also must we alter our view of what constitutes Ch. carnea in the Old World.
In geographically heterogeneous central Europe, for example, at least three
acoustically distinctive morphs of Ch. carnea occur sympatrically, or nearly
so. One of these, referred to here as "Basel camnea" (Fig. 5), appears to be
the most common lowland form of the species and has been found in
meadows and oak scrub from France and Germany south at least to the
Atlas Mountains of northern Africa. Its song consists of a relatively long
sequence of three to fifty or more volleys of abdominal vibration; the partner
of a courting individual responds in kind only after the sequence is complete,
rather than alternating volley for volley as does Ch. plorabunda. A totally
different call characterizes Ch. carnea from an area south of the Swiss Alps,
near Italy: this morph, "Ticino carnea," produces a long train of short volley
pulses that are issued so rapidly that they blend to create a continuous, mod-
ulated tone (Fig. 5). Finally, from another region of Switzerland called the
"Pfynwald" (pine forest) between the two major chains of the Alps, a third,
darker green morph of Ch. carnea occurs sympatrically with Basel carnea.
It sings in a manner similar to Ticino carnea, but differs from the latter in

26 Florida Entomologist 68(1) March, 1985

CH. CARNEA (EUROPE)

Fig. 5. Oscillographs as in Figures 3 and 4, of the carnea complex of
Chrysoperla in central Europe. Basel and Pine carnea are sympatric with
one another in an area of Switzerland called the Pfynwald.
its shorter calls and more sharply delineated volleys (Fig. 5). This form,
known as "pine carnea," is a European counterpart of Ch. downesi, and like
its North American analogue is restricted to coniferous habitats (Duelli
pers. comm.). Obviously, extensive divergence of call characteristics is oc-
curring in a parallel manner in Chrysoperla of both hemispheres, but
whether or not full species status has been attained by the various song
morphs cannot be decided until we can document the existence of reproduc-
tive isolation among them.

LACEWING SONGS AND REPRODUCTIVE ISOLATION
In light of what I have just said about courtship and mating, one can
think of several possible explanations for why lacewings in the carnea-
plorabunda-downesi complex sing as they do and with such extraordinary
variety. Principally, these explanations fall into two categories, sexual
selection and species recognition-although note that some workers con-
vincingly argue that the distinction between the two is more convenient than
real (Lloyd in press). A third possibility is that both sexual selection and
selection for reproductive isolation combine together in some way to pro-
duce the patterns that we see in nature. Each of these hypotheses will be
examined in detail below.
That song repertoires of green lacewings could originate and diversify
in response to sexual selection has been suggested, or even assumed, by
Tauber and Tauber (1982a). Female or (less likely) male choice of vigorous
males (females) capable of signaling often and enthusiastically would en-
courage fixation of genes responsible for elaborate call patterns, and these
patterns might come to vary from species to species either by chance or in
response to the different sound transmission properties of the habitats oc-
cupied by different lacewing taxa. In addition, acoustic competition among

Insect Behavioral Ecology-'84 Henry

members of one sex for access to the other could accelerate and intensify
this process, because the loudest or most persistent singers will mate more
often and therefore contribute genetically to the next generation out of
proportion to their numbers. If this view is correct, the sex showing the
sexually selected features should also exhibit high variance among indi-
viduals in reproductive success-that is, some individuals should mate much
more often than others and produce many more viable offspring from those
matings. Thus, a high degree of polygamy should characterize the sex in
question. In lacewings, calling behavior is equally well developed in both
sexes, suggesting that males should be polygynous and females polyandrous
for sexual selection to have produced the observed patterns. Both sexes of
Ch. plorabunda and Ch. downesi do exhibit polygamy: individual males in
the lab will continue to copulate at < 24 hour intervals indefinitely (Table
1), while females regularly copulate an average of four times (range = 1 to
7, see Table 2) before senescing. However, if these mating patterns are ex-
amined for the number of progeny produced per copulation, a different
picture emerges. With virgin males as consorts, most females fertilize well
over half of their lifetime egg production with the sperm of just one-
usually her first-partner; a second male, accepted as a mate only after 80
percent of the female's previous batch of eggs has been oviposited, sires
most of the remaining progeny (Table 2). In fact, one Ch. downesi female
laid 991 fertile eggs in 47 days after mating just once, while the corre-
sponding record for Ch. plorabunda is 891; these numbers equal or exceed
any in the literature, for any number of matings (Rousset 1984). The most
surprising discovery, however, is that males share this pattern with females,
despite their capacity to mate many times more often. Although the data
are few, they strongly suggest that only two females derive high fertiliza-
tion potential from a given male, regardless of his copulatory prowess
(Table 1). This finding agrees with Sheldon and MacLeod's (1974) that
spermatogenesis in Ch. plorabunda is completed during the pupal stage,

TABLE 1. SUCCESS AT SIRING PROGENY BY MALES OF Chrysoperla downesi
EACH MATED AT 24-HOUR INTERVALS.

Consecutive
matings Fertile eggs produced per copulation
(every 24 hours) male DM-lx male DM-V-2

first 680 510
second 549 27
third 4 645
fourth 50 85
fifth 1 42
sixth 49 9
seventh 0 15
eighth 1 2
ninth 0 12
tenth 0 0

Total 1334 1347

Florida Entomologist 68 (1)

March, 1985

before any mating can have taken place: males therefore cannot manu-
facture new sperm to keep pace with copulations. Thus the maximum num-
ber of offspring that can be produced by either a male or a female
Chrysoperla spp. during its lifetime is probably 1,200 to 1,400 individuals
(Tables 1 and 2). One conclusion to be drawn from this sketchy picture of
lacewing reproductive biology is that both sexes are only mildly polygamous,
at least in a practical, functional sense. Consequently, sexual selection alone
should not act very strongly to promote morphological or behavioral modi-
fication in these insects, and it should affect both sexes more or less equally.
As an explanation for lacewing singing, it seems inadequate.
The second hypothesis proposes that lacewing calls serve in species rec-
ognition and reproductive isolation. This would explain why the calls of
closely related species are so different, but begs the question of how these
differences arose in the first place. However, if calls can be shown to be
the principal barriers to reproduction between close relatives, then pre-
sumably the divergence of call characteristics is related in some causal
manner to the actual speciation process in Chrysoperla. The best evidence
pertaining to these issues once again comes from the sibling species Ch.
plorabunda and Ch. downesi in North America (Henry 1984b and in press);
we can then hope that this information is more generally applicable to other
singing Chrysoperla. Since Ch. plorabunda and Ch. downesi will interbreed
if given no other option in a laboratory situation, reproductive isolation
between them was tested by offering the insects choices that each might en-
counter in nature. One approach was to play back tape recordings of con-
specific and non-conspecific calls to individual females of each species, and
to note whether or not the insect in question responded to a particular
recording by dueting with it. Although a few individuals refused to respond
to any such disembodied signal, the majority (30 of 37 plorabunda, 32 of 37
downesi) established prolonged duets exclusively with tapes of conspecific
calls and ceased calling in the presence of signals of the other species
(Henry in press: Table 1). A more realistic test allowed a female lacewing

TABLE 2. FERTILE EGGS PRODUCED PER MATING FOR Chrysoperla plorabunda
AND Ch. downesi FROM EASTERN NORTH AMERICA. ONLY INDI-
VIDUALS PRODUCING 500 OR MORE EGGS DURING THEIR LIFETIMES
ARE INCLUDED.

Maximum number of
Average number of fertile fertile eggs for an
eggs produced per individual individual

Chry- Chry-
Chrysoperla Chrysoperla soperla soperla
plorabunda downesi plorabunda downes'
First mating 468 273 (7) 563 256 (14) 891 991
Second mating 327 272 (6) 330 + 244 (9) 777 824
Third mating 147 156 (4) 142 84 (3) 359 238
Fourth mating 16 ---. (1) 274 ---(1) 22 274

Lifetime total 852 172 (7) 825 196 (14) 1207 1286

Insect Behavioral Ecology-'84 Henry

to choose a sexually receptive male of her own or the other species as a mate.
In these experiments, the correct (i.e., conspecific) choice was made in 47
out of 49 trials, and the two heterospecific matings were obviously abnormal
in the sense that they involved older females of Ch. downesi which had
previously shown no sign of sexual receptivity (Henry in press: Table 2).
All of these data are consistent with the songs of green lacewings being
behavioral barriers to copulation between closely related taxa. This in-
terpretation is strengthened by the absence of any real postmating barriers
to mating, at least between Ch. plorabunda and Ch. downesi: interspecific
matings are fertile, and hybrids can reproduce among themselves or with
their parents. In addition, other possible premating isolating mechanisms
like differences in habitat or breeding season (Tauber and Tauber 1977a, b)
have subsequently been shown to be imperfect and not capable of preventing
hybridization (Henry 1980a, 1982a, 1984b and in press). Thus it would seem
that songs exist as the only obvious source of interspecific reproductive
isolation, at least in some sibling species of Chrysoperla.
To understand how call differences arose between populations of
Chrysoperla in the first place, it is probably necessary to invoke sexual
selection in combination with selection for reproductive isolation: hy-
pothesis three. As West-Eberhard (1983) has pointed out, the syrr'-i'm of
these two types of selection, mixed with a pinch of chance (mutation) and a
dash of geographical isolation, can theoretically encourage especially rapid
speciation in certain kinds of semi-social organisms. More will be said about
this idea after I have discussed the status of the other song morphs of
Chrysoperla, at the end of this paper.

THE GENETICS OF REPRODUCTIVE ISOLATION
Since different species of Chrysoperla can be induced to hybridize, and
since their calls normally prevent this from happening, one can study what
happens to distinctive, species-specific songs under a hybridization program
and get some idea of the genetic basis of reproductive isolation. When this
is done for the species pair Ch. plorabunda and Ch. downesi (Henry 1984b
and in press), all first-generation (Fl) hybrid individuals produce invariant
calls that are precisely intermediate in their several features between those
of the parents, while the songs of second-generation (F2) hybrids are much
more variable among individuals and even match the call phenotypes of one
or the other of the parents about four percent of the time (Henry in press:
fig. 2). Progeny of backcrosses show variable call features skewed toward
those characteristic of whichever species is serving as a parent, and the few
F3 hybrids analyzed sang much like F2's. Clearly, these data indicate that
singing patterns in this pair of sibling species are under the control of a
simple polygenic system consisting of perhaps two to four homologous loci,
each of which is homozygous (fixed) for alleles that are different in the two
taxa. Homozygosity within each species is inferred from lack of variance in
parental and F1 call phenotypes, while polygeny best explains the noif-on
distribution of those phenotypes in the F2 generation. As expected for
species in which both males and females sing similarly, sex linkage in the
inheritance of song traits is absent: reciprocal crosses yielded identical re-
sults.
Thus, behaviorally-based reproductive isolation between closely related

30 Florida Entomologist 68(1) March, 1985

sibling or cryptic species of Chrysoperla rests on a very simple genetic
foundation. This absence of much genetic differentiation between bona fide
species probably characterizes all aspects of the Chrysoperla genome, since
the ease of laboratory hybridization between Ch. plorabunda and Ch. downesi
and among several western American song morphs (Table 3) certainly re-
flect overall genetic similarity. Even those non-behavioral features by
which Ch. plorabunda and Ch. downesi differ have been shown to be under
the control of simple gene systems: for example, the intensity of green
pigmentation is regulated by two alleles at one locus, while responses to
photoperiod are mediated by gene substitutions at two loci (Tauber and
Tauber 1977a, b). Given these data, it seems that full species status within
the carnea-plorabunda-downesi complex of Chrysoperla can be attained by
local populations without the kind of genetic revolution assumed necessary
by many evolutionary biologists (Mayr 1963, Lewontin 1974, Ayala 1975).
In addition, those features responsible for keeping the species separate are
clearly non-adaptive, implying that speciation can be at once simple in a
genetic sense, rapid in an evolutionary one, and random or stochastic in its
ultimate cause.

SONG MORPHS OF Chrysoperla

If song phenotype varies significantly and in a continuous manner over
the geographical range of a lacewing species, one can simply assume that a
character dine exists and label the species polytypic. On the other hand,
discrete, non-continuous variation in song type suggests the existence of
partial or complete reproductive isolation among populations possessing
those different songs. Sympatric occurrence of two or more of such "song
morphs" without apparent hybridization among them argues strongly for
the recognition of the morphs as separate biological species, no matter how
physically similar to one another they might be. In lacewings of the genus
Chrysoperla, what was once thought to represent polymorphism within
single, widely distributed species like Ch. carnea or Ch. plorabunda is now
looking more and more like local species proliferation, especially in light of
findings presented here and elsewhere on the functional significance and

TABLE 3. EXTENT OF SUCCESSFUL HYBRIDIZATION BETWEEN VARIOUS SPECIES,
SONG MORPHS AND GEOGRAPHICAL POPULATIONS WITHIN THE Chry-
soperla plorabunda-downesi COMPLEX FROM NORTH AMERICA, EX-
PRESSED AS A RATIO (SMALL SAMPLE SIZE) OR PERCENTAGE (n >
10). PRODUCTION OF VIABLE F1 LARVAE CONSTITUTED SUCCESS IN
THESE EXPERIMENTS.

.....------ ..... ..........-------------..M ale parent -_. ---------...........-.---------- --
SP1 P2 P3 D1 D2 D3
2 P1 70-90% 4/9 a10% 1/3 0/1
pP2 3/9 z80% 0/5 0/2 0/2
P3 0/5 5/5 0/1
D1 z30% 70-90% -
SD2 1/1 1/3 6/6 -
D3 0/2 0/1 10/10

Insect Behavioral Ecology-'84 Henry 31

genetic basis of singing behavior and on the occurrence of cryptic song
morphs in the mountains of North America and central Europe.
It has now been documented that Ch. plorabunda and Ch. downesi each
maintains its species-specific, invariant call type over vast geographical
stretches of North America. For example, as partially shown in Fig. 5,
individuals exhibiting "typical" Ch. plorabunda signals (P1) have been
collected as far west as extreme western Idaho; east of there, from Canada
at least to Virginia, no variation in this standard pattern has been detected
(unpublished data). Typical Ch. downesi songs (Dl) are even more wide-
spread, occurring from eastern Canada and New England at least to
western Montana, northern Idaho, and the Cascade Mountains of southern
Oregon southward through most of the Sierra Nevada of California (Fig.
5). However, in geographically heterogeneous areas such as the northern
Rocky Mountains and the rugged mountainous regions of Oregon and Cali-
fornia, the typical song morphs of Ch. plorabunda and Ch. downesi are
mixed with or replaced by other song types that are physically identical to
one or the other of the common species but whose calls are distinctive (Figs.
3 and 4). As described earlier, the calls of these morphs are characterized
by easily defined differences that are consistent over wide geographical
areas, and where several different call types exist together, intermediate
forms indicative of hybridization are absent. For example, the P2 morph of
Ch. plorabunda, whose call is the most similar of any morph to its "parent"
species (Fig. 3), completely replaces Ch. plorabunda in a strip nearly 1,000
miles long at high elevations in the coniferous forests of Oregon and Cali-
fornia (Fig. 6). On the other hand, the more acoustically distinctive morph
D2 of Ch. downesi co-occurs microsympatrically with the nominate form
from northwestern Montana to central California, while the two rarer and
most unusual singers, P3 and D3, show overlapping or coincident distribu-
tions with P2 and D2, respectively, in different parts of central and southern
California.
The situation across the Atlantic is parallel to that in North America.
Ch. carnea, synonymous with the "Basel carnea" song morph described in
this paper, is geographically widespread and acoustically invariant through-
out its range; so far, it has been collected all over lowland Europe and north
Africa. Yet in the mountains of Switzerland, it is replaced south of the Alps
by the very distinctive song morph called Ticino carnea and occurs sym-
patrically in the Pfynwald pinewoodss) with downesi-like pine-carnea (Fig.
7). Just as in North America, sympatry is always correlated with better-
defined call differences: pine-carnea sings very differently from Basel
(standard) carnea, yet shows considerable similarity in its call structure
to Ticino carnea on the other side of the Alps (Fig. 5). Morphological and
hybridization studies by P. Duelli and electrophoretic ones by L. Bullini
(both unpublished) strongly suggest that Basal and pine carnea, at least,
are well-defined taxa deserving full species status; the precise taxonomic
relationship of Ticino to Basel carnea is less clear, although paradoxically
they seem less similar to one another by electrophoretic measures than are
pine- and Basel-carnea in the Pfynwald. Overall, however, electrophoresis
confirms what we have already concluded from morphology and inter-
fertility tests, that the whole carnea complex in Europe shows very little
genetic differentiation, just like the plorabunda-downesi complex in North

Florida Entomologist 68 (1)

S 400km

SONG MORPHS: PLORABUNDA-DOWNESI GROUP
Fig. 6. Map of western North America, showing where the various song
morphs of Ch. plorabunda and Ch. downesi have been collected. Stippled
areas are 2000 or more meters in elevation, while black represents mountains
higher than 3000 meters.
America. Indeed, extraordinary similarity links both complexes to one an-
other across the Atlantic: note, for example, an earlier and dramatic
pictorial demonstration of indistinguishable male genitalia in Ch. plora-
bunda, plorabunda's P2 song morph, and Ticino carnea (Henry 1983: fig. 2).

CONCLUSIONS AND SPECULATION
What does all this mean? In brief, we have learned that Chrysoperla

March, 1985

Insect Behavioral Ecology-'84 Henry

O BASEL and OAK CARNEA
* TICINO CARNEA
A PINE CARNEA

Fig. 7. Map of Switzerland, showing collection sites of three song morphs
of Ch. carnea. The Pfynwald is in southwestern Switzerland, where Basel
and Pine carnea occur together. Stippled areas are above 3000 meters.

lacewings sing, that singing is not confined to males, and that mating will
not take place unless the two partners duet with one another; that indi-
vidual males can mate many times more often than females, but been use of
irreversible sperm depletion exhibit maximum lifetime reproductive po-
tentials that are no higher than those of females; and that the PoP- pro-
duced during courtship are complex, stereotyped and species-specific and
in the sibling species pair Ch. plorabunda and Ch. downesi serve as the
major barrier to hybridization in the field. Further laboratory analysis has
revealed that the basis for call differences between species like Ch. plora-
bunda and Ch. downesi is polygenic yet still relatively simple, indicating
that complete reproductive isolation and hence speciation in the genus need
not involve much genetic change. And field work in rugged mountainous
regions of North America and Europe has uncovered numerous cryptic
(morphologically indistinguishable) populations within the carnea-plora-
bunda-downesi complex that possess their own unique song properties and
that may very well represent reproductively isolated species. I think that
all of these pieces of information fit together to produce an interesting,
harmonious evolutionary picture.
It seems probable that calling behavior and cladogenesis (species forma-
tion) are linked together in this complex of Chrysoperla green lacewings.
Rather than postulating long term historical isolation (Futuyma and Mayer
1980) or adaptive responses to local ecological conditions (Maynard Smith
1966, Bush 1975, 1982, Tauber and Tauber 1982b) as prerequisites for the
development of reproductive isolation between populations, I think it is
sufficient here to invoke changes in songs as the principal catalysts of the
speciation process (see Lloyd 1984 for a similar view). Imagine a single
ancestral lacewing species, in which elaborate singing behavior during

34 Florida Entomologist 68(1) March, 1985

courtship has become established and strongly canalized in both sexes by
the action of sexual selection, mediated through both female and male
choice. Based on the genetic and behavioral data for Ch. plorabunda and
Ch. downesi, one would expect that a simple, chance allelic substitution at
a call-controlling locus in one individual of that ancestral species could
radically alter the phenotype of its song, resulting in its being rejected as a
mate by conspecifics. Occasionally, however, local environmental conditions,
leading to very low population density, might force a conspecific to copulate
with the heterozygous mutant individual, producing a larger population
consisting of 50 percent hybrid singers. If, as seems likely, alleles for the
call feature are co-dominant, the F2 generation will include a small number
of pure mutant singers along with larger numbers of ancestral and hybrid
types. These extreme singers will choose to mate only among themselves
because of the necessity of dueting smoothly with their partners. Continued
assortative mating and sexual selection over several more generations should
cause the population eventually to consist only of ancestral and mutant
singers. This conclusion follows from two properties of hybrid individuals
with intermediate call phenotypes. On the one hand, such hybrids will be
more likely than purebred genotypes to make mistakes during courtship,
because they share call genes with both mutant and ancestral types. On the
other hand, it should be remembered that hybrids will produce a certain
number of extreme phenotypes no matter with whom they mate. In other
words, hybrid callers represent an unstable class of individuals that will
decay toward the more stable extremes over time. Since lacewings are also
gregarious insects, sexual selection in the form of social selection (West-
Eberhard 1983) probably accelerates the process of divergence as prefer-
ences develop within the population for singers with extreme and therefore
more attractive calls. The result is two species that sing in different ways,
for no apparent adaptive or historical reason.
The complexes of cryptic, sibling species of Chrysoperla lacewings can
be viewed as end products of this process. Concentration of species numbers
in mountainous regions suggests that geographical isolation of populations
is another necessary ingredient, but it is also possible that geographical
heterogeneity simply favors low population density at local sites, thereby
encouraging normal individuals to mate with mutants and produce hybrids.
Whatever the details, speciation is rapid and prolific and involves little
reorganization of the genome. Behavioral repertories orchestrate cladogene-
sis, and are often the only sign that the process has occurred.

ACKNOWLEDEMENTS

First, I would like to thank James Lloyd and the Florida Entomological
Society for inviting me to the 1984 symposium on insect behavioral ecology,
for which this contribution was prepared. Some of the work described here
was supported through the Systematics Program of the National Science
Foundation (DEB 77-12443 and DEB 79-11537), while the balance was
completed with aid from the Research Foundation of the University of
Connecticut. Many people contributed their ideas, critical skills, and even
physical labor to the manuscript: warm thanks are extended to Ray Pupedis,
Peter Duelli, Phil Adams, Ellis MacLeod, Ding Johnson, Steve Pacala, Ted
Taigen, Gary Shea, Christine Busher, Jane O'Donnell, Will Cook, and Burma

Insect Behavioral Ecology-'84 Henry

Stelmak. The E. N. Huyck Preserve in Rensselaerville, New York served as
principal field site throughout the many years of this study.

REFERENCES
ALEXANDER, R. D. 1956. A comparative study of sound production in in-
sects, with special reference to the singing Orthoptera and Cicadidae
of the eastern United States. Ph.D. dissertation, Ohio State Uni-
versity. 529 pp.
1975. Natural selection and specialized chorusing behavior in
acoustical insects. Pages 35-77 In Pimentel D. ed., Insects, Science and
Society. Academic Press, N. Y.
AvISE, J. C. 1974. Systematic value of electrophoretic data. Systematic
Zoology 23: 386-409.
AYALA, F. J. 1975. Genetic differentiation during the speciation process.
Evolutionary Biology 8: 1-78.
-- 1982. The genetic structure of species. Pages 60-82 In Milkman, R.,
ed., Perspectives on Evolution. Sinauer Associates, Sunderland, Mass.
BARDACH, J. E., AND J. H. TODD. 1970. Chemical communication in fish.
Pages 205-40 In J. W. Johnston, Jr., D. G. Moulton, and A. Turk,
eds., Advances in Chemoreception, vol. 1, Communication by Chemical
Signals.
BICKLEY, W. E., AND E. G. MACLEOD. 1956. A synopsis of the nearctic
Chrysopidae with a key to the genera (Neuroptera). Proc. Ent. Soc.
Washington 58: 177-202.
BLAIR, W. F. 1974. Character displacement in frogs. American Zool. 14:
1119-25.
BULLINI, L., AND R. CIANCHI. 1984. Electrophoretic studies on gene-
enzyme systems in chrysopid lacewings. Pages 48-56 In Canard, M.,
Y. S6m6ria, and T. R. New, eds., Biology of Chrysopidae. W. Junk,
The Hague.
BUSH, G. L. 1975. Modes of animal speciation. Annu. Rev. Ecol. and Syst.
6: 339-64.
1982. What do we really know about speciation? Pages 119-28 In
Milkman, R., ed., Perspectives on Evolution. Sinauer Associates,
Sunderland, Mass.
DANFORTH, C. H. 1950. Evolution and plumage traits in pheasant hybrids,
phasianus X chrysolosus. Evolution 4: 301-315.
DARWIN, C. 1871. The Descent of Man and Selection in Relation to Sex.
Modern Library, N. Y.
DUELLI, P. 1980. Preovipository migration flights in the green lacewing
Chrysopa carnea (Planipennia, Chrysopidae). Behav. Ecol. and
Sociobiol. 7: 239-46.
EISENBERG, J. F., AND D. G. KLEIMAN. 1972. Olfactory communication in
mammals. Annu. Rev. Ecol. and Syst. 3: 1-32.
FERGUSON, G. 1977. Display and communication in reptiles: an historical
perspective. American Zool. 17: 167-76.
FUTUYMA, D. J., AND G. C. MAYER. 1980. Non-allopatric speciation in
animals. Syst. Zool. 29: 254-71.
GOULD, S. J. 1980. Is a new and general theory of evolution emerging?
Paleobiology 6(1) : 119-30.
GRANT, P. R. 1972. Convergent and divergent character displacement. Biol.
J. of the Linnean Soc. 4: 39-68.
HAGEN, K. S., E. F. SAWELL, JR., AND R. L. TASSAN. 1970. The use of food
sprays to increase effectiveness of entomophagous insects. Pages
59-81 Proceedings of the Tall Timbers Conference on Ecological

36 Florida Entomologist 68(1) March, 1985

Animal Control.
HALLIDAY, T. R. 1978. Sexual selection and mate choice. In Krebs, J. R.,
and N. B. Davies, eds., Behavioral Ecology, Pages 180-213. Blackwell
Scientific Publications, Oxford.
HENRY, C. S. 1979. Acoustical communication during courtship and mating
in the green lacewing Chrysopa carnea (Neuroptera: Chrysopidae).
Ann. Ent. Soc. America 72(1): 68-79.
-- 1980a. The courtship call of Chrysopa downesi Banks (Neuroptera:
Chrysopidae) : its evolutionary significance. Psyche 86(2-3): 291-97.
1980b. The importance of low-frequency, substrate-borne sounds in
lacewing communication (Neuroptera: Chrysopidae). Ann. Ent. Soc.
America 73 (6) : 617-21.
- 1982a. Reply to Tauber and Tauber's "Sympatric speciation in
Chrysopa: Further discussion." Ann. Ent. Soc. America 75(1): 3-4.
1982b. Reproductive and calling behavior in two closely related,
sympatric lacewing species, Chrysopa oculata and Chrysopa chi
(Neuroptera, Chrysopidae). Proc. Ent. Soc. Washington 84: 191-203.
1983. Acoustic recognition of sibling species within the holarctic
lacewing Chrysoperla carnea (Neuroptera: Chrysopidae). Syst. Ent.
8: 293-301.
.1984a. Temperature-induced changes in the calls of the green
lacewing, Chrysoperla plorabunda (Neuroptera: Chrysopidae).
Psyche 90(4) : 343-60.
1984b. The sexual behavior of green lacewings. Pages 101-10 In
Canard, M., Y. Semeria, and T. R. New, eds., Biology of Chrysopidae.
W. Junk, The Hague.
(in press). Sibling species, call differences, and speciation in green
lacewings (Neuroptera: Chrysopidae: Chrysoperla). Evolution.
HOLZEL, H. 1970. Zur generischen Klassifikation der palaarktischen Chry-
sopinae. Eine neue Gattung und zwei neue Untergattungen der
Chrysopidae (Planipennia). Zeits. Arbeits. Oest. Ent. 22: 43-52.
JACOBSON, M. 1972. Insect Sex Pheromones. Academic Press, N. Y. 382 pp.
KING, M-C., AND A. C. WILSON. 1975. Evolution at two levels: molecular
similarities and biological differences between humans and chimpan-
zees. Science 188: 107-16.
LANDE, R. 1980. Microevolution in relation to macroevolution. Paleobiology
6: 235-38.
LEWONTIN, R. C. 1974. The Genetic Basis of Evolutionary Change.
Columbia University Press, N. Y.
LLOYD, J. E. 1969. Flashes of Photuris fireflies: their value and use in
recognizing species. Florida Ent. 52: 29-35.
1984. On deception, a way of all flesh, and firefly signaling and
systematics. Oxford Surv. Evol. Biol. 1: 48-84.
1984. On deception, a way of all flesh, and firefly signaling
and systematics. Oxford Surv. Evol. Biol. 1: 48-84.
MAYNARD SMITH, J. 1966. Sympatric speciation. American Naturalist 100:
637-50.
MAYR, E. 1963. Animal Species and Evolution. Belknap Press, Cambridge,
Massachusetts.
MILLER, L. A., AND J. OLESEN. 1979. Avoidance behavior in green lacewings.
1. Behavior of free-flying green lacewings to hunting bats and ultra-
sound. J. Comp. Physiol. 131: 113-20.
MORRIS, G. K. 1980. Calling display and mating behavior of Copiphora
rhinoceros Pictet (Orthoptera: Tettigoniidae). Animal Behaviour 28:
42-51.
MURRAY, B. G. 1971. The ecological consequences of interspecific territorial

Insect Behavioral Ecology-'84 Henry

behavior in birds. Ecology 52(3) : 414-23.
NEW, T. R. 1975. The biology of Chrysopidae and Hemerobiidae (Neuro-
ptera), with reference to their usage as biocontrol agents: a review.
Trans. Roy. Ent. Soc. London 127: 115-40.
PRINCIPI, M. M., AND M. CANARD. 1984. Predatory behavior of larvae and
cannibalism. Pages 92-101 In Canard, M., Y. S6mBria, and T. R.
New, eds., Biology of Chrysopidae. W. Junk, The Hague.
PUNZo, F. 1985. Insect learning: behavioral ecological implications. Florida
Ent. 68(17): 89-104.
ROUSSET, A. 1984. Reproductive physiology and fecundity. Pages 116-29 In
Canard, M., Y. S6m6ria, and T. R. New, eds., Biology of Chrysopidae.
W. Junk, The Hague.
ROVNER, J. S., AND F. G. BARTH. 1981. Vibratory communication through
living plants by a tropical wandering spider. Science 214: 464-66.
RUTOWSKI, R. L. 1982. Mate choice and lepidopteran mating behavior.
Florida Ent. 65: 72-82.
SHELDON, J. K., AND E. G. MACLEOD. 1974. Studies on the biology of the
Chrysopidae. IV. A field and laboratory study of the seasonal cycle
of Chrysopa carnea Stephens in central Illinois (Neuroptera: Chry-
sopidae). Trans. American Ent. Society 100: 437-512.
SILBERGLIED, R. E. 1977. Communication in Lepidoptera. In Sebeok, T. A.,
ed., How Animals Communicate. Indiana University Press, Bloom-
ington, Indiana.
SIVINSKI, J. 1984. The behavioral ecology of vermin. Florida Ent. 67(1):
57-67.
SMITH, R. C. 1922. The biology of the Chrysopidae. Memoirs of the Cornell
University Agricultural Experiment Station 58: 1287-1372.
TAUBER, C. A. 1974. Systematics of North American chrysopid larvae:
Chrysopa carnea group (Neuroptera). Canadian Ent. 106: 1133-53.
AND M. J. TAUBER. 1977a. Sympatric species based on allelic
changes at three loci: evidence from natural populations in two
habitats. Science 197: 1298-99.
-- AND M. J. TAUBER. 1977b. A genetic model for sympatric speciation
through habitat diversification and seasonal isolation. Nature 268:
702-705.
AND M. J. TAUBER. 1982a. Sympatric speciation in Chrysopa:
further discussion. Ann. Ent. Soc. America 75(1) : 1-2.
-- AND M. J. TAUBER. 1982b. Evolution of seasonal adaptations and
life history traits in Chrysopa: response to diverse selective pressures,
pp. 51-72. In Dingle, H. and J. P. Hegmann, eds., Evolution and
Genetics of Life Histories. Springer-Verlag, N.Y.
TAUBER, M. J., AND C. A. TAUBER. 1981. Seasonal responses and their geo-
graphic variation in Chrysopa downesi: ecophysiological and evolu-
tionary considerations. Canadian J. Zool. 59: 370-76.
THIELCKE, G. 1969. Geographic variation in bird vocalizations. Pages 311-39
In Hinde, R. A., ed., Bird Vocalizations: Their Relation to Current
Problems in Biology and Psychology: Essays Presented to W. H.
Thorpe. Cambridge University Press, Cambridge.
THORNHILL, R. 1980. Competitive, charming males and choosy females:
was Darwin correct? Florida Ent. 63(1): 5-30.
AND J. ALCOCK. 1983. The Evolution of Insect Mating Systems.
Harvard University Press, Cambridge, Mass.
TJEDER, B. 1960. Neuroptera from Newfoundland, Miquelon, and Labrador.
Opuscula Ent. 25: 146-49.
WAAGE, J. K. 1983. Sexual selection, ESS theory and insect behavior: some
examples from damselflies (Odonata). Florida Ent. 66(1): 19-31.

38 Florida Entomologist 68 (1) March, 1985

WALKER, T. J. 1964. Cryptic species among sound-producing ensiferan
Orthoptera (Gryllidae and Tettigoniidae). Quart. Rev. Biol. 39:
345-55.
1974. Character displacement and acoustical insects. American
Zool. 14: 1137-50.
WEST-EBERHARD, M. J. 1983. Sexual selection, social competition, and
speciation. Quart. Rev. Biol. 58(2): 155-83.
WILSON, E. 0. 1975. Sociobiology. Belknap Press, Cambridge, Massachu-
setts.
--- AND W. L. BROWN. 1956. New parasitic ants of the genus Kyidris,
with notes on ecology and behavior. Insectes Sociaux 3(3): 439-54.
WING, S., J. E. LLOYD, AND T. HONGTRAKUL. 1983. Male competition in
Pteroptyx fireflies: wing-cover clamps, female anatomy, and mating
plugs. Florida Ent. 66(1) : 86-91.

Insect Behavioral Ecology-'84 Buskirk and Sherman 39

THE INFLUENCE OF LARVAL ECOLOGY ON
OVIPOSITION AND MATING STRATEGIES
IN DRAGONFLIES

RUTH E. BUSKIRK AND KAREN J. SHERMAN*

SYNOPSIS
Reproductive behavior of adult dragonflies can be related to ecological
requirements of their aquatic larvae. Based on the literature on larval
ecology, we suggest patterns of female oviposition behavior that should place
offspring in favorable situations, appropriate for the species. The predicted
patterns, with respect to microhabitat use and spatial distribution of eggs,
largely agree with our observations of oviposition behavior in seven species
of pond dragonflies. By knowing the behavior of ovipositing females, we can
predict the most feasible strategies for male dragonflies to maximize their
breeding success. Whether males employ territorial defense and whether
they guard their mates following copulation is related to the manner of
oviposition of the female in sixteen species of dragonflies surveyed. In order
to test these predictions further, we call for more detailed observations of
female oviposition behavior and other features of adult mating systems that
are directly related to survival of young offspring.

Dragonflies spend most of their life cycle as aquatic larvae or nymphs
(Corbet 1962, Paulson and Jenner 1971), and pond and lake studies indicate
there is up to 99.9% mortality during these aquatic stages (Benke and
Benke 1975). In dragonflies of the temperate zone, individuals are in the
adult stage only a few weeks, a small fraction of their total lifetime. Fol-
lowing emergence as adults, females produce batches of mature eggs, mate,
select appropriate oviposition sites and deposit their eggs. Eggs may be
dipped or flicked into water or mud, or they may be inserted into vegetation
with the female's piercing ovipositor (Corbet 1962, 1980, Paulson 1969).
Males patrol the oviposition habitats, may defend areas against other males,
and attempt to mate with any solitary female they see.
In recent years odonates have been the subject of several excellent be-
havioral ecology studies in which mating systems were investigated with a
consideration of individual strategies and variability within a species (Alcock
1979, 1982, Campanella 1975, Campanella and Wolf 1974, Fincke 1982,
Pezalla 1979, Sherman 1983b, Ubukata 1975, Ueda 1979, Waage 1979, Wolf
and Waltz 1984). Research has concentrated primarily on the adult males
and, from observations of individually marked animals, how their successful
strategies vary with age, population density, and resource configuration.
For example, at an earlier symposium in this series, Waage (1983) calcu-
lated the costs and benefits of various territoriality strategies in males of
Calopteryx damselflies.

*Ruth Buskirk is a research scientist at the University of Texas at Austin, with research
interests in behavioral ecology and the biological effects of geophysical phenomena. Karen
Sherman is a postdoctoral fellow at the University of Washington, whose dissertation research
focused on dragonfly reproductive strategies. Current addresses: Buskirk, P. O. Box 7456,
University of Texas, Austin, TX 78712; Sherman, Department of Epidemiology SC-36, Uni-
versity of Washington, Seattle WA 98195.

Florida Entomologist 68 (1)

March, 1985

The objective of this paper is to draw attention back to the selection
pressures acting on the dragonfly egg and early nymphal instars. We wish
to focus on features of adult mating systems that are more directly related
to survival of young offspring. In particular, we are concerned with where
and how the female deposits her eggs. Our approach is to consider selective
pressures on the eggs and young nymphs in order to predict the optimum
oviposition behavior of the female and then the related behavior of the male.
Female dragonflies have limited ability to influence the survival of off-
spring after egg deposition since no parental care is provided. However,
there should be strong selection pressure to choose oviposition sites in which
nymph survival is high. Such sites should be characterized by physical con-
ditions that are favorable for development (temperature, oxygen concen-
tration, water movement), that provide sufficient food for nymphs, that
minimize predation and that minimize intra- and interspecific nymphal
competition. Females must use general habitat cues to evaluate oviposition
site quality since they are unable to use more direct assessments such as
predator density or relative food supply. Females do not inspect all areas of
a pond, particularly if it is large, before beginning to oviposit. Thus, they
are not necessarily selecting the best oviposition sites, but initially are
probably choosing adequate sites.

RELATION OF OVIPOSITION PATTERNS TO LARVAL REQUIREMENTS
Several aspects of female oviposition behavior would affect survival of
eggs and young nymphs, including temporal patterns, microhabitat choice,
and spatial dispersal (i.e., clumping) of eggs. In addition, the behavior of
females would be affected by the two primary causes of interrupted oviposi-
tion, i.e., predation on the female and disturbance by conspecific males.
Temporal aspects of oviposition primarily involve selective forces on the
adult female. For instance, the rate of egg release is temperature dependent
(McVey 1984) and thus varies with time of day. Male activity periods affect
the probability that oviposition will be interrupted by males. Male distribu-
tion, therefore, is related to the time of day females can deposit most eggs.
These temporal aspects affect the larvae only indirectly, in that rapid
oviposition rates at one site might lead to overcrowding in the nymphs.
Because we are interested in oviposition patterns that affect success of
eggs and early nymphal instars, this discussion will concentrate on micro-
habitat and spatial aspects, rather than temporal patterns.

MICROHABITAT SELECTION

Females should place eggs into habitats that are most suitable for egg
and early nymph survival. Because late instar nymphs are more mobile and
probably select their own microhabitats (Corbet 1962), female behavior
probably has less direct impact on survival in these stages. Merritt and
Cummins (1978) have classified odonate nymph habits into three broad
categories: burrowers, sprawlers and climbers. Burrowers (primarily mem-
bers of the families Petaluridae, Cordulegastridae and Gomphidae) live in
burrows in the sediments of streams, ponds and lakes. Sprawlers (most
Libellulidae, some Corduliidae) inhabit the surface of sediments, debris and
vegetation while climbers (Aeshnidae, Celithemis, Epicordulia, Leucorrhinia)

Insect Behavioral Ecology-'84 Buskirk and Sherman 41

move vertically on stem surfaces. However, within each category there may
be considerable differences in the activity pattern of different species (Corbet
1962, Kime 1974). Johnson and Crowley (1980) suggested that nymphs can
be functionally classified into two groups: cryptic "sit and wait" nymphs
and actively searching nymphs. More active nymphs should be found in
vegetated habitats which offer greater protection from visual predators
while "sit and wait" nymphs should be found in more open habitats. Thus,
females of species that have active nymphs should prefer vegetated habitats
for oviposition.
The two most common sources of mortality for eggs and nymphs are
predation and desiccation. Oviposition in emergent vegetation in littoral
habitats would offer more protection to eggs and young larvae from visual
fish predators. These heavily vegetated areas in shallow water, however,
are more likely to dry up before the completion of larval development.
Oviposition in open water at the center of a pond would reduce the chances
of desiccation. The balance point between these two selective factors varies
from species to species. For example, in dragonflies with potentially rapid
growth rates, the larvae have a greater probability of maturing before the
habitat dries up. In addition, the warmer water temperatures of shallow,
vegetated areas or small pools would facilitate rapid growth. These groups
would be expected to select small pools or pond margins for oviposition.

SPATIAL DISTRIBUTION OF EGGS

The consequences of spatial patterns of egg deposition in salamanders
were studied by Wilbur (1977). He calculated that when there was a con-
stant probability of egg loss to predators, then the expected number of
surviving offspring would be a function of the total eggs produced, not a
consequence of their distribution among different nests. He determined,
however, that the mathematical variance in the number of surviving off-
spring increased with higher numbers of eggs per nest. Individual females
with a few nests each containing many eggs would have lower reproductive
success when an entire nest was lost to predation than females with more
nests and fewer eggs per nest. There can be, therefore, an advantage to
reducing the number of eggs deposited per single nest,, as long as the cost
of making new nests is not high.
These calculations also apply to dragonflies. They too have a complex
life cycle and a balance between use of temporary and permanent ponds
that is determined by the species' vulnerability to predation and desiccation
(Wilbur 1980). For dragonfly females the cost of "making multiple nests"
would lie in finding or traveling to another pond or in movement to another
site in the same pond. Females in populations with high rates of predation
on eggs should have been selected to divide their egg mass by depositing
smaller batches in each of several suitable locations. Species in which eggs
are less subject to predation, and in which adult females are not heavily
predated during oviposition, are predicted to deposit more of their egg
mass at a single point within a pond.

PREDICTED AND OBSERVED OVIPOSITION PATTERNS FOR POND DRAGONFLIES
From the selection pressures just discussed emerge predictable models

42 Florida Entomologist 68(1) March, 1985

of adult dragonfly behavior. In this section we first present the array of
oviposition patterns "available" to females, defined in terms of microhabitat
and spatial distribution, and predict specific patterns for various larval
requirements. We then test the predictions with observations on several
species of pond dragonflies. No detailed comparative studies on the behavior
of individual females have been reported previously, nor are there many
experimental studies of nymphal ecology. Detailed investigations of nymphal
mortality, its causes, and the location of an individual female's oviposition
bouts would be needed to test these predictions rigorously. Therefore, we
must rely on a few published studies of larval ecology to make predictions,
and we present new observations on female behavior along with some
previously reported work (Table 1).
Some of the oviposition options available to females, depending upon
pond size and density of vegetation, are sketched in Figure 1. The simplest
strategy would be for a mated female to select a suitable patch of habitat
and deposit her entire batch of fertilized eggs as quickly as possible (see A,
Fig. 1). In order to avoid having all her eggs devoured immediately by
fish attracted to the oviposition movements, the female could oviposit briefly
at several sites within the pond (see option B, Figure 1). Placing eggs in
small bodies of water, where fish are less likely to occur, or at heavily
vegetated sites, where nymphs would be less visible, could reduce fish preda-
tion on the larvae (C). When nymphal overcrowding and intraspecific
competition are probable, more offspring would thrive if eggs were spread
out over the suitable habitat (D). To avoid desiccation, the female could
choose larger bodies of water or more open water (E), but if rapid larval
development were possible, growth would be faster in smaller pools or
shallower areas with warmer water (F). Females of species with more
active, climbing larvae are expected to oviposit near or on aquatic vegeta-
tion (G), while species with nymphs that are sprawlers, "sit and wait"
predators, may oviposit in more open water (H).
From observations of oviposition behavior we can test the predictions in
several species. The widespread libellulid Pantala hymenea is a strong-
flying species and is extremely opportunistic in selecting oviposition sites.
In warm conditions larval development for this species is most rapid (Corbet
1962), and the generalist feeding, active sprawlers are apparently not often
food-limited (Heyer et al. 1975). We would predict pattern F (small bodies
of water) and, especially in variable habitats, a reduction of reproductive
variance by distributing a single batch of eggs among many sites. In fact,
mated females of P. hymenea oviposit briefly in small pools of water and
fly in tandem to many oviposition sites. The male releases his mate momen-
tarily for her to dip her abdomen one to fifteen times in small bodies of
water, that range from water-filled footprints to swimming pools.
In Table 1 larval susceptibility to predation is inferred from the con-
sistent breeding failure of Anax junius and Plathemis lydia in study ponds
with fish (Johnson and Crowley 1980). Vulnerability to predators was also
demonstrated by the extreme increase in dominance of Pachydiplax longi-
pennis in fish exclusion experiments by Morin (1984), while lack of preda-
tion pressure on larvae was suggested for Perithemis tenera in the same
experiments. In the three species in Table 1 considered most vulnerable to
fish predation, the females oviposit only in heavily vegetated areas; in fact,

Insect Behavioral Ecology-'84 Buskirk and Sherman

0 3
m
m --
S'.

2a3

aa
<<-! ,

C,

') V

03
t *'-

c0 r
>> >>l

Q)

m

C,

pq

bi C 'a

WW (W
+3 +3
gf f

Q ~( 0 0

F 00

02
-a

0
FI a

bo
;> 0

; 0
-1 J!+
C3

s z

r 0
0 Q

ia ;

Idu ^

C3
NC3
NO

cu
,

10
0
uD
4tf-

OS, .4~

kh.0
COW

S0a

va,

-!.

aoo

c3 0
kFIa

H

of o I

i .3 .3

02

vi

N

3 '3
o
a

0

Eg
r12

bo a)

BJ
bf .-

LO
LO
10

+s

02
-r0

Cca

),~
'*
4T3 +3I
3c r

Se 9
Is
V '0

00

M
02
03

41
0i
0

'3
a0

,o
(ft~
*^ S

'3
>

Scd

020
0 0

0
00

020
a-

0 0J
Zc31
1H|

c3a

e3 ^

c^.1

Florida Entomologist 68 (1)

March, 1985

A SIMPLE

B AVOID EGG PREDATION

C 7 AVOID NYMPH PREDATION

D T7 AVOID NYMPH CROWDING

E AVOID POND DESICCATION

F SPEED NYMPH DEVELOPMENT

G 73 7 CLIMBING NYMPHS

H SPRAWLING NYMPHS

Fig. 1. Schematic diagrams of possible oviposition patterns, with respect
to pond size, density of vegetation, and spatial distribution of eggs for pond
dragonflies. Individual options B through H would be predicted for species
with the different larval requirements indicated.

females of A. junius oviposit endophytically, usually in tandem position
with their mates. Spring emergence species such as Ladona deplanata have
more synchronous larval development than other dragonflies (Benke and
Benke 1975, Benke et al. 1982). As predicted in such species with potentially
more larval competition (pattern D), the female of L. deplanata moves
erratically around the perimeter of the pond as she oviposits. Nymphs of
Libellula species are sprawlers, and the female tends to oviposit in more
open water and more widespread sites, often wandering outside the com-
paratively large territory held by her mate.
As many of the examples in Table 1 indicate, there is agreement between
the observed oviposition behavior of pond dragonflies and the basic larval
requirements. A good deal of variability exists among and even within
females in a species (e.g., Paulson 1969, Sakagami et al. 1974, Sherman
1983a). With more intensive observations of marked individuals in the
future no doubt even more variability will be documented. We may find
that individual behavior patterns are a mixture of distinct stable strategies
(Brockman 1980) or perhaps a set of conditional strategies (Cade 1980).
In general, the existing information on larval ecology suggests a relation-
ship to the oviposition pattern of adult females of the species.

Insect Behavioral Ecology-'84 Buskirk and Sherman 45

RELATIONSHIPS OF MALE BEHAVIOR TO OVIPOSITION PATTERNS
A broad spectrum of male behavior, with considerable variability within
a species has been described for odonates (Alcock 1979, 1982, Corbet 1962,
1980, Jacobs 1955, Moore 1952, Panjunen 1966, Parr 1983, Pezalla 1979,
Ueda 1979, Waage 1983, Wolf and Waltz 1984). Much of the variabilityjn
pond dragonfly behavior lies in two characteristics: the extent to which
males occupy or defend specific sites and the extent to which males defend
their mates following copulation. Just as in many other animals (Emlen and
Oring 1977, Thornhill and Alcock 1983), the probable selective pressures on
odonates are elucidated when the occurrence of females is considered a type
of resource and male behavior viewed as a resource-based strategy (Campa-
nella 1975, Sherman 1983a, b).
If the distribution of reproductive females is predictable in space and
time, males that defend a preferred oviposition site can be expected to have
an advantage over males that do not. Males that have more oviposition
sites in their defended areas tend to obtain more matings. Frequently, males
vigorously defend small areas where females visit the pond at specific times
of the day, for example, Plathemis lydia (Campanella and Wolf 1974),
Perithemis tenera (Jacobs 1955), and Erythrodiplax funerea (Buskirk,
personal observations). The number of matings obtained by territorial males
also depends upon the density of males at the pond (Pajunen 1966, Campa-
nella and Wolf 1974, Parr 1983, Sherman 1983a, Wolf and Waltz 1984). At
very low densities, males flying over larger areas to seek females may be
more successful than localized or territorial males. At very high male
densities, the proportion of time that a male spends in flight increases
(Fried and May 1983), and territorial males could spend so much energy
in fighting that they would obtain no matings.
Detailed observations on several species indicate that the temporal and
spatial distribution of females is related to the pattern of pond use by males
(Campanella and Wolf 1974, Pezalla 1979, Sherman 1983a). Comparisons
of the mating systems of many species show the same relationship. In
Figure 2 sixteen species of pond dragonflies are categorized behaviorally, on
the basis of our observations supplemented with some data from the litera-
ture. Female oviposition (spatial distribution of eggs) for each species is
placed on a scale from "scattered" (few abdomen dips at each site, many
widespread sites) to "clustered" (entire clutch deposited at one site).
Highly clustered oviposition is exemplified by a Perithemis tenera female
that, following copulation, dipped continuously in one 15-cm opening of
water in an algal raft for 165 seconds then left the pond. On the other ex-
treme a Pantala hymenea female dipped her abdomen no more than 12
times, usually 1-4 times, in emergent vegetation and moved a few meters
between each oviposition bout. In Figure 2, circles indicate species with
males that display site-specific territorial defense. Note that all species in
which females have highly clustered oviposition patterns show some terri-
toriality. Species in which oviposition is spatially scattered tend not to show
site-specific defense by males.
The second aspect of male behavior, the extent to which males defend
their mates following copulation, is also related to female oviposition pat-
terns in Figure 2. For several species of odonates (Alcock 1979, 1982,
Sherman 1983, Waage 1979, 1983) it has been shown that when a male de-

Florida Entomologist 68 (1)

March, 1985

46

0)

w
-j

W
<

t
CD

z
Z

C
<
C3
w
I-

OVIPOSITION SITES
Fig. 2. Relationship between female oviposition behavior and male be-
havior for pond dragonflies. Oviposition pattern ranges from scattered (few
abdominal dips at each of many distant sites) to clustered (many abdominal
dips at a single site). Male defense of his mate following copulation can
range from none to hover-guarding to tandem, contact guarding. Species
observed by the authors (localities: CR=Costa Rica, MN=Minnesota,
SC=South Carolina, TX=Texas) are: A-Aeschna canadensis (MN, also
Whitehouse 1941), B-Anax junius (SC, TX), C-Anax longipes (SC),
D-Erythemis simplicicollis (SC, TX), E-Erythrodiplax funerea (CR),
F-Ladona deplanata (SC), G-Lepthemis vesiculosa (CR), H-Libellula
auripennis (SC), I-Libellula croceipennis (TX, also Williams 1976), J-
Libellula luctuosa (TX, also Campanella 1975, Pezalla 1979), K-Pachy-
diplax longipennis (SC, TX), L-Pantala flavescens (TX, also Sakagami
et al. 1974), M-Pantala hymenea (CR, TX), N-Perithemis tenera (TX,
also Jacobs 1955), O-Plathemis lydia (TX, also Jacobs 1955, Campanella
and Wolf 1974), P-Tramea carolina (SC). Circles indicate species in which
males show site defense.
fends the female after mating, she deposits more eggs (presumably fer-
tilized by his sperm) than when there is no defense. This male behavior
may take the form of non-contact guarding in which the male hovers above
the ovipositing female and vigorously chases away other odonates that come
near, or contact guarding when the male joins in tandem flight with the
female and they move together to oviposition sites. Note in Figure 2 that

SCATTERED CLUSTERED

Insect Behavioral Ecology-'84 Buskirk and Sherman 47

most species with clustered oviposition sites for individual females tend to
involve non-contact, hover-guarding by the males. For those species in which
females show more widely dispersed oviposition, the data suggest that two
possible strategies may be followed. In some species (e.g., Pantala, Tramea)
males utilize contact guarding and defend their mates in the tandem posi-
tion. On the other hand, it appears that in some circumstances (Fig. 2: A,
Aeschna canadensis; G, Lepthemis vesiculosa) males abandon mate guarding
entirely, when female oviposition is widely scattered and far from the
copulation site. This strategy appears to characterize a number of stream
dragonflies, such as some gomphids and cordulegasterids, as well.
A limitation of the conclusions from Figure 2 is that the data points
are over-simplified for many species. In a study of the libellulid Sympetrum
parvulum, Ueda (1979) found that territorial males hover-guard over their
mates, but in wandering males the incidence of tandem mate-guarding in-
creased as population density increased. Variability in contact guarding
with male density also appears in Pantala flavescens and Tramea carolina
(authors' observations). For several species in Figure 2, then, population
density and momentary sex ratios at the pond can affect the extent of male
territoriality, mate guarding and female movement. Despite variability
within species, however, the major correlations between male and female
behavior patterns remain clear.
An additional complication is that male behavior has a short-term direct
effect on female oviposition behavior. Females of species that tend to lay
many eggs at one point in the pond when undisturbed or when guarded show
a more widely dispersed pattern if disturbed regularly by conspecific males.
On the other hand, observations on some species suggest that the hovering
male may "herd" the female and induce her to stay longer in his territory
(e.g., Campanella and Wolf 1974). A male, however, can rarely confine a
female to his territory. Behavioral interactions between the sexes contribute
to the variability seen in oviposition patterns.

DISCUSSION
From this brief survey we speculate that the mating behavior of adult
dragonflies can indeed be related to ecological requirements of the aquatic
egg and larval stages. The spatial configuration and microhabitat use of
ovipositing females can be predicted from knowledge of larval ecology
(Figure 3). The extent of territoriality and mate-guarding behavior in
males of a species is related to the oviposition pattern of the females.
Natural selection favors male behavioral strategies that provide maximum
oviposition by their mates in suitable habitats. Territorial defense by males
is only advantageous when females are predictable in time or space.
The major selective factors on adult dragonfly behavior are shown as
solid arrows in Figure 3. Larval ecology for each species determines which
female oviposition patterns) will convey most reproductive success. Micro-
habitat selection by ovipositing females should place eggs in locations that
are appropriate for the habits and behavior of the nymphs. Males seek
mates at the oviposition habitat. Territorial decisions by males at the pond
should be made on the basis of female predictability in space and time and
of male population density. Whether a male employs territorial defence or
guards his mate following copulation depends upon the movement patterns

48 Florida Entomologist 68(1) March, 1985

Demographic
Predation Variables

FEMALE MALE
NYMPH m OVIPOSITION = BEHAVIOR
ECOLOGY BEHAVIOR

Environmental
Variables

Fig. 3. Relationships between variables affecting reproductive success in
pond dragonflies. The major selective factors on adult dragonflies are shown
as solid arrows. Short-term factors affecting reproductive behavior are in-
dicated as open arrows.

of the ovipositing female. There is variability within a species, and often
within an individual, for components of the model in Figure 3, with the
greatest variability in male behavior. Previous studies of marked individual
dragonflies indicate that there are extreme differences in reproductive suc-
cess (number of matings) among males in a population (e.g., Sherman
1983a, Wolf and Waltz 1984).
Short-term factors affecting reproductive behavior, including population
parameters and habitat variables, are depicted as open arrows in Figure 3.
These variables affect mating behavior, but in this scheme they are only
ancillary evolutionary factors. Behavior of the adults is ultimately shaped
by the ecological needs of eggs and young larvae.
In addition to behavioral studies of individual male odonates, therefore,
we call for closer study of the relationships between adult mating systems
and selective pressures on the eggs and young nymphs. A detailed look at
female oviposition behavior can provide the framework necessary to link
these different aspects of reproductive success. Previous observations of
oviposition microhabitat can be combined with new studies of the movements
of marked females.
There is a practical use that can be made of such information. Larval
odonates have been used for monitoring environmental quality (Carle 1979,
Voshell and Simmons 1978), and in some stream species with narrow
habitat requirements, a shift in adult species composition can be seen in
association with relatively small changes in water flow and vegetation struc-
ture. Adult shifts thus can also serve as indicators of water quality. Be-
cause of the great variability in male behavior, there is actually more in-
formation available from observations of female oviposition.
The information on presence of adults, however, must be used with
discretion. For example, in a survey in Australia (Watson et al. 1982),
odonates of all life stages decreased in abundance immediately downstream
from a source of sewage effluent, and though the adult fauna recovered
further downstream, larvae of some species remained absent. Thus, popula-
tion analyses based on adult stages only, particularly adult males, will not
necessarily provide information on long-term effects of environmental change

Insect Behavioral Ecology-'84 Buskirk and Sherman 49

on odonates. Note that locality records of adult male odonates of an en-
dangered species (e.g., Bick 1983) may not provide confirmation that the
species is breeding successfully at that site. Perhaps as we learn more about
larval requirements, as well as the associated oviposition patterns and male
behavior, it may be possible to monitor habitat change using observations
of the adults of certain indicator species.

ACKNOWLEDGEMENTS

We greatly appreciate logistical support for field observations from the
Savannah River Ecological Laboratory (South Carolina), the Brackenridge
Field Laboratory (Texas), and the Organization for Tropical Studies (Costa
Rica). We thank Jim Lloyd for inviting participation in the symposium and
to the social behavior group at Cornell University for nourishment at the
early stages of this research.

LITERATURE CITED
ALCOCK, J. 1979. Multiple mating in Calopteryx maculata (Odonata:
Calopterygidae) and the advantage of noncontact guarding by males.
J. Nat. Hist. 13: 439-46.
1982. Postcopulatory mate guarding by males of the damselfly,
Hetaerina vulnerata Selys (Odonata: Calopterygidae). Anim. Behav.
30: 99-107.
BENKE, A. C., AND S. S. BENKE. 1975. Comparative dynamics and life his-
tories of coexisting dragonfly populations. Ecology 56: 302-17.
---, P. H. CROWLEY, AND D. M. JOHNSON. 1982. Interactions among
coexisting larval Odonata: an in situ experiment using small en-
closures. Hydrobiologica 94: 121-30.
BICK, G. H. 1983. Odonata at risk in conterminous United States and
Canada. Odonatologica 12: 109-26.
BROCKMAN, H. J. 1980. Diversity in the nesting behavior of mud-daubers
(Trypoxylon politum Say; Sphecidae). Florida Ent. 63: 53-64.
CADE, W. 1980. Alternative male reproductive behaviors. Florida Ent. 63:
30-45.
CAMPANELLA, P. J. 1975. The evolution of mating systems in temperate
zone dragonflies (Odonata: Anisoptera) II. Libellula luctuosa
(Burmeister). Behaviour 54: 278-310.
-- AND L. L. WOLF. 1974. Temporal leks as a mating system in a
temperate zone dragonfly (Odonata: Anisoptera). I: Plathemis lydia
(Drury). Behaviour 51: 49-87.
CARLE, F. L. 1979. Environmental monitoring potential of the Odonata,
with a list of rare and endangered Anisoptera of Virginia, United
States. Odonatologica 8: 319-23.
CORBET, P. S. 1962. A biology of Dragonflies. Witherby, Ltd., London.
- 1980. Biology of Odonata. Annu. Rev. Ent. 25: 189-217
EMLEN, S. T., AND L. W. ORING. 1977. Ecology, sexual selection and the
evolution of mating systems. Science 197: 215-23.
FINCKE, O. M. 1982. Lifetime mating success in a natural population of
the damselfly, Enallagma hageni (Walsh) (Odonata: Coenagrion-
idae). Behav. Ecol. Sociobiol. 10: 293-302.
FOLSOM, T. C., AND N. C. COLLINS. 1982. Food availability in nature for
the larval dragonfly Anax junius (Odonata: Aeschnidae). Freshwater
Invert. Biol. 1: 33-40.
FRIED, C. S., AND M. L. MAY. 1983. Energy expenditure and food intake of

Florida Entomologist 68 (1)

March, 1985

territorial male Pachydiplax longipennis. (Odonata: Libellulidae).
Ecol. Ent. 8: 283-92.
HEYER, W. R., R. W. MCDIARMID, AND D. L. WEIGMANN. 1975. Tadpoles,
predation, and pond habitats in the tropics. Biotropica 7: 100-11.
JACOBS, M. E. 1955. Studies on territorialism and sexual selection in
dragonflies. Ecology 36: 566-86.
JOHNSON, D. M., AND P. H. CROWLEY. 1980. Odonate "hide and seek":
Habitat-specific rules? Pages 569-79 In W. C. Kerfoot ed. Evolution
and ecology of zooplankton communities. Univ. Press of New England.
Hanover, New Hampshire.
KIME, J. 1974. Ecological relationships among three species of aeshnid
dragonfly larvae (Odonata, Aesnidae). Ph.D. thesis, University of
Washington, Seattle.
McVEY, M. E. 1984. Egg release rates with temperature and body size in
Libellulid dragonflies (Anisoptera). Odonatologica 13: 377-85.
MERRITT, R. W., AND K. W. CUMMINS. 1978. An introduction to the aquatic
insects of North America. Kendall Hunt, Dubuque, Iowa.
MOORE, N. W. 1952. On the so-called "territories" of dragonflies (Odonata-
Anisoptera). Behaviour 4: 101-3.
MORIN, P. J. 1984. The impact of fish exclusion on the abundance and
species composition of larval odonates: results of short-term experi-
ments in a North Carolina farm pond. Ecology 65: 53-60.
NEEDHAM, J. G., AND M. J. WESTFALL. 1955. A manual of the dragonflies of
North America (Anisoptera). Univ. Calif. Press, Berkeley.
PAJUNEN, V. I. 1966. The influence of population density on the territorial
behaviour of Leucorrhinia rubicunda. Ann. Zool. Fenn. 3: 40-52.
PARR, M. J. 1983. An analysis of territoriality in Libellulid dragonflies
(Anisoptera: Libellulidae. Odonatologica 12: 39-57.
PAULSON, D. R. 1969. Oviposition in the tropical dragonfly genus Mi-
crathyria (Odonata, Libellulidae). Tombo 12: 12-16.
AND C. E. JENNER. 1971. Population structure in overwintering
larval Odonata in North Carolina in relation to adult flight season.
Ecology 52: 96-107.
PEZALLA, V. M. 1979. Behavioral ecology of the dragonfly Libellula pulchella
Drury (Odonata: Anisoptera). American Midi. Nat. 102: 1-22.
SAKAGAMI, S. F., H. UBUKATA, M. IGA, AND M. J. TODA. 1974. Observations
on the behavior of some Odonata in the Bonin Islands, with con-
siderations on the evolution of reproductive behavior in the Libel-
lulidae. J. Fac. Sci. Hokkaido Univ. 19: 722-57.
SHERMAN, K. J. 1983a. The evolution of reproductive strategies in a dragon-
fly, Pachydiplax longipennis. Ph.D. Dissertation. Cornell University.
-- 1983b. The adaptive significance of post copulatory mate guarding
in a dragonfly, Pachydiplax longipennis. Anim. Behav. 31: 1107-15.
THORNHILL, R., AND J. ALCOCK. 1983. The evolution of insect mating sys-
tems. Harvard Univ. Press, Cambridge, Mass.
UBUKATA, H. 1975. Life history and behavior of a corduliid dragonfly,
Cordulia aenia amurensis Selys. II. Reproductive period with special
reference to territoriality. J. Fac. Sci. Hokkaido Univ. Ser. 6, Zool.
19: 812-33.
UEDA, T. 1979. Plasticity of the reproductive behavior in a dragonfly,
Sympetrum parvulum Barteneff, with reference to the social relation-
ship of males and the density of territories. Res. Popul. Ecol. 21:
135-52.
VOSHELL, J. R., JR., AND G. M. SIMMONS, JR. 1978. The Odonata of a new
reservoir in the southeastern United States. Odonatologica 7: 67-76.
WAAGE, J. K. 1979. Adaptive significance of post copulatory guarding of

Insect Behavioral Ecology-'84 Buskirk and Sherman 51

mates and nonmates by male Calopteryx maculata (Odonata). Behav.
Ecol. Sociobiol. 6: 147-54.
-- 1983. Sexual selection, ESS theory and insect behavior: some ex-
amples from damselflies (Odonata). Florida Ent. 66: 19-31.
WATSON, J. A. L., A. H. ARTHINGTON, AND D. L. CONRICK. 1982. Effect of
sewage effluent on dragonflies (Odonata) of Bulemba Creek, Brisbane,
Australia. Australian J. Mar. Freshwater Res. 33: 517-28.
WHITEHOUSE, F. C. 1941. British Columbian dragonflies (Odonata) with
notes on distribution and habits. American Midl. Nat. 26: 488-557.
WILBUR, H. M. 1977. Propagule size, number, and dispersion pattern in
Ambystoma and Asclepias. American Nat. 111: 43-68.
1980. Complex life cycles. Annu. Rev. Ecol. Syst. 11: 67-93.
WILLIAMS, C. E. 1976. Courtship display in Belonia croceipennis (Selys),
with notes on copulation and oviposition (Anisoptera: Libellulidae).
Odonatologica 5: 283-7.
WOLF, L. L., AND E. C. WALTZ. 1984. Dominions and site-fixed aggressive
behavior in breeding male Leucorrhinia intact (Odonata: Libel-
lulidae). Behav. Ecol. Sociobiol. 14: 107-15.

52 Florida Entomologist 68(1) March, 1985

APPROACHES TO INSECT BEHAVIOR OF INTEREST
TO BOTH NEUROBIOLOGISTS AND
BEHAVIORAL ECOLOGISTS

FRANZ HUBER*

I. SYNOPSIS

The manifold habitat-dependent behavioral tactics insects have evolved
are the common subject of research by both behavioral ecologists and be-
haviorally oriented neurobiologists. Behavioral ecologists prefer to study
undisturbed individuals and populations in the field. They describe behaviors
of different species and try to elucidate how they contribute to reproductive
success. Neurobiologists concentrate on substrates and mechanisms that
underly distinct and quite often simple behavioral acts, and most commonly,
they do this in restrained animals. This paper is an effort by an insect
neurobiologist toward a mutualistic coexistence between behavioral ecology
and neurobiology by the discussion of topics of insect behavior that can
enhance future cooperation, especially in the field of communication systems.

II. INTRODUCTION

Animal behavior can be investigated at various levels within its organi-
zational hierarchy, according to the problems of interest and the questions
being asked. However, quite often new insights and new ideas are achieved
when scientists working at these different levels begin to talk to each other
and make an attempt to exchange concepts, methods and results. This
should become particularly important for behavioral ecologists and neuro-
biologists, especially when the latter concentrate on the neural bases of
adaptive behaviors and habitat-dependent tactics that the former study in
the field. But there are considerable differences in the problems being at-
tacked and in the research strategies used to solve them.
As schematized in figure 1, both intra- and interspecific behavioral inter-
actions provide a common source for research and could start a fruitful co-
operation between behavioral ecologists and neurobiologists. Behavioral
ecologists focus, for example, on how different species select their mates,
compete in mate selection, hunt for prey, avoid or fool predators, forage, or
take care of offspring, and how they organize social groups. They like to
know how these behaviors contribute to survival and reproductive success,
and how they have evolved. Their discoveries raise questions for themselves
and also for neurobiologists, such as, what sensory modalities are involved
in these tactics, and how do sensory, central nervous and effector systems
function in order to produce the demonstrated behaviors.
Neurobiologists, by mainly working under controlled laboratory condi-
tions and quite often with behaviorally restrained animals, search for com-

*Franz Huber is Scientific Member of the Max Planck Society and Director within the
Institute for Behavioral physiology in Seewiesen. In 1953, he received his Ph.D. under the
late Werner Jacobs and the late Karl von Frisch at the University of Munich; from 1954-1963,
he worked as Assistant and Associate Professor of Animal Physiology at the University of
Tilbingen; from 1963-1973 he was Full Professor and Head of the Department for Animal
Physiology at the University of K61n. His research concentrates on the neural bases of insect
behavior. Mailing address: Max-Planck Institut fir Verhaltensphysiologie, Abtl. Huber D 8131
Seewiesen, Federal Republic of Germany.

Insect Behavioral Ecology-'84 Huber

MUTUAL INTERACTIONS BETWEEN BEHAVIORAL ECOLOGY
AND NEUROBIOLOGY

raises questions for

Fig. 1. Common sources of research interests in behavioral ecology and
neurobiology, and mutual interactions.

ponents involved in sensorimotor performances that are expressed in dif-
ferent behaviors. It is their ultimate goal to study how visual, acoustical or
chemical signals are produced and how these signals are perceived and en-
coded in sensory and nervous elements to provide a causal basis for intra-
and interspecific interactions. Their discoveries at the molecular, cellular
and multicellular levels may in turn provide feedback to behavioral ecolo-
gists and have impact on their research.
It is the purpose of this article to start a dialogue between the two
groups by introducing problems of insect behavior attacked by insect
neurobiologists that may guide future research in behavioral ecology, and
even to enhance cooperation.

III. INSECTS AS MODEL SYSTEMS
Insects offer great opportunities for behavioral ecologists (see, Blum and
Blum 1979; Thornhill and Alcock 1983; Gwynne and Morris 1983) to study
the strategies and tactics of sexual selection, reproductive competition and
the evolution of mating systems. For neurobiologists (see Huber 1983 a,b;
Camhi 1984) insects offer a rich and complex behavioral repertoire and a
nervous system that can be attacked at the single cell level, allowing the
so-called "identified single cell approach" (Hoyle 1983). The topics and ex-
amples chosen in this article may help initiate some cross-talk between
neurobiologists and behavioral ecologists and assist scientists working in
these areas to better coordinate their research activities.

54 Florida Entomologist 68(1) March, 1985

A. NEUROBIOLOGICAL LIMITATIONS IN INSECT BEHAVIOR
a. Escape and Hunting
It is most critical for a prey animal to detect the predator's attack as
early and to respond as fast as possible. On the other hand, an efficient
hunter must develop tactics to avoid early recognition and escape by the
prey.
Toads are among natural predators of cockroaches (Camhi 1984). Dur-
ing the lunge of their tongue air is moved which stimulates filiform hair
sensillae located on the cockroach's anal cerci. These mechanosensitive struc-
tures are activated by air particle displacement and their spatial arrange-
ment provides directional information to the roach (Nicklaus 1965; Westin
et al. 1977; Camhi 1984). Roeder (1959) made precise behavioral measure-
ments in the cockroach of the time course and velocity of the escape response
to air puffs. He compared these measurements with data on those elements
in the system that could be monitored electrophysiologically, i.e. cereal
afferent fibers which originate in the mechanosensitive cereal sensillae, ab-
dominal giant interneurons and the leg motor system. Roeder found that
the quickest behavioral response expressed by movement of the body of a
slightly restrained roach occurred around 20 to 30 ms after the stimulus
onset, whereas the time estimated on the basis of electrophysiological data
was around 16 to 20 ms. The two times agreed within an order of magni-
tude strongly indicating that the cockroach's fastest escape is limited by the
time courses of the underlying sensory, interneuronal and motor systems.
More recently Tautz (1977, 1978) and Tautz and Markl (1978) studied
behavior of the caterpillar of Barathra brassicae (Noctuidae). This larva
responds with cessation of movement to airborne vibrations caused by the
wingbeat of an approaching predatory wasp. The response is lost after re-
moval of all eight mechanosensitive filiform hair sensilla situated on the
caterpillar's thorax. The sensors are activated by air-particle displacement
and are especially sensitive to wingbeat frequencies in the range of 100 to
400 Hz at unspecified amplitudes. Within this range they respond at dis-
tances of about 70 cm. Behavioral and electrophysiological data led the
authors to predict that an efficient hunter can avoid early recognition by
these prey larvae in two different ways: it (1) can shift its wingbeat fre-
quency out of the sensitive range of the filiform hairs, as documented in
bigger wasps parasiting caterpillars (Rathmayer 1978), where the wing-
beat frequency lies below 100 Hz; or it (2) can reduce the force of the air
vibrations it produces to below the threshold of the caterpillar's sensors. In
some small wasps parasitic on the same larvae (Habrobracon sp., Bracon-
idae), it was calculated from known mechanical and electrophysiological
properties of the sensory hairs that the caterpillar will not respond until
the wasp approaches closer than 4.5 cm.

b. Signalling and signal detection in the time domain
Insects display a great variety of signals for sexual communication,
mate competition, territorial display and for interactions within social
groups. Depending upon the species, signals may involve chemical substances
(pheromones), reflected light, bioluminescent flashes, airborne sounds, and
even low amplitude vibrations which are common alarm signals in ant

Insect Behavioral Ecology-'84 Huber 55

colonies (Markl 1967, 1968, 1970; Markl and Fuchs 1972). Here I shall con-
centrate on acoustic communication in which airborne sounds are used. Two
signal parameters are most important; the sound frequency spectrum and
the temporal organization (rhythm) of sounds. The latter is also a fre-
quency and is important for signal detection in the time domain.
In timballing cicadas and in some stridulating katydids, trains of short
and highly damped sound pulses are produced at high repetition rates. We
would like to know how the sender generates such fast sound pulse rates and
also whether the receiver is able to transduce such fast amplitude modula-
tions into groups of nerve impulses which encode the rate.
Most cicadas timbal at rates of appr. 200 Hz (=Hertz=cps) (Pringle
1954; Hagiwara and Watanabe 1956). This rate is generated by an alterna-
tion of the paired timbal systems each of which signals with appr. 100 Hz
(Moore and Sawyer 1966; Reid 1971; Young 1972; Simmons 1977; Simmons
and Young 1978). More recently it has been shown by high speed movie
pictures (Moore and Kausch 1975) and by combined recordings of sounds
and timbal muscle potentials (Moore and Sawyer 1966; Young and Joseph-
son 1983 a, b; Moore et al. in prep.) that within each timbal-motion a train
of sound pulses is broadcast. The pulse rate within each train goes far
beyond 200 Hz. In all timballing cicadas the timbal is not a uniform plate,
but is subdivided into parts with rather soft and flexible membranes inter-
spersed with a sclerotized plate and a series of 2-12 sclerotized ribs. When
the timbal muscle contracts it pulls the timbal plate inward. This motion
progresses in discrete steps then from rib to rib. Each rib produces one of
the sound pulses of a train (figure 2). In the periodical cicada Magicicada
cassini up to 9 of its dozen or so ribs operate in succession during a single
inward click of the timbal, and this results in a sound pulse rate of 900 Hz
and a sound frequency spectrum of 4-9 KHz at nearly 100 dB (decibel) at
the source. Thus, the sender is able to generate fast pulse rates because of
the biomechanical properties of its timbal. The number of pulses produced
by a timbal may vary among sounds produced by a single male, but the
pulse rate of the timbal ribs varies very little; dramatic rhythm rates at
much lower frequencies are produced by turning the whole paired mecha-
nism on and off. Frequency and amplitude modulation of carrier frequencies
are also quite common among these complex sounds (Moore, pers. comm.).
The receiver, a male or female conspecific, may be able to resolve the
high timbal rib pulse rates, but at present we have not performed the
relevant quantitative behavioral tests in the field and laboratory which
could unequivocally demonstrate such high pulse rate detection. However,
indirect evidence that these cicadas may not encode pulse rates beyond
200 Hz (at signal durations of less than about 10 ms) comes from whole
auditory nerve and single auditory interneuron recordings (Huber et al.
1980). The response of many auditory receptors to playbacks of conspecific
songs strongly indicates that the receptors only respond to the overall
timbal movement and associated rhythm of ca. 200 Hz (figure 3), and do
not clearly resolve the succession of pulses produced by the ribs. Auditory
interneurons of Magicicada septendecim recorded within the fused thoracic
central nervous system encode only the buzzy song as a whole but not the
detailed temporal structure. In the species Magicicada cassini such inter-
neurons copy the slower tick rate but not the fast buzz rate in the calling

56 Florida Entomologist 68(1) March, 1985

1V
i

2 ms
Fig. 2. Sound production in cicadas. Upper trace: Sequence of brief and
highly damped sound pulses produced by inward buckling of the timbal in
a Magicicada cassini male. The 4 pulses seen correspond to the inward
buckling of each of 4 ribs; the subsequent and faint pulse (arrow) is gen-
erated when the timbal springs back into its resting (outward) position
(Weber, T., unpublished data). Lower trace: Two consecutive muscle po-
tentials recorded from the ipsilateral timbal muscle while the male was pro-
ducing sound as shown in the upper trace, another group of sound pulses
would follow the second potential. Each single muscle potential (shown)
precedes muscle contraction (not shown). Each muscle contraction is fol-
lowed by a fast inward movement of the corresponding timbal with the
associated sounds (upper trace). Time scale horizontal, voltage scale for
sound pulses, vertical (Huber, unpublished data).

song. If it turns out in future work that even single auditory receptors can
not follow such fast rates, then a part of the temporal information laid
down by the sender is lost at the receiver's ear, and the explanation for the
occurrence of ribs may come from another context.
A quite similar problem has been reported in the very loud buzz of the
katydid Neoconocephalus robustus (Josephson and Halverson 1971). Here
the male stridulates by rubbing one forewing against the other. The wing
stroke frequency and resulting pulse rate reaches to 200 Hz or even beyond,
producing a very high and broad sound frequency spectrum. Also in these
katydids the generation of songs with high pulse rates depends on special
biomechanical and muscular properties (Josephson 1973, 1984).
The question arises: Does the receiver encode such pulse rates? Morris
and Fullard (1983) investigated the discrimination power in some ticking
and buzzing Conocephalus species in the time domain using a phonotactic
assay. They found that songs or random noises that lacked a distinct ampli-
tude modulation mimicking the pulse rate still remained highly attractive
to females. This could mean that the receiver either "pays no attention" to
the pulse rate of the song in conspecific recognition or its sensory system
can not resolve fast rates. A comparison with electrophysiological results
obtained in the genus Gryllus will perhaps be helpful for understanding
possible limitations of the auditory system in the time domain.

Insect Behavioral Ecology-'84 Huber 57

75 dB

1 1 r FI" F

20 ms
Fig. 3. Time resolution in the auditory nerve of cicadas (Okanagana
rimosa). Upper trace: Series of bursts of calling sound pulses. Lower trace:
Summed activity recorded from the whole auditory nerve. Note that each
sound burst is correlated with a single but summed auditory nerve signal.
Each sound burst reflects 2 pulses from one timbal followed by 2 or 3 pulses
from the other timbal by a single inward movement of each of the two
timbals. Playback of these bursts elicited a response of similar duration in
the auditory nerve after a latency of approximately 20 msec, as indicated
by the arrows. Sound intensity 75 dB SPL, time scale between black bars
20 msec (modified from Huber et al. 1980).

Auditory receptors in Gryllus bimaculatus and campestris copy sound
pulses of the chirped calling songs (narrow, relatively low sound frequency
spectrum) with a burst of nerve impulses, and impulse frequency within
each burst encodes sound intensity (Esch et al. 1980). At 90 dB the impulse
frequency encoded reaches about 300 Hz. Assuming that similar intensity
encoding is present in Conocephalus, then a sound pulse rate of 200 Hz, with
an interpulse interval of 5 ms between consecutive pulse onsets, would result
in a discharge frequency of auditory receptors of 200 Hz-assuming that
each receptor responds with a single impulse to each single sound pulse.
However, if the receptor also encodes the intensity of the sound in a manner
like that of Gryllus, then consecutive bursts of nerve impulses would overlap
and partly interfere. A consequence would be to reduce high fidelity tem-
poral resolution of the pulse rate by the receptors. So far no auditory system
studied lacks intensity encoding. Thus, Conocephalus is perhaps forced to
make a compromise between pulse rate copying and intensity encoding.
Morris and Fullard's results suggest a change from using the high pulse
rate as an important temporal parameter.

B. NEUROMUSCULAR ECONOMY AND COURTSHIP BEHAVIOR
As often reported (Hoyle 1964, 1970), the insect nervous system has
rather few motor neurons and many muscles are innervated by not more
than 3. In addition, several muscles serve multiple motor actions such as
walking, flying or stridulating; they are called bi- or even multi-functional
(Wilson 1962; Elsner 1968, 1975). The courtship behavior of the male grass-
hopper Stenobothrus rubicundus provides an excellent illustration of such
"multifunction" (Elsner 1974). Two kinds of stridulation are displayed
during courtship: conventional hind leg stridulation, characterized by al-

Florida Entomologist 68 (1)

March, 1985

ternating up- and down-strokes at a frequency of 12 Hz (producing short
bursts with a broad sound frequency spectrum); wing stridulation with the
unfolded fore- and hindwings raised and lowered periodically at a frequency
of 70 Hz (producing even shorter bursts with a broad sound frequency
spectrum)-this rate of movement is also characteristic of flight. During
wing stridulation the sclerotized veins of the hindwings are beaten against
each other to produce sound (figure 4). Between leg stridulation at the start
of the courtship and wing stridulation at the end, the male interposes a mid-
dle part where approximately every third downstroke of the hindleg is
prolonged and subdivided into 3-10 short-lasting subpulses which follow the
frequency of wing stridulation, though the wings are still folded.
Recordings of sounds and muscular activity showed that during leg
stridulation other motor units within the same muscles became active than
were active during the final wing stridulation. In addition, their phase
relationships changed. Thus, the male has produced two different motor
outputs. However, during the prolonged downstrokes of the middle part,

A

25 ; 12 8. B5

129.4 5mv

pill -Down

0.1 sec

0.5 sec
Fig. 4. Neural economy during courtship in a grasshopper. A. Sound
patterns: Groups of double pulses during a single up- and downstroke with
the hindleg (1); middle part of the courtship display with series of up- and
downstrokes where two downstrokes are prolonged and split up into short
sub-pulses at the rate of wing stridulation (2, 3); sound produced during
wing stridulation (4). B. Combined recordings of muscular activities (up-
per three traces) and of sound produced during single upstroke and a con-
secutive but prolonged downstroke (lower trace). The metathoracic muscles
involved are indicated by numbers: 133 tergotrochanter muscle (elevator
of hindleg and hindwing); 125 1st pleurocoxal muscle promoterr and ab-
ductor of the coxa (leg)); 128 2nd basalar muscle promoterr of the coxa
and depressor of the hindwing); 129 subalar muscle (remotor of the coxa
and depressor of the hindwing). Potential calibration at the right. 133
and 125 are both drownstroke muscles which operate as synergists; during
the prolonged downstroke of the leg muscle 128 joins the group. Arrows
point different motor units used by 125 and 128. In wing stridulation (not
shown) 128 and 129 operate synchronously as wing depressors (modified
after Elsner 1974).

Insect Behavioral Ecology-'84 Huber 59

motor units that served leg stridulation were superimposed with those used
for wing stridulation. This indicates firstly, that the male can not suppress
the second motor pattern before the first has ended, and secondly that a
transition occurs at the neuromuscular level before the actual wing move-
ments are performed. Each set of motor units is driven by its own generator
network, e.g. by a limited number of neurons which may be multifunctional
as well. Thus, the middle part of courtship expresses two output-specific but
overlapping rhythmical processes.
Acridid grasshoppers are well known to be most responsive to the species-
specific temporal structure of the song (Helversen 1972; Helversen and
Helversen 1975, 1983). In Stenobothrus rubicundus it can not be excluded
that the different temporal arrangement of sound pulses in the middle part
of courtship is used as an important time element for male recognition, or
that it is just a byproduct of sender mechanisms arising from the simplicity
of neuromuscular systems.

C. CONSPECIFIC SONG RECOGNITION IN CRICKETS:
BEHAVIORAL AND NEUROBIOLOGICAL STUDIES
One fundamental problem in animal communication is recognition, ex-
pressed for instance in species or even individual recognition and also in
tactics where one species mimics another as documented in fireflies (Lloyd
1965,1975, 1980, 1981,1983).
Recognition is thought to be based upon sensory and neuronal detector
systems, often called 'templates'. These are viewed as nervous machineries
that match certain behaviorally relevant stimulus configurations, named
'sign stimuli' by the students of animal behavior. Sign stimulus detection
requires a filter process within the peripheral and or central nervous system.
Recognition of conspecifics often involves more than a single sensory
modality. Male butterflies are aroused by female pheromones and then find
the stimulus source with the help of sensory input generated by wind and
by visual cues (Preiss and Kramer 1983). Male acridid grasshoppers use
sound to attract females from a distance, but courtship is elicited by a set
of visual stimuli (Riede et al. 1979). In many crickets and especially
katydids both airborne sounds and substrate vibrations direct the female
to the singing male at closer distances (Latimer and Schatral 1983; Keuper
and Kiihne 1983). Corresponding to these behavioral observations, neurons
identified within the auditory pathway respond both to airborne sound and
substrate vibrational signals. The substrate signal increases the accuracy of
the response to the temporal structure of the sound (Kalmring et al. 1983).
a. Behavioral approach to recognition
From field observations with marked female crickets we know that they
approach calling males of the same species even in populations where more
than one species is present (Popov et al. 1974). Other stimulus modalities
such as visual targets or chemical substances may contribute (Weber et al.
in prep.; Stout et al. in prep.), but the acoustic channel is sufficient.
By studying female cricket phonotaxis on a walking compensator (a
spherical treadmill designed to record the walking mode and walking direc-
tion in a closed loop system) we learned about the demands for conspecific
song recognition and sound source localization (Wendler et al. 1980, Weber

60 Florida Entomologist 68(1) :` March, 1985

et al. 1981, Thorson et al. 1982, Schmitz et al. 1982, 1983, Pollack et al.
1984). Figure 5 shows a drawing of the experimental setup used and some
of the results. With the female positioned on the treadmill in the dark or
under homogenous light conditions in a soundproof room (neither of which
reflects the natural habitat), we discovered a rather simplistic template for
calling song recognition in females of Gryllus campestris (Thorson et al.
1982). Females were aroused and began phonotactic tracking on the tread-
mill if the model calling song had the natural sound frequency spectrum of
4 to 5 kHz (even without the higher harmonics present in the natural song)
and was organized in syllables (pulses) spaced between 25 to 55 ms inter-
vals, reflecting syllable or pulse repetition rates between 18 to 40 Hz. No
other intrachirp and interchirp parameters were critical, except that each
syllable (pulse) had a critical duration of more than 200 usec (compared
with 15 to 20 ms in the natural song) and the silent intrasyllable (intra-
pulse) pause had a critical duration of at least 4 ms (natural values ca. 10
to 15 ms). Even playbacks of natural calling songs in reverse did not pre-
vent female phonotaxis, that is, the receiver did not discriminate the differ-
ent amplitude modulation (chirp envelope) when played forward or back-
ward. When given a choice between similar model songs, females always
chose the one that was a few decibels louder. At equal intensity, calibrated
close to the female's body on the treadmill, they meandered between the two
loudspeaker positions (1350 separation), indicating that they were unable
to discriminate. Tone bursts of similar frequencies but without syllables
(pulses) of natural chirp length and chirp repetition rate did not start
females of Gryllus campestris walking, and they were not tracked at all.
When the females were given a choice between them and 'pulsed chirps'
they tracked the latter, even if the sound intensity in the former was in-
creased. However, recently different results were obtained with females of
Gryllus bimaculatus (Doherty in prep.) Here some clear burst trackers were
found..
A rather surprising result was obtained with these chirping crickets:
Some females tracked model songs built of continuous trills at their natural
pulse frequency, duration and repetition rates. It appears as though the
receiver may not always require the chirp organization of the song-at least
not under treadmill conditions (but see Stout et al. 1983').
This raises the question: Is the 'simplicity' of the template in the receiver
related to the particular life style of Gryllus campestris? This mid-European
species occupies habitats of hilly grassland where individuals are often
rather widely spaced, especially toward the end of their reproductive season.
No other species of Gryllus lives in this habitat in this area. Males call in
front of their burrows between late April and early July at a time when no
other insect is singing. Birds sing during this time, and they can be heard
by crickets, but their most abundant predators are spiders, carabid beetles,
lizards and shrews. It could be that the absence of sympatric species has
facilitated the evolution of a rather simplistic template.
The species most closely related to Gryllus campestris is the southern-
European Gryllus bimaculatus, which does not overlap with G. campestris.
Their relationship is documented by the fact that they can be hybridized in
the laboratory (v. HSrmann-Heck 1957). Individuals of Gryllus bimaculatus
live alone or in smaller groups and adults fly. Gryllus campestris adults are

Insect Behavioral Ecology-'84 Huber 61

flightless. Males of the two species produce nearly identical calling songs
and females of Gryllus campestris on the treadmill don't discriminate be-
tween them (Thorson et al. 1982). With simultaneous playback of the two
natural songs from two loudspeakers (1350 apart), nearly equalized in
intensity, the females meander between the speaker positions as though
faced with two conspecific songs (figure 5).
A B

++++ 2

b
L1 m

Sf
X b c b b 4

1 min
Fig. 5. Phonotaxis of female of Gryllus campestris (1, 2) and tracking
of conspecific and heterospecific (Gryllus bimaculatus) songs (3, 4).
A. Schematic view of the sound proof room with the spherical treadmill and
a female cricket positioned on top carrying a reflecting foil on the pronotum
(not visible). IR infrared sensing and detecting devices; x, y positions of
motors which drive the sphere; L1, L2 positions of the two loudspeakers
(135 degrees apart). B. Recordings of the females' phonotactic behavior
(tracking) when stimulated with model calling chirps (1, 2) or playbacks
of natural calling sounds (3, 4). (1) G. campestris response to pulsed
chirps (4 pulses per chirp at 5 kHz) and 5 kHz tone bursts of same total
duration as the pulsed chirps in a sequential paradigm. The female tracks
only pulsed chirps not tone bursts. Relative loudspeaker positions (L1, L2)
are indicated by straight lines superimposed over the corrective meandering
during playback of sounds, and their angular position is indicated left
(range 00 to 3600); symbols above and below show signal from each speaker
(L1 above, L2 below), line means no signal vs four pulsed chirps, or solid
tone bursts. (2) G. campestris response with model pulsed chirps from the
left then right sides followed by pulsed chirps and tone bursts presented in
a choice paradigm, and followed by pulsed chirps from the left. The female
always tracks pulsed chirps. (3) G. campestris response with natural con-
specific calling song (indicated by c on the side from which it is played)
and the similar calling song of G. bimaculatus (indicated by b, similarly) in
a sequential paradigm. The female tracks both songs equally accurately, and
without sound she circles (right). (4) G. campestris with the same natural
sounds as in (3). First the conspecific song (c) is played on the right, then
the left, then both songs simultaneously three times but changing sides each
time, then the heterospecific song (b) on the left. When given a choice be-
tween conspecific and heterospecific calling song at nearly equal sound in-
tensities the female is unable to discriminate between the two and meanders
between the two loudspeaker positions. (A, B composed from Weber et al.
1981, and from Thorson et al. 1982).

62 Florida Entomologist 68 (1) March, 1985

Popov (Popov et al. 1974, 1975) reported that the two species overlap
in some part of southern Russia. There the calling songs are heard at the
same time of the day, but hybrids were not recorded in the field. This sug-
gests that in sympatric populations of these species the templates become
more specific. Popov tested individuals from areas of sympatry and found
that females of Gryllus bimaculatus were less discriminatory with respect
to the number of syllables (pulses) per chirp than those of Gryllus cam-
pestris. Our results obtained with mid-European Gryllus campestris (Weber
et al. 1981, Thorson et al. 1982) contradict Popov's findings with sympatric
Gryllus campestris but confirm his results with Gryllus bimaculatus.
Further field and laboratory studies are needed to establish the concept that
sympatry may sharpen the templates.
Crickets are poikilothermic and their body temperature follows ambient
temperature. As first shown by Walker (1957) ambient temperature effects
the syllable (pulse) repetition rate in trilling and chirping tree crickets
(mostly Oceanthus spp.): it increases with increasing temperature. Simi-
larly, in chirping crickets of the genus Gryllus both syllable (pulse) and
chirp repetition rate change with temperature (Kriechbaum 1983, Doherty
and Huber 1983, Doherty in press). Therefore, we became interested in
temperature effects on female phonotaxis in order to get some insight into
temperature coupling between sender and receiver. Females of Gryllus
bimaculatus were acclimated to either 150C, 22C or 300C in an incubator,
and then tested on the treadmill for phonotaxis in the equally-acclimated
sound-proof room. A computer-controlled program of model calling songs
with different syllable (pulse) and chirp repetition rates and durations was
delivered. As shown in figure 6, 'cold' females (150C) shifted their prefer-
ence to lower syllable (pulse) repetition rates, and 'hot' females (300C)
preferred higher syllable (pulse) repetition rates. These results may indicate
that sender and receiver are temperature-coupled in Gryllus as well as in
Oceanthus, that is, that mechanisms for song generation in the sender and
those for recognition in the receiver are influenced by temperature in a
similar way. How such effects are achieved within the nervous system re-
mains to be discovered, and we need much more information on the nature
of songs produced and responses elicited at different natural temperatures
throughout the ranges of several species of crickets.

b. Neurobiological approach to phonotaxis
Song recognition and sound source localization, both expressed in the
phonotactic behavior of crickets, require auditory organs capable of air-
borne sound perception. In crickets and katydids auditory organs are located
in the proximal part of each fore tibia. In crickets each ear contains 50 to
60 receptor cells arranged in a row along the crista acustica (Eibl 1978),
and as recently demonstrated by Oldfield (1982, in prep.) in an Australian
katydid, such an arrangement reflects a peripheral tonotopic organization.
Each receptor within the whole organ is tuned to a certain sound frequency,
and behaviorally important frequencies may even be represented by more
than one receptor. The receptor cells are in close contact with the inner wall
of the 'acoustic trachea', a sound conducting pathway (Kleindienst et al.
1981), which itself has close contact with the larger posterior tympanum.
The smaller anterior tympanum in crickets is underlain by a thicker sheath

Insect Behavioral Ecology-'84 Huber 63

100- o u- A a 15 0C
3 o---o 220C

80- /' ...... 0o 300C
I *
20- i f' \5
\
o
t 6 0 "
0 0 1 \

40 1 /A A 3 \

I20- 3 0 7 15\0C
/ / dV22oC '.....,
2;,-A -.-/..------3 -o---,---- 'b----, ^

20 30 40 50 60 70 80 90
SP
Fig. 6. Temperature dependent phonotaxis in females of Gryllus bi-
maculatus with model chirped sounds. Ordinate: % of female tracking time
per stimulus time; abscissa: Sound pulse rates (SP), ranging from 20 to
90/1000 sees (11 to 50 pulses or syllables/sec). Open triangles are mean
values for several females that tracked at 150C (chirp period 700 msec),
open squares are mean values for several females that tracked at 22C
(chirp period 500 msec), open circles are mean values for several females
that tracked at 30C (chirp period 350 msec). Hatched bars at the bottom
diagram proportionate syllable interval periods for the natural males' songs
at the listed temperatures (modified from Doherty, in press).

of cells and is not in direct contact with the acoustic trachea and the re-
ceptors. However, both tympana add to hearing (Huber et al. in press), but
most effective is the posterior tympanum (see Larsen and Michelsen 1978).
Airborne sound reaches the tympana directly from the outside and the pos-
terior tympanum also from inside via the acoustic trachea that connects to
a spiracular opening in the prothoracic-mesothoracic body wall. Sound in-
duced oscillations of the tympana are required for hearing (Kleindienst
et al. 1983, Huber et al. 1984). These oscillations depend on sound pres-
sure differences and phase differences of the sound waves impinging on
the tympanum from both sides (Hill and Boyan 1977, Kleindienst et al.
1981, 1983). Such a 'pressure gradient receiver' (Michelsen and Nocke
1974) can be considered as a novel evolutionary design in animals with
small body sizes whose behaviorally relevant sound waves are proportionally
much greater in length relative to overall body length than for large
vertebrates, and whose diffraction by the body is therefore minimized.
Female Gryllus with one auditory organ destroyed are still able to 'rec-
ognize' their conspecific song. In the field (Klopffleisch 1973) and on the
treadmill (Huber et al. 1984) they begin to walk as soon as a normal
calling song is broadcast. They generally circle toward the side of thp intact

64 Florida Entomologist 68(1) March, 1985

ear according to the rule 'turn toward the ear most strongly stimulated'.
During this circling they also exhibit short episodes of tracking with typical
meandering, however, at an unexpected erroneous angle with respect to the
sound source. This meandering is thought to be associated with 'lateraliza-
tion' or 'binaural comparison' (Rheinlaender and Blitgen 1982) in two-
eared crickets. But peripheral binaural comparison is not available in one-
eared crickets, so the persistence of meandering in one-eared crickets de-
mands a new evaluation of the assumptions and data regarding the under-
lying mechanisms of meandering (Huber et al. in 1984).
Females who lose a foreleg during postembryonic development, even if
subsequently regenerated, lack both tympana on a regenerated leg. His-
tological examination shows that within this leg the tracheal organization
and the auditory receptor arrangement necessary for audition are also
missing. Nevertheless, these females track calling songs nearly as precisely
as those with both ears intact (Huber et al. 1984). There is no doubt
that binaural information normally guides the female most correctly to the
singing male; but even with monaural input some form of localization re-
mains. Phonotactic orientation in one-eared females is even 'improved' after
an initial period of monaural hearing deficit. This points to previously un-
recognised plasticity within the auditory pathway.
In a search for the neural mechanisms underlying sound source localiza-
tion, especially conspecific song recognition, we first turned our attention
to the functional capacities of the ears. When recording from single and
partly identified auditory receptors at the axonal level we found some axons
'tuned' (i.e. most sensitive) to the calling song sound frequency spectrum
of 4 to 5 kHz (figure 7), others tuned to the courtship song sound frequency
spectrum of 13 to 16 kHz, and still others sensitive to a broader range of
frequencies (3 to beyond 30 kHz) (Esch et al. 1980, Hutchings and Lewis
1981). Thus, the spectral domain of the cricket ear is not just designed for
intraspecific demands. It also has the capacity to perceive sound signals
outside the range produced by that species such as songs of other species or
sounds emitted by predators, sometimes even hunting bats (Popov and
Markovich 1982). Indeed, negative phonotaxis (interpreted as avoidance)
was reported in Teleogryllus by Moiseff et al. (1978) and by Pollack and
Hoy (1981).
Bioacousticians discovered rather early that many cricket species, even
those living sympatrically, emit calling songs comprised of a narrow band
of carrier frequencies within a rather limited overall range of sound fre-
quencies, usually between 2 and 7 kHz (Popov et al. 1974). The small over-
all range may be associated with the biomechanical properties of the wings
that radiate sound (Michelsen and Nocke 1974). This prospect raises the
question: Is the ear able to discriminate sound frequencies within this over-
all narrow band? The best example known to us is that of the two partly
sympatric species, Teleogryllus oceanicus and commodus. Males of T.
oceanicus call in the range of 4.2 to 4.7 kHz, those of T. commodus in the
range of 3.3 to 3.7 kHz (Hill 1974). Their sound spectra are separated by
only 300 to 500 Hertz. Hill (1974) showed by behavioral studies that fe-
males can discriminate model songs that differ only by these frequencies.
One mechanism by which discrimination of sound frequencies can be
improved beyond that already available at the receptor level involves

Insect Behavioral Ecology-'84 Huber 65

A B AC
ML
dB 100
SPL

80N\ /H

4.5 kHz 70dB I
H^602 3 4 6 8 s12 18

80dB

300 opm

60 ms 90 dB
PC

Fig. 7. Song encoding capacities of one primary auditory fiber in females
Gryllus campestris, and central projections of that fiber. A. Model chirps
with 4 pulses (syllables) at three different sound pressure levels (upper
traces) and responses of the same single identified primary auditory fiber
tuned to 4.5 kHz (lower traces), recorded semi-intracellularly. Its threshold
curve is shown in the inset of B. Note that this fiber copies the sound pat-
tern by bursts of nerve impulses, and encodes sound intensity by an in-
creasing number of spikes per burst, and by slightly increasing spike fre-
quency. B. Left half of the prothoracic ganglion seen in the horizontal plane
with the central part of the auditory fiber stained with the fluorescent dye
Lucifer yellow. The terminations of the fiber cover the area called the
auditory neuropile. AC, PC, anterior and posterior connectives respectively;
ML, midline of the ganglion; LN, prothoracic leg nerve that contains the
bundle of 55 to 60 auditory fibers (modified after Huber 1983).

'sharpening the tuning' (narrowing the sensitivity range) of neurons by
side-band inhibiton. Sound frequencies outside the best excitatory frequency
of the cell elicit an auditory input to the cell which has an inhibitory effect.
Side-band inhibition was demonstrated in cricket auditory pathways using
two-tone stimulation techniques and by recording from identified central
auditory neurons (Boyan 1981, Oldfield and Hill 1983, Hutchings and Lewis
1984, Boyd et al. 1984). Thus, cricket auditory pathways are in principle
capable of discriminating sounds which differ in pitch by a few hundred
Hertz. It is quite possible that such a mechanism-which deserves studies
at the network level-not only has evolved in interspecific contexts separating

66 Florida Entomologist 68(1) March, 1985

mate selection from predator avoidance, but also in the context of individual
song differences as in male/male aggressive signals etc. (Alexander 1961,
Lloyd 1980b).
The following discoveries, probably of adaptive significance, should be
evaluated in this context. In females of Gryllus bimaculatus, Boyan (1980,
1981) recorded and identified central neurons in the brain that were tuned
best to the conspecific courtship song carrier frequency (figure 8). When
the female's ear was stimulated simultaneously with two tones, these neurons
exhibited a clear suppression of synaptic potentials and spike responses to
the courtship song frequencies if the second tone was near the calling song
carrier frequency.
An example of behavioral correlate in a field situation is that females
approaching a calling male by phonotaxis receive only input from receptors
tuned to the 4 to 5 kHz band at a distance, because the song's sound energy
is greater in this band and is less attenuated by the surrounding vegetation
(Popov et al. 1974). Upon closer approach the ear will also receive higher
harmonics of the male's call (8 to 10 or 15 kHz). Some auditory receptors
are also tuned to lower and higher sound frequencies in addition to middle
ones; when close to the male the high frequency pathway is activated in
addition to the low frequency pathway, though less so because of the lower
energy of the higher harmonics. On the treadmill, when stimulated with
model calling songs of the natural pattern but at carrier frequencies higher
than 8 kHz, females tracked the sound source with an "erroneous" angle as
though they were orienting to a male at a location other than the speaker
(Thorson et al. 1982). This angle is carrier frequency dependent and has a
value of more than 90 degrees at frequencies above 12 kHz. Such frequencies
are present in the male's call when a female has approached closely. Since
we never observed a change in the phonotactic course in the field, it is
reasonable to postulate that a suppression of responses in neurons sensitive
to these higher frequencies which mediate "erroneous" angle tracking-or
even negative phonotaxis-improves the accuracy of phonotaxis and is of
adaptive significance, and/or is overridden by tactual, visual, or chemical
cues at such close ranges.
Bioacousticians have further shown that calling songs of different spe-
cies could easily be discriminated by their different temporal pattern, in-
cluding intra- and interchirp parameters (Alexander 1961; Popov et al.
1974, Otte and Alexander 1983). At the level of the ear, electrophysiological
recordings provide no clear-cut evidence of a species-specific temporal tuning
of auditory receptors. In other words, none of the 50 to 60 auditory sense
cells responded preferentially or exclusively to the conspecific pattern. In-
stead, the receptors copy in their impulse discharge all kinds of patterns,
including chirps and trills, bird songs, and tones as well as other environ-
mental noises broadcast within their relatively broad sensitive range. In
the temporal domain, then, the ear is designed to fit intraspecific demands;
however, the acuostic channel is open also for interspecific interactions.
Searching for 'temporal template properties' in the central neurons
requires an experimental strategy which is based upon results of behavioral
studies such as those reported earlier. We may discover neurons tuned to
conspecific song carrier frequencies that in addition encode the rhythm

Insect Behavioral Ecology-'84 Huber

SB
A
R 85 dB
10

NNC
OT \i 6 CS COS

0 5
SIN

2 3 5 7 101315 20
AN "Frequency [kHz3

\ P 100 --- 15kHz
100 em / 1 kHz

c 10 kHz
CEC t 50

0n `- %kHz
0
CT CT=TT TT*5 TT*10 TT+20
Fig. 8. Local brain auditory interneuron response suppression in females
of Gryllus bimaculatus. A. Right half of brain viewed from front with
camera lucida drawing of the interneuron inside (IN) ; NC, cell body; arrow
points to intracellular recording site; ON, lateral ocellar nerve; OT, optic
tract; AN, antennal nerve; CEC, circumesophageal connective. B. Supra-
threshold response of the neuron in the range of 2 to 20 kHz, equal sound
pressure levels, with optimal response to the sound frequency band of the
courtship song (COS) and a much weaker response to the band of the
calling song sound frequency spectrum (CS); points on curve represent
means, bars standard deviations. C. Response to an artificial courtship sound
alone and with a second lower frequency sound played simultaneously at
different intensities. The neuron is first stimulated with a control-tone (CT,
15 kHz, dashed line = 100% of response) in the range to which the cell is
optimally responsive. Subsequently adding a tone of either 1 kHz (closed
circles) or 10 kHz (open circles, upper curve) only weakly suppresses the
response relative to that at 15 kHz; however, if the added tone (TT) is 5
kHz (in the range of the carrier frequencies of the calling song, open
circles, lower curve) the response to 15 kHz is significantly suppressed
especially at increasing intensities of the TT. CT, control tone only; CT =
TT, sound intensities of CT and TT are equal; TT +5, +10, +20, TT was
delivered with 5, 10 or 20 dB higher intensity than the CT. With TT +20
db at 5 kHz, mimicking the natural situation, the response to 15 kHz is
reduced more than 80% (modified after Boyan 1981).
pattern of that conspecific song. Only then we can call such neurons part of
the recognizerr template'.
We started our experimental strategy within the prothoracic ganglion
and continued toward the brain. Using intracellular recording and marking

68 Florida Entomologist 68(1) March, 1985

1mm

B CB IPSI CB CONTRA BOTH

100 ms
Fig. 9. Morphology and physiology of the prothoracic intraganglionic
Omega-neuron in Gryltus campestris females. A. Lucifer yellow fill of one
member of the mirror-image pair of the Omega-neurons type I. The faint
autofluorescent stripe in the middle marks the midline of the prothoracic
ganglion. IPSI, the half of the ganglion where the neuronal cell body,
neurite and ipsilateral field of arborizations are located (input-area of the
neuron); CONTRA, axon traversing from ipsilateral to contralateral side
of the ganglion and terminating in the contralateral field of arborizations
(output-area of the neuron). B. Response of the neuron to monaural stimula-
tion by a model calling song with a central carrier frequency of 5 kHz at
80 dB SPL, of 4 chirps. CB IPSI, patterned discharge of burst of nerve
impulses riding on top of depolarization waves with ipsilateral stimulation,
each burst reflects the response to one syllable (pulse). CB CONTRA,
contralateral stimulation shows syllable related inhibitory postsynaptic
potentials (downward directed deflections). BOTH, binaural stimulation
(input from both ears simultaneously) gives patterned discharge bursts
with a slight reduction in the number of impulses per burst due to contra-
lateral inhibition. These recordings were obtained intracellularly at the
input region of the neuron (modified after Wohlers and Huber 1982).

of neurons with the fluorescent dye Lucifer yellow (see figures 9 and 10)
we delimited a family of mirror-image nerve cells which responded to the
calling song stimulus (Wohlers and Huber 1978, 1982). Among them were
several cells tuned to the carrier frequency of 4 to 5 kHz that repeated the
pulse and chirp rates by bursts of nerve impulses, figures 9 and 10. One
pair of cells, called Omega-neurons (figure 9), forms a two cell network

Insect Behavioral Ecology-'84 Huber 69

SPL n 18
(dB) 70 s
p 12
60
AN1 k6
s
0 20 60 120
30 -4.5 kHz 70 dB

D 90 n
SPL
(dB) EoE-II
70 -1 k 8 u
AN2 ri
60 01
0 40 80 120 160
50 5 kHz 70dB
2 3 5 6 8to 1121 Time [ms]
kHz

Imm

Fig. 10. Morphologies and responses of two types of plurisegmental
ascending prothoracic auditory interneurons in Gryllus campestris females.
A. Lucifer yellow fill of one member of the mirror-image pair of the ascend-
ing neuron type I (AN1); seen in the horizontal plane with cell body,
neurite crossing the ganglionic midline, and densely packed arborization
field with axon passing to the head ganglion on the contralateral side.
B. Lucifer yellow fill of one member of the mirror-image pair of ascending
neuron type II (AN2); seen in the horizontal plane with cell body, thin
neurite which crosses the ganglionic midline and arborizes in the larger
dendritic field that extends to the entrance of the auditory nerve, and with
axon passing to the head ganglia on the contralateral side. In both A and B
the autofluorescent stripe marks the midline (ML) of the prothoracic
ganglion. C. AN1 is tuned to the calling song carrier frequency in the range
of 4 to 5 kHz, as seen by the threshold curves (solid and dashed lines). It
copies the syllable pattern within each calling chirp with high fidelity at all
sound intensities, as seen in the poststimulus response pattern histogram.
Each single histogram correlates to one syllable of the four pulses (syllables)
at 4.5 kHz and 70 dB SPL. D. AN2 is more broadly.tuned and sensitive
below 60 dB SPL in the range between 3 and 16 kHz, as seen by four
threshold curves (solid lines). The response pattern histogram is shown also
and documents that AN2 copies the calling chirp as a whole. Stimulus was
presented with 5 kHz, 4 pulses and 70 dB SPL. Threshold curves in C and
D: Ordinate, sound intensity in dB SPL to elicite a response just above
threshold; abscissa, sound frequency range covering 2 to 16 kHz. Histograms
in C and D: Ordinate, number (n) of nerve impulses (spikes) sampled
consecutively for each 2 ms time period during a stimulus series of 20
model calling chirps; abscissa, time in msec where 0 means onset of the
first syllable (pulse) of the chirp. (A, B, modified after Wohlers and Huber
1982; C, D, modified after Stout and Huber 1981).
with reciprocal inhibition (Wohlers and Huber 1978, 1982; Selverston et al.
1985). Each cell receives excitatory input from only one ear and is inhibited
by input from the other ear conveyed by the mirror-image partner cell.
Since the terminals of the primary auditory fibers end within the pro-
thoracic ganglion in what is called the ipsilateral neuropile, with no

70 Florida Entomologist 68(1) March, 1985

crossing to the contralateral side, binaural reciprocal inhibition by the
Omega-cells is one mechanism by which a cricket may enhance directional
information processing in the calling song range (Kleindienst et al. 1981).
In principle, such a network is also able to assist pattern recognition (Wiese
1978, 1983, 1984), if the dynamic properties of the network are such that
only phonotactically effective patterns are copied and transmitted with
fidelity. Recently, Selverston et al. (1985), killed one Omega-cell in female
crickets by photoinactivation, and thus withdrew it from the interaction
network. Under these conditions the remaining Omega-cell received and
transmitted excitatory input from 'its ear' but the inhibitory input from the
contralateral ear was missing. With another single test on our walking
compensator a female with one Omega-cell killed was expected to change
direction finding during phonotaxis, because a part of binaural comparison
was lost. This female exhibited clear and precise tracking of the sound
source before Omega-cell killing, and after killing showed only a slight
asymmetry in tracking to the expected side of the intact Omega-cell. We
have a hint, but many more such behavioral experiments are needed to
draw firm conclusions.
Two other pairs of mirror-image central neurons with axons ascending
to the brain (figure 10) are particularly interesting in our context of song
recognition. One pair, the AN1 type cells, are best tuned to the calling song
carrier frequency and copy the natural song pattern at all intensities. Each
member of the pair receives auditory input from only one ear, from the ear
whose primary auditory fiber terminals overlap with the dentritic field of
the AN1 cell. It is conceivable that this cell is involved in one-ear song
recognition, however, it can not be called an intrinsic element of the 'rec-
ognizer template' because it is not selectively tuned only to the pattern of
the conspecific song (Stout and Huber 1972; Schildberger 1984). Another
pair of ascending neurons-AN2 type cells-(figure 10) is equally or more
sensitive at higher sound frequencies and responds to the calling chirp with
a rather continuous discharge of impulses throughout the whole chirp. This
neuron does not copy the pulse repetition rate but encodes the chirp rate. It
could play a role in those females which track bursts, as well as function in
negative phonotaxis and in avoidance of predators.
There are several possibilities for the manifestation of a recognizerr
template' at the neural level. Perhaps there are neurons present at higher
levels of the auditory pathway that are tuned to the conspecific temporal
pattern and thus reflect phonotactic demands, or the template may be more
widespread, perhaps laid down in a multiganglionic network, which due to
cooperation of its members represents the recognizerr' etc. We favor the
first possibility because very recently Schildberger (1984) identified a class
of brain neurons that showed a clear correlation in their response (evalu-
ated by the number of spikes per chirp) to the phonotactically effective
range of pulse repetition rates. These neurons have band-pass properties
and it is such a class of brain cells which we consider at present to be the
most intimate part of the recognizerr template' (figure 11).
Further studies will show how these brain neurons are connected to each
other within the auditory pathway, and how their band-pass properties are
achieved. We expect to unravel mechanisms at the cellular and synaptic
levels that provide a basis for temporal tuning, and furthermore, to get

Insect Behavioral Ecology-'84 Huber 71

information about how these neurons elicit walking and orientation char-
acteristics known to be present in phonotaxis (Weber et al. in prep., Stout
et al. in prep.). On a longer scale, other questions can be attacked. For in-
stance, are the neurons providing the substrates for recognition affected by
temperature in much the same way as shown for phonotaxis? Are there
sex-dependent similarities and differences, including temperature matching
within sender and receiver? Is temperature matching merely a function of
chance and eventual close proximity? A further step may even deal with
'trade-off' phenomena recently studied in phonotaxis of Gryllus bimaculatus
(Doherty in prep.) with the outcome that at the limits of effective neural
functional ranges different temporal parameters are weighted by the fe-
males. We finally hope to contribute as neurobiologists to bridging the gap
that still exists between us and behavioral ecologists in this most fascinating
field of communication strategies.

CONCLUDING REMARKS

I have tried to outline mutual interests and ways for future cooperation
between behavioral ecologists and neurobiologists by discussing examples of
adaptive behaviors and the present state of knowledge concerned with their
neural bases. What the reader should understand is that the time has come
when these two fields would greatly benefit from cross talk and cooperation
with each other. Insect neurobiology has yet a rather limited influence on
insect behavioral ecology because many of the behavioral strategies studied
by ecologists are too complex to be attacked by neurobiologists with the
present day technical know how and expertise. But this will change as
neurobiologists become better able to study behaviors at the neuronal level
in freely moving animals, and important beginnings in this were made
during the last decade (Huber 1983 a,b).
But there is still another reason for the existing gap between behavioral
ecology and neurobiology, which is of mental origin. There is a limited
cross-talk between the two groups, too little "looking across the fence into
each others flowering garden". We must experience each others problems,
must find a common language to discuss possible ways to solve them, and no
longer only step side by side. One way toward this desired goal is for neuro-
biologists to try to become apprentices in behavioral ecology, and, vice versa,
behavioral ecologists should not hesitate to dive a bit deeper into the pond
where neurobiologists play. This needs patience and open minds on both sides.
What behavioral ecologists and evolutionary biologists could perhaps
gain from neurobiologists-aside from the problems they are interested in-
is to learn to look for tactics that can be used experimentally to make pre-
dictions which can be falsified or verified. At present, neurobiologists are
confronted with and suffer from a plethora of speculations and hypotheses
formulated by behavioral ecologists, often without a thought for an experi-
mental solution. On the other hand, what neurobiologists could gain from
behavioral ecologists is to become open to the diversities in strategies and
tactics animals have evolved, and to be confronted with field work and the
comparative method. Then perhaps we can all gain a closer understanding
of the real biological questions. This is a big challenge for both, let's start.

Florida Entomologist 68 (1)

March, 1985

Spikes/Chirp

%

I

Phonotaxis
S[
C

50
Syllable

A 0 0

0 0 0

74 98
Repetition Interval [ms]

Fig. 11. Temporal selectivity of brain neurons in females of Gryllus
bimaculatus in relation to song recognition, and generalized phonotactic
tracking range for both G. bimaculatus and G. campettris (hatched area).
Upper: Diagram of model calling chirps (arranged vertically), chirp
energy equal (nearly equal length of chirps, chirp repetition intervals 500
msec). Within each chirp the duty cycle was kept 50%, that is the individual
syllable (pulse) durations were equal to the intersyllable (interpulse) dura-
tions. The model chirps were presented at 5 kHz and 80 dB SPL. They dif-
fered in the number of syllable (pulses) per chirp (from 25 at the left to 3
at the right), in syllable (pulse) duration (length of the black bars) and
intersyllable (interpulse) duration, and most importantly in syllable (pulse)
repetition interval (ranging from 8 to 98 msec). Lower: Phonotactic re-
sponse range (hatched area) of females to these chirps (listed above). 100%
= maximal phonotactic (right ordinate) or maximal neuronal response
(left ordinate in terms of spikes recorded from any female per chirp dura-
tion). The phonotactic range was replotted from Thorson et al. 1982, and
from Doherty in prep. Superimposed on the hatched area are data points
that represent responses of identified brain neurons. Open and closed
triangles represent recordings from brain neurons that respond maximally
to short syllable (pulse) repetition intervals (range 8 to about 50 msec

26
26

I
I
I

Insect Behavioral Ecology-'84 Huber 73

ACKNOWLEDGEMENTS
This paper is dedicated to Thomas E. Moore, Richard D. Alexander, and
Thomas J. Walker, close friends for many years, who as behavioral ecolo-
gists and evolutionary biologists influenced my own research in insect neuro-
biology. They made me aware of how important a cross-talk can become. I
am also most grateful to James E. Lloyd, who has encouraged me to think
about mutualistic coexistence between behavioral ecologists and neurobiolo-
gists and has invited me to attend his symposium. He and T. E. Moore have
critically read this paper and given valuable advice. Finally, I thank my
colleagues in our Abteilung, former guests and students, and in addition my
technical staff.

LITERATURE CITED
ALEXANDER, R. D. 1961. Aggressiveness, territoriality, and sexual behavior
in field crickets (Orthoptera: Gryllidae). Behaviour 17: 130-223.
BLUM, M. S., AND N. A. BLUM. 1979. Sexual selection and reproductive
competition in insects. Academic Press, New York, San Francisco
London.
BOYAN, G. S. 1980. Auditory neurons in the brain of the crickets, Gryllus
bimaculatus (DeGeer). J. Comp. Physiol. 140: 81-93.
.1981. Two-tone suppression of an identified auditory neurone in
the brain of the cricket, Gryllus bimaculatus (DeGeer). J. Comp.
Physiol. 144: 117-25.
BOYD, P., R. KUHNE, S. SILVER, AND B. LEWIS. 1984. Two-tone suppression
and song coding by ascending neurones in the cricket Gryllus
campestris L.. J. Comp. Physiol. 154: 423-30.
CAMHI, J. M. 1984. Neuroethology. Nerve cells and the natural behavior of
animals. Chapter 4: A case study in neuroethology: The escape sys-
tem of the cockroach, 79-105, Sinauer Associate Inc. Publishers,
Sunderland, Massachusetts.
DOHERTY, J. A. 1984. Temperature coupling and 'trade-off' phenomena in
the acoustic communication of the cricket, Gryllus bimaculatus
(DeGeer). J. Exp. Biol. (in press).
DOHERTY, J. A., AND F. HUBER. 1983. Temperature effects on acoustic com-
munication in the cricket, Gryllus bimaculatus (DeGeer). Verh.
Dtsch. Zool. Ges. A 188, G. Fischer-Verlag, Stuttgart.
EIBL, E. 1978. Morphology of the sense organs in the proximal parts of the
tibiae of Gryllus campestris L. and Gryllus bimaculatus DeGeer
(Insecta, Ensifera). Zoomorph. 89: 185-205.
ELSNER, N. 1968. Die neuromuskultlren Grundlagen des Werbeverhaltens
der roten Keulenheuschrecke, Gomphocerippus rufus L.. Z. Vergl.
Physiol. 60: 308-50.
-- 1974. Neural Economy: Bifunctional muscles and common pattern

with above 75% response), open and filled squares represent recordings
from neurons that respond best to longer syllable (pulse) repetition inter-
vals (range 34 to 98 msec above 75% response). Open and filled circles
represent neurons that match in their response most closely the range of
phonotactically effective syllable repetition intervals (26 to 60 msec above
75% response). Each symbol at every syllable (pulse) repetition interval is
a data point from a different animal. Chirp diagrams lie directly above
their proportionate syllable (pulse) repetition intervals (after Schildberger
1984).

74 Florida Entomologist 68 (1) March, 1985

elements in leg and wing stridulation of the grasshopper, Stenobothrus
rubicundus Germ. (Orthoptera: Acrididae). J. Comp. Physiol. 89:
227-36.
-- 1975. Neuroethology of sound production in gomphocerine grass-
hoppers (Orthoptera: Acrididae) II. Neuromuscular activity under-
lying stridulation. J. Comp. Physiol. 97: 291-322.
ESCH, H., F. HUBER, AND D. W. WOHLERS. 1980. Primary auditory neurons
in crickets: Physiology and central projections. J. Comp. Physiol.
137: 27-38.
GWYNNE, D. T., AND G. K. MORRIS. 1983. Orthopteran Mating Systems.
Sexual competition in a diverse group of insects. Westview Press,
Boulder, Colorado.
HAGIWARA, S., AND A. WATANABE. 1956. Discharges in motoneurons of a
cicada. J. Cell. and Comp. Physiol. 47: 415-28.
HELVERSEN, D. VON. 1972. Gesang des Miinnchens und Lautschema des
Weibchens bei der Feldheuschrecke Chorthippus biguttulus (Orthop-
tera: Acrididae). J. Comp. Physiol. 81: 381-422.
-- AND O. VON HELVERSEN. 1975. Verhaltensgenetische Untersuchungen
am akustischen Kommunikationssystem der Feldheuschrecken
(Orthoptera: Acrididae). II. Das Lautschema von Artbastarden
zwischen Chorthippus biquttulus and Ch. mollis. J. Comp. Physiol.
104: 301-23.
---- AND 1983. Species recognition and acoustic localization in
acridid grasshoppers. A behavioral approach. In: Neuroethology and
Behavioral Physiology (eds. F. Huber and H. Markl) pp. 95-107,
Springer Verlag, Berlin, Heidelberg.
HILL, K. G. 1974. Carrier frequency as a factor in phonotactic behavior of
female crickets (Teleogryllus commodus). J. Comp. Physiol. 93: 7-18.
.1974. Acoustic communication in the Australian field crickets
Teleogryllus commodus and T. oceanicus (Orthoptera: Gryllidae).
Thesis, University of Melbourne.
HILL, K. G., AND G. S. BOYAN. 1977. Sensitivity to frequency and direction
of sound in the auditory system of crickets (Gryllidae). J. Comp.
Physiol. 121: 79-97.
HORMANN-HECK, S. VON. 1957. Untersuchungen fiber den Erbgang einiger
Verhaltensweisen bei Grillenbastarden (Gryllus campestris L. Gryllus
bimaculatus DeGeer). Z. Tierpsychol. 14: 137-83.
HOYLE, G. 1964. Exploration of neuronal mechanisms underlying behavior
in insects. In: Neural Theory and Modeling (ed. R. F. Reiss) pp.
346-376, Stanford University Press, Stanford, California.
.1970. Cellular mechanisms underlying behavior-Neuroethology.
Adv. Insect Physiol. 7: 349-444.
---. 1983. On the way to neuroethology: The identified neuron approach.
In: Neuroethology and Behavioral Physiology (eds. F. Huber and
H. Markl) pp. 9-25, Springer-Verlag, Berlin, Heidelberg.
HUBER, F. 1983a. Implications of insect neuroethology for studies on
vertebrates. In: Advances in Vertebrate Neuroethology (eds. J. P.
Ewert, R. R. Capranica, D. J. Ingle) Series A: Life Sciences 56: 91-
138, Plenum Press Corporation.
- 1983b. Neural correlates of orthopteran and cicada phonotaxis. In:
Neuroethology and Behavioral Physiology (eds. F. Huber and
H. Markl) pp. 108-35, Springer-Verlag, Berlin, Heidelberg.
HUBER, F., D. W. WOHLERS, AND T. E. MOORE. 1980. Auditory nerve and
interneurone responses to natural sounds in several species of cicadas.
Physiol. Ent. 5: 25-45.
----, H. U. KLEINDIENST, T. WEBER, AND J. THORSON. 1984. Auditory

Insect Behavioral Ecology-'84 Huber

behavior of the cricket. III. Tracking of male calling song by
surgically and developmentally one-eared females, and the curious
role of the anterior tympanum. J. Comp. Physiol. 155: 725-738.
HUTCHINGS, M., AND B. LEWIS. 1981. Response properties of primary
auditory fibers in the cricket, Teleogryllus oceanicus (Le Guillou).
J. Comp. Physiol. 143: 129-34.
AND -- 1984. The role of two-tone suppression in song coding
by ventral cord neurones in the crickets, Teleogryllus oceanicus (Le
Guillou). J. Comp. Physiol. 154: 103-12.
JOSEPHSON, R. K. 1973. Contraction kinetics of the fast muscles used in
singing by a katydid. J. Exp. Biol. 59: 781-801.
1984. Contraction dynamics of flight and stridulatiory muscles of
tettigoniid insects. J. Exp. Biol. 108: 77-96.
--- AND R. C. HALVERSON. 1971. High frequency muscles used in sound
production by a katydid. I. Organization of the motor system. Biol.
Bull. 141: 411-433.
KALMRING, K., R. KUHNE, AND B. LEWIS. 1983. The acoustic behaviour of
the bushcricket Tettigonia cantans. III. Coprocessing of auditory and
vibratory information in the central nervous system. Behavioural
Processes 8: 213-228.
KEUPER, A., AND R. KUHNE. 1983. The acoustic behaviour of the bushcricket
Tettigonia cantans. II. Transmission of airborne sound and vibration
signals in the biotope. Behavioural Processes 8: 125-145.
KLEINDIENST, H. U., U. T. KOCH, AND D. W. WOHLERS. 1981. Analysis of
the cricket auditory system by acoustic stimulation using a closed
sound field. J. Comp. Physiol. 141: 283-296.
-- D. W. WOHLERS, AND O. NAESBYE-LARSEN. 1983. Tympanal mem-
brane motion is necessary for hearing in crickets. J. Comp. Physiol.
151: 397-400.
KLOPFFLEISCH, K. 1973. Ethologische Untersuchungen an der Feldgrille,
Gryllus campestris L., an natiirlichen Standorten. Staatsarbeit der
Universitit zu KSln.
KRIECHBAUM, W. 1983. Zeitstruktur des Lockgesanges bei Gryllus campes-
tris L.. Diplomarbeit der Universitit Tiibingen.
LARSEN, O. N., AND A. MICHELSEN. 1978. Biophysics of the ensiferan ear.
III. The cricket ear as a four input system. J. Comp. Physiol. 123:
217-27.
LATIMER, W., AND A. SCHATRAL. 1983. The acoustic behaviour of the bush-
cricket Tettigonia cantans. I. Behavioural responses to sound and
vibration. Behavioural Processes 8: 113-24.
LLOYD, J. E. 1965. Aggressive mimicry in Photuris: Firefly femmes fatales.
Science 149: 653-54.
1975. Aggressive mimicry in Photuris fireflies: Signal repertoires
by femmes fatales. Science 187: 452-53.
- 1980a. Male Photuris fireflies mimic sexual signals of their females'
prey. Science 210: 669-71.
-- 1980b. Sexual selection: Individuality, identification, and recogni-
tion in a bumblebee and other insects. Florida Ent. 64: 89-118.
-- 1981a. Mimicry in the sexual signals of fireflies. Sci. Am. 245 (7) :
138-45.
1981b. Firefly mate-rivals mimic their predators and vice versa.
Nature 290: 498-500.
LLOYD, J. E. 1983. Bioluminescence and communication in insects. Annu.
Rev. Entomol. 28: 131-60.
MARKL. H. 1967. Die Verstiindigung durch Stridulationssignale bei
Blattschneiderameisen. I. Die biologische Bedeutung der Signale.

76 Florida Entomologist 68 (1) March, 1985

Z. Vergl. Physiol. 57: 299-330.
1968. Die Verstandigung durch Stridulationssignale bei Blattsch-
neiderameisen. II. Erzeugung und Eigenschaften der Signale.
Z. Vergl. Physiol. 60: 103-50.
1970. Die Verstandigung durch Stridulationssignale bei Blatt-
schneiderameisen. III. Die Empfindlichkeit fir Substratvibrationen.
Z. Vergl. Physiol. 69: 6-37.
-- -, AND S. FUCHS. Klopfsignale mit Alarmfunktion bei Rofameisen
(Camponotus, Formicidae, Hymenoptera) Z. Vergl. Physiol. 76:
204-25.
MICHELSEN, A., AND H. NOCKE. 1974. Biophysical aspects of sound com-
munication in insects. Adv. Insect Physiol. 10: 247-96.
MOISEFF, A., G. S. POLLACK, AND R. R. HOY. 1978. Steering responses of
flying crickets to sound and ultrasound: Mate attraction and predator
avoidance. Proc. Natl. Acad. Sci. 75: 4052-56.
MOORE, T. E., AND R. T. SAWYER. 1966. The mechanism of cicada timbal
action (Insecta: Homoptera: Cicadidae). American Zool. 5: 509.
- AND J. KAUSCH. 1975. 17-year cicadas (Magicicada septendecim L.).
Publications of the University of Michigan, Museum of Zoology,
Movie, colour and sound, half hour.
MORRIS, G. K., AND J. H. FULLARD. 1983. Random noise and congeneric dis-
crimination in Conocephalus (Orthoptera: Tettigoniidae). In: Orthop-
teran Mating Systems (eds. D. T. Gwynne and G. K. Morris) pp.
73-96, Westview Press, Boulder, Colorado.
NICKLAUS, R. 1965. Die Erregung einzelner Fadenhaare von Periplaneta
americana in Abhiingigkeit von der Gr6iie und Richtung der Auslen-
kung. Z. Vergl. Physiol. 50: 331-62.
OLDFIELD, B. P. 1982. Tonotopic organization of auditory receptors in
Tettigoniidae (Orthoptera: Ensifera). J. Comp. Physiol. 147: 461-69.
-- AND K. G. HILL. 1983. The physiology of ascending auditory
interneurons in the tettigoniid Caedicia simplex (Orthoptera: En-
sifera): Response properties and a model of integration in the
afferent auditory pathway. J. Comp. Physiol. 152: 495-508.
OTTE, D., AND R. D. ALEXANDER. 1983. The Australian Crickets (Orthop-
tera: Gryllidae). Monograph 22. The Academy of Natural Sciences
of Philadelphia. Allen Press Inc., Lawrence, Kansas.
POLLACK, G. S., AND R. R. HOY. 1981. Phonotaxis in flying crickets: Neural
correlates. J. Insect Physiol. 27: 41-45.
F. HUBER, AND T. WEBER. 1984. Frequency.and temporal pattern-
dependent phonotaxis of crickets (Teleogryllus oceanicus) during
tethered flight and compensated walking. J. Comp. Physiol. 154: 13-26.
PoPov, A. V., V. F. SHUVALOV, I. D. SVETLOGORSKAYA, AND A. M. MARKOVICH.
1974. Acoustic behavior and auditory system in insects. In: Mechano-
reception (ed. J. Schwartzkopff) Abh. Rhein. Westf. Akad. Wiss 53:
281-306.
AND A. M. MARKOVICH. 1975. Spectrum of the calling songs,
phonotaxis and the auditory system in the cricket, Gryllus bimaculatus
DeGeer. J. Evol' uzionnoj Biochimii i Fisiologii 11: 453-60.
POPOv, A. V., AND A. M. MARKOVICH. 1982. Auditory interneurones in the
prothoracic ganglion of the cricket, Gryllus bimaculatus DeGeer. II.
A high-frequency ascending neurone (HF, AN). J. Comp. Physiol.
146: 351-59.
PREISS, R., AND E. KRAMER. 1983. Stabilization of altitude and speed in
tethered flying gypsy moth males. Influence of (+) and (-) dis-
palure. Physiol. Ent. 8: 55-68.
PRINGLE, J. W. S. 1954. A physiological analysis of cicada song. J. Exp.

Insect Behavioral Ecology-'84 Huber 77

Biol. 31: 525-60.
RATHMAYER, W. 1978. Venoms of Sphecidae, Pompilidae, Mutilidae and
Bethylidae. In: Handbook of Experimental Pharmacology (ed.
S. Bettini) 48: 661-90, Springer-Verlag, Berlin, Heidelberg, New
York.
REID, K. H. 1971. Periodical cicadas: Mechanism of sound production. Sci-
ence 172: 949-51.
RHEINLAENDER, J., AND G. BLXTGEN. 1982. The precision of auditory
lateralization in the cricket, Gryllus bimaculatus. Physiol. Ent. 7:
209-18.
RIEDE, K., F. HUBER, AND D. VARJU. 1979. Quantitative analysis of the
spatial relationships between male and female grasshoppers during
courtship. Naturwissenschaften 66: 370.
ROEDER, K. D. 1959. A physiological approach to the relation between prey
and predator. Smithson. Misc. Coll. 137: 287-306.
SCHILDBERGER, K. 1984. Temporal selectivity of identified auditory neurons
in the cricket brain. J. Comp. Physiol. 155: 171-185.
SCHMITZ, B., H. SCHARSTEIN, AND G. WENDLER. 1982. Phonotaxis in Gryllus
campestris L. (Orthoptera, Gryllidae). I. Mechanism of acoustic
orientation in intact female crickets. J. Comp. Physiol. 148: 431-44.
AND 1983. Phonotaxis in Gryllus campestris L.
(Orthoptera, Gryllidae). II. Acoustic orientation in female crickets
after occlusion of single sound entrances. J. Comp. Physiol. 152:
257-64.
SELVERSTON, A. L., H. U. KLEINDIENST, AND F. HUBER. 1985. Synaptic con-
nectivity between cricket auditory interneurons as studied by selec-
ive photoinactivation. J. Neurosci. (in press).
SIMMONS, P. 1977. Neuronal generation of singing in a cicada. Nature 270:
243-45.
-- AND D. YOUNG. 1978. The tymbal mechanism and song patterns of
the bladder cicada, Cystosoma saundersii. J. Exp. Biol. 76: 27-45.
STOUT, J. F., AND F. HUBER. 1972. Responses of central auditory neuron of
female crickets (Gryllus campestris L.) to the calling song of the
male. Z. Vergl. Physiol. 76: 302-13.
1981. Responses to features of the calling song by ascending
auditory interneurones in the cricket Gryllus campestris. Physiol.
Ent. 6: 199-212.
--, C. H. DEHAAN, AND R. W. MCGHEE. 1983. Attractiveness of the
male Acheta domesticus calling song to females. I. Dependence on
each of the calling song features. J. Comp. Physiol. 153: 509-21.
TAUTZ, J. 1977. Reception of medium vibration by thoracal hairs of cater-
pillars of Barathra brassicae L. (Lepidoptera, Noctuidae). I. Mechan-
ical properties of the receptor hairs. J. Comp. Physiol. 118: 13-31.
1978. Reception of medium vibration by thoracal hairs of cater-
pillars of Barathra brassicae L. (Lepidoptera, Noctuidae). II. Re-
sponse characteristics of the sensory cell. J. Comp. Physiol. 125:
67-77.
TAUTZ, J., AND H. MARKL. 1978. Caterpillars detect flying wasps by hairs
sensitive to airborne vibration. Behav. Ecol. Sociobiol. 4: 101-10.
THORNHILL, R., AND J. ALCOCK. 1983. The Evolution of Insect Mating
Systems. Harvard University Press, Cambridge Mass.
THORSON, J., T. WEBER, AND F. HUBER. 1982. Auditory behavior of the
cricket. II. Simplicity of calling-song recognition in Gryllus, and
anomalous phonotaxis at abnormal carrier frequencies. J. Comp.
Physiol. 146: 361-78.
WALKER, T. J. 1957. Specificity in the response of female tree crickets

Florida Entomologist 68 (1)

March, 1985

(Orthoptera, Gryllidae, Oecanthinae) to calling songs of the males.
Ann. Ent. Soc. Am. 50: 626-36.
WEBER, T., J. THORSON, AND F. HUBER. 1981. Auditory behavior of the
cricket. I. Dynamics of compensated walking and discrimination
paradigms on the Kramer Treadmill. J. Comp. Physiol. 141: 215-32.
WENDLER, G., M. DAMBACH, B. SCHMITZ, AND H. SCHARSTEIN. 1980. Analysis
of the acoustic orientation behavior in crickets (Gryllus campestris
L.). Naturwissenschaften 67: 99.
WESTIN, J., J. J. LANGBERG, AND J. M. CAMHI. 1977. Responses of giant
interneurons of the cockroach Periplaneta americana to wind puffs of
different directions and velocities. J. Comp. Physiol. 121: 307-24.
WIESE, K. 1978. Recurrent inhibition within the acoustic pathway of
Gryllus bimaculatus as the basis of temporal filtering. Verh. Dtsch.
Zool. Ges. A 168, G. Fischer-Verlag, Stuttgart.
1983. Akustische Kommunikation bei der Grille. Eigenschaften
der bilateral inhibitorischen Kommissur in der HSrbahn. Verh. Dtsch.
Zool. Ges. A 180, G. Fischer-Verlag, Stuttgart.
1984. Neural enhancement of directionality in sensory pathways of
two arthropods, the cricket and the crayfish. In: Localization and
Orientation in Biology and Engineering (eds. D. Varju and H. U.
Schnitzler) pp. 186-191, Springer-Verlag, Berlin, Heidelberg.
WILSON, D. M. 1962. Bifunctional muscles in the thorax of grasshoppers.
J. Exp. Biol. 39: 669-77.
WOHLERS, D. W., AND F. HUBER. 1978. Intracellular recording and staining
of cricket auditory interneurons (Gryllus campestris L., Gryllus bi-
maculatus DeGeer). J. Comp. Physiol. 127: 11-28.
AND 1982. Processing of sound signals by six types of
neurons in the prothoracic ganglion of the cricket, Gryllus campestris
L. J. Comp. Physiol. 146: 161-73.
YOUNG, D. 1972. Neuromuscular mechanism of sound production in Aus-
tralian cicadas. J. Comp. Physiol. 79: 343-62.
-, AND R. K. JOSEPHSON. 1983a. Mechanisms of sound production and
muscle contraction kinetics in cicadas. J. Comp. Physiol. 152: 183-95.
- AND 1983b. Pure-tone songs in cicadas with special refer-
ence to the genus Magicicada. J. Comp. Physiol. 152: 197-207.

AC

-13 kH2

LN

80dB

300 mWn

i,-,rii- 90 dB

PC I
PC ^

ML
I

1 mm

CB IPSI

ar~j~Vk~V
man

CB CONTRA
m m m

-r

p -

100 ms

BOTH

c I I I
SPL ? I
\ ,.

ANI ,
0 l ,* ..

,\ -. k

D_ I

O B -*, I, *I ,

Id "" - .. .
... I 1I

AN.' '
6 \0

7 J ., 6. '1i II
kHz

0 A0 60 120
&.5 -kH 70 US

(.u uso
5 kHz 70dB
Timn [tns]

Insect Behavioral Ecology-'84 Dewsbury

FROM FLIES TO MICE-AND BACK AGAIN

DONALD A. DEWSBURY*

As the laws of nature must be the same for all beings, the
conclusions furnished by this group of insects must be ap-
plicable to the whole organic world...
Henry Walter Bates, 1910, p. 348

SYNOPSIS
Our goal in science is to generate principles of generality. Information
on insects has been important in the development of our program of research
on rodent reproductive behavior; I hope that the reverse effect might also
be realized. Mammalian copulatory patterns can be classified with respect
to locking, thrusting, multiple intromissions, and multiple ejaculations; ap-
plication of a similar classification scheme might be useful with insects. In
both insects and rodents variation in male genitalia and accessory glands
appear correlated with reproductive behavior. Patterns of sperm competition
are focal to the evolution of mating systems. Because there are important
species differences in sperm competition, detailed study of both the basic
pattern and the dynamics of sperm competition in a variety of species is
warranted. There is evidence for female mate choice in both insects and
rodents. Further, in both insects and rodents the capacity of males to pro-
duce ejaculates is limited; this implies a role for male choice in mate selec-
tion. Principles of the greatest generality will be developed if investigators
can synthesize information from a wide range of taxa.

Our goal in developing a science of behavior is to elucidate principles of
broad generality. Such principles ought to be applicable, with appropriate
caution, across a wide range of taxa. However, there is an increasing trend
in the biological sciences for students of different groups of animals to
isolate themselves from other such groups. I believe that such isolation is
detrimental to the search for general principles and that there is much to
be gained by efforts to break down such barriers. Hopefully, the result will
be hybrid vigor-not hybrid sterility.
I am a student of the evolution of mammalian reproductive behavior.
However, one of my trade secrets is that many of my research ideas have
come from the insect literature. In many respects the insect literature is
ahead of that on mammals. This may be because of the short generation time
of most insects, the low cost of maintenance, and the enormous diversity of
biological material. Nevertheless, there may be some areas in which en-
tomologists might benefit from considering the mammalian literature. I
shall explore some parallels between the reproductive behavior of rodents
and insects, some examples of cross fertilization, and some possibilities for
future cross fertilization.

*Donald A. Dewsbury is Professor of Psychology at the University of Florida. He received
his Ph.D. from the University of Michigan in 1965 and is a Past-Pres'dent of the Animal
Behavior Society His research is focused on the evolution of mammalian reproductive be-
havior. Address: Department of Psychology, University of Florida, Gainesville, Florida, 32611.

80 Florida Entomologist 68(1) March, 1985

PATTERNS OF COPULATORY BEHAVIOR
We begin with the matters of description and classification. I have re-
viewed the mammalian literature and proposed that mammalian patterns
can be classified according to four criteria: locking, thrusting, multiple
intromissions and multiple ejaculations (Dewsbury 1972). I know of no
similar effort for insects but I believe that such a system might usefully be
employed.
A lock is a mechanical tie between the male and female genitalia and is
found in such mammalian species as dogs, short-tailed shrews, golden mice,
and grasshopper mice (Dewsbury 1972). Analogous genital locking is found
in insects as well. Most notable are the locks of the familiar Florida love
bug that last an average of 56 h (Thornhill 1976).
Whereas many mammalian species, such as virtually all primates, display
intravaginal thrusting, many others cease thrusting when insertion is
achieved. Descriptions of such post-insertion thrusting appear less common
for insects but thrusting does occur in various species such as giant water
bugs (Smith 1980), damselflies (Waage 1979), and crickets (Loher and
Rence 1978).
In many species of rodents and primates there is a pattern of multiple
intromissions prior to ejaculation. Typically, the male repeatedly mounts
the female, gains insertion, and dismounts without ejaculating. Such mul-
tiple intromissions are prerequisite to ejaculation. I know of no comparable
pattern in insects.
Finally, whereas the copulatory activity of a male-female pair in some
mammalian species is terminated with the occurrence of the first ejaculation,
in other species pairs continue to copulate for several ejaculations. Repeated
copulations are common in some insect species, such as giant water bugs
(Smith 1980). The most dramatic example of a single ejaculation species is
the honey bee, in which males make a "suicidal" donation of the genitalia
as a "plug" and die soon thereafter (Michener 1974).
With such a system we can classify copulatory patterns and search for
evolutionary trends. Similar systems could be developed for a variety of
taxa.
A notable characteristic of insect copulatory patterns is the ability to
alter copulation duration. For example, in the southern green stink bug
copulation ranges from 5 minutes to 14 days (McLain 1980). Mammalian
species appear much less variable in this regard. Perhaps copulation dura-
tion is more "hard wired" in mammals than in some insects!

ANATOMY
Taxonomists of both insects and mammals have used the structure of
male genitalia as important characters in classification. As noted by Lloyd
(1979) such characters may be used to predict behavioral variability.
Fortunately, Emmet Hooper and his associates (e.g., Hooper & Musser
1964) did a thorough job of describing penile anatomy in the rodents of the
superfamily Muroidea. A decade ago we noted that variations in penile
anatomy were correlated with behavioral variability. In species in which
males either lock or thrust (and which have a "simple" glans penis) the
glans is thicker, relative to length, than in other species (Dewsbury 1974,

Insect Behavioral Ecology-'84 Dewsbury

1975). Further, the spines that line the glans surface are enlarged in lock-
ing species. In the decade since this proposal we have studied additional
species and successfully predicted other copulatory patterns from penile
anatomy. We also noted that males that lock possess a reduced complement
of accessory glands in the reproductive tract.
The males of most rodent species deposit a copulatory plug with each
ejaculate (Hartung and Dewsbury 1978, Baumgardner et al. 1982). Inter-
estingly, locking species, with their reduced reproductive tracts, deposit no
such plugs. Males of nonlocking species also have penile spines. We have
proposed that they, together with the multiple intromission pattern de-
scribed earlier, function in removing plugs and sperm-sometimes those of
other males (Dewsbury 1981a).
In conducting these analyses we marched in parallel with developments
on insects. In insects, coupling is mediated by a bewildering array of
mandibles, genital claspers, antennae, and modified legs (Wing et al. 1983).
In some butterflies a product of the male accessory glands appears to cement
the pair together (Leopold 1976). In Lytta nuttalli specialized dorsal and
ventral penile spines catch onto folds in the vaginal wall to maintain
coupling (Gerber et al. 1972).
Males of many insect species deposit a copulatory plug. Often these are
the remnants of spermatophores, as in Pteroptyx fireflies (Wing et al. 1983).
The reproductive tracts of some insect species are simplified in ways parallel
to those of rodents. However, whereas in rodents it is species with long
copulations (i.e., locks) that have simplified tracts, in insects it appears to
be those with brief copulations. This may be because, in insects, long copula-
tions are associated with the passage of complex accessory gland secretions
(Gerber et al. 1971, Wing 1984).
Male insects too are adapted to remove plugs and sperm from the female
tract. Most dramatic are the specialized adaptions for sperm removal of the
damselfly penis described by Waage (1979, 1983). Lloyd (1979) wrote of a
"veritable Swiss Army knife of gadgetry" in insect anatomy designed to
function in such a context.
Clearly, in both insects and rodents, reproductive behavior and anatomy
have evolved interactively. The adaptations of rodents and insects often
appear parallel.

SPERM COMPETITION
Among the very few papers most influential to the development of our
research program was G. A. Parker's (1970) "Sperm Competition and its
Evolutionary Consequences in the Insects." With that paper, Parker both
provided a basis for interpreting many curious patterns of insect reproduc-
tive behavior and stimulated students of other taxa to study analagous
phenomena. Parker defined sperm competition as "the competition within a
single female between the sperm from two or more males for the fertilization
of ova" (p. 527). The species-typical pattern of sperm competition is of
great importance to the evolution of mating strategies. Females may
manipulate sperm and males must compete within rules set by females
(Lloyd 1979).
To study sperm competition, one needs to determine the paternity of the
offspring resulting from an episode of copulatory activity. In insects this

82 Florida Entomologist 68(1) March, 1985

has generally been done either with marker genes or by sterilizing one of
two males so that the eggs that do not hatch are assumed to be sired by
him. One must control for the order of mating by the two males, the timing
of the matings, and the differential fertilizing capacity of the sperm from
the males of the two genotypes. The most common result of insect studies
is that the last male to copulate enjoys a differential advantage. However,
there are many species with a first-male advantage, such as Culicoides
melleus (Linley 1975), the parasitic wasp Nansonia vitripennis (Holmes
1974), and the southern green stink bug (McLain 1980). Gwynne (1984)
interprets this variability in relation to the pattern of non-promiscuous
mating efforts and/or parental effort characteristic of the males of each
species. Similarly, Smith (1980) noted that a last male advantage is to be
expected in species, such as giant water bugs, in which there is appreciable
paternal investment. If males are to make a large paternal investment, it is
critical that it be for their own offspring.
Although Parker predicted minimal sperm competition in mammals,
there is good evidence of multiple-male copulations by female rodents in the
field and we were stimulated to investigate the phenomenon (Dewsbury
1984). We have used marker geses affecting both coat color and transferring,
as assessed with electrophoresis. In one series of studies, females mated with
each of two males for an equal number of ejaculations with the timing and
order of mating controlled. In several studies we have found no order effects
in either deer mice or laboratory rats. It did not matter whether a male was
the first or last to ejaculate. Clearly, plugs do not prevent subsequent in-
seminations. By contrast, Levine (1967) found a first-male advantage in
house mice. There is some indication of a last-male advantage in golden
hamsters, although caution must be used in interpreting these results
(Oglesby et al., 1981). We found a last-male advantage in prairie voles
(Dewsbury and Baumgardner 1981). Prairie voles appear to display sub-
stanial male paternal investment and appear sometimes monogamous in the
field (Getz and Carter 1980). Whether this correlation between sperm com-
petition pattern and male investment will prove general must be determined
by future data.
The pattern of sperm competition is one of the most important char-
acteristics of the reproductive system of any species. Yet it is known for
too few species. In their recent book, Thornhill and Alcock (1983) repeatedly
used phrases like "Suppose we assume a 'last-male-to-mate advantage'," (p.
249), "If sperm precedence occurs" (p. 261), and "We shall assume that it
occurs" (p. 334) in interpreting insect mating patterns. It is critical that
more basic studies of sperm competition be conducted in insects.
More research on the dynamics of sperm competition would also be of
interest. In rodents, we have found that the relative number of ejaculates
deposited by two males is critical in determining litter composition. Imposi-
tion of a two-hour delay between males was found to have little effect. How-
ever, it appears that a male deer mouse can essentially cancel another male's
ejaculate by mating with the same female within one minute-presumably
by disrupting sperm transport. Detailed studies of such dynamics can pro-
vide a more solid basis for understanding the pressures affecting the evolu-
tion of mating systems.

Insect Behavioral Ecology-'84 Dewsbury 83

FEMALE CHOICE
Charles Darwin (1871) distinguished between natural selection and
sexual selection, selection for traits that increase an individual's success in
getting mates. Darwin proposed two components: intrasexual selection, gen-
erally male-male competition, and female choice. His logic was summarized
by Thornhill (1980a) and has generally been accepted by most biologists.
In contemporary terms, it is the difference in parental investment (Trivers
1972) that is responsible for the pattern of "ardent" males and "coy"
females. As reviewed by Thornhill, male-male competition has often been
studied, although I might add that the full consequences have not always
been fully documented (Dewsbury 1982a). Female choice has been more
difficult to demonstrate. Thornhill's concerns related to 1) the nature of
the evidence of female choice, 2) the view that those traits that may be
preferred appear unrelated to fitness, and 3) that choice may not be
heritable.
There is much evidence suggestive of female choice in insects. For
example, Thornhill (1980b) showed that female hangingflies appear to
prefer males with prey of an optimal size. The benefit to the female is well
documented. Borgia (1981) demonstrated choice by female dung flies, for
large males and related it to 1) reduced harm in struggles, 2) more rapid
copulation and oviposition, and 3) ability to escape danger during copula-
tion. Partridge (1980) showed that female fruitflies permitted to choose
among males produced offspring that were more fit in tests of intraspecific
competition than did females not permitted to choose. Mate choice in the
two-spot ladybird appears heritable (Majerus et al. 1982).
The methodological problems outlined by Thornhill (1980a) are im-
portant. If preference is tested in a naturalistic situation one cannot be
certain that association is due to true female choice rather than male
coercion or some other factor. If choice is tested in artificial situations
greater control is possible, but the very artificiality is open to criticism.
Our approach is one of convergent methodologies. We are trying to study
mate choice in a variety of situations in the hope that consistent findings
will be generated. In a semi-natural enclosure, for example, female deer
mice approach dominant males more frequently than subordinate males
(Dewsbury 1981b). However, this could be because dominant males tend to
be quite active and accessible, whereas subordinates often appear to be
hiding. We are currently studying female choice in a small cage with the
dominant and subordinate males anesthetized. Should the results of both
situations be consistent, we believe we will have a strong case for female
choice.
We have also studied female choice in a test chamber in which two males
are tethered at opposite ends. Female prairie voles, but not female montane
voles, prefer to mate with a male with which they have copulated previously
rather than a novel male. The experience of copulation appears to "stamp
in" a preference in the monogamous species, prairie voles, but not the non-
monogamous species, montane voles. An analagous effect of mating ex-
perience on female choice has been found in Drosophila (Pruzan 1976,
O'Hara et al. 1976).

84 Florida Entomologist 68(1) March, 1985

MALE CAPACITY AND MALE CHOICE
Much biological thought appears to indicate that in promiscuous species
the cost of sperm is trivial and male choice should be nonexistent (e.g.,
Bateman 1948; Dawkins 1976). We have argued that the correct unit for
consideration is the ejaculate, and that males are limited in their capacity
to produce ejaculates (Dewsbury 1982b). This limitation suggests the ex-
istence of some degree of male choice even in promiscuous species.
Rutowski (1982a) drew similar conclusions from his studies of moths
and butterflies: "Because males are limited in their ability to make these
secretions they also should be selective in whom they court and inseminate"
(p. 78). There are numerous demonstrations in insects of the limited ability
of males to produce ejaclates; examples include oriental fruit moths (George
and Howard 1968), mosquitoes (Jones 1973), fruitflies (Markow et al.
1978), and spruce budworms (Outram 1971). There is also evidence of male
selectivity in mate choice. For example, in the white checkered butterfly,
males selectively court young and large females and this appears adaptive
(Rutowski 1982b). Male Mormon crickets, which produce large spermato-
phores, reject many potential mates, preferring more fecund females
(Gwynne 1981).
We reasoned that male voles ought to mate preferentially with unmated
rather than mated females. By so doing, they could avoid the certain con-
sequences of sperm competition. We studied the preference of male voles
for tethered mated versus unmated females. Male prairie voles spent more
time with and copulated more with unmated rather than mated females.
Because such differences could be due in part to active female resistance by
mated females we repeated the study with the two females anesthetized
and obtained similar results. Although male prairie voles preferred to mate
with unmated females, there was no such preference in montane voles. In
another study we permitted one male vole to copulate with one, two, or four
receptive females. As might be expected, prairie vole males tended to con-
centrate their copulations on a smaller number of females, whereas montane
voles mated less selectively (Fuentes and Dewsbury 1984).
Recency of mating is also a factor in male choice in insects. Male Droso-
phila melanogaster preferentially court virgin rather than inseminated fe-
males, even when the females are decapitated and do not extrude their
genitalia (Cook and Cook 1975), Virgin females emit pheromones that
stimulate males to court; males produce a pheromone that inhibits court-
ship (Tompkins and Hall 1981a, 1981b). Male flour beetles prefer virgin to
fertilized females (Graur and Wood 1982).
The large investments by male insects in spermatophores are of nutri-
tional benefits to the females (e.g., Boggs 1981, Boggs & Gilbert 1979). In
rodents sperm not involved in fertilization may function in facilitating
embryo development (Chaykin and Watson 1983, Watson et al. 1983).
Male newts that have mated repeatedly and have a reduced capacity to
produce spermatophores also decrease the intensity of their courtship dis-
plays and thus might be termed "honest salesmen" (Halliday and Houston
1978). Rutowski (1979) argued for a similar phenomenon in the checkered
white butterfly. Although mosquitoes appear "honest" (Jones 1973) (and
may not even resume courting when supplies have recuperated), species such

Insect Behavioral Ecology-'84 Dewsbury 85

as coddling moths (Howell et al. 1978) and spruce budworms (Outram 1971)
appear less honest.
Although we found that the sperm counts of male deer mice in successive
ejaculates within a session became progressively lower, we could not get
males to deliver numbers of ejaculates in a range where the decreased
sperm counts would have functional consequences for pregnancy initiation
or sperm competition. We termed male deer mice "honest ejaculators"
(Dewsbury & Sawrey 1984).

CONCLUSION

Our quest is for principles of generality. In this quest we should employ
whatever information appears useful, whatever the source. Our research
program on rodents has profited from some knowledge of research on in-
sects. There are some striking similarities and some important differences in
the patterns of copulatory behavior, anatomy, sperm completion, female
choice, and male capacity and choice in insects and rodents. My hope is that
students of insect behavior might learn something of interest from our
studies of rodents. If Henry Walter Bates was correct that the laws of
nature must be the same for all beings then the applicability of conclusions
from rodents, like those from insects, may be quite broad.

ACKNOWLEDGMENTS
This research was supported by grant BNS-8200689 from the National
Science Foundation.

LITERATURE CITED
BATEMAN, A. J. 1948. Intra-sexual selection in Drosophila. Heredity 2:
349-68.
BATES, H. W. 1910. The naturalist on the River Amazons. John Murray,
London.
BAUMGARDNER, D. J., T. G. HARTUNG, D. K. SAWREY, D. G. WEBSTER, AND
D. A. DEWSBURY. 1982. Muroid copulatory plugs and female re-
productive tracts: A comparative investigation. J. Mammal. 63:
110-17.
BOGGS, C. L. 1981. Selection pressures affecting male nutrient investment at
mating in Heliconiine butterflies. Evolution 35: 931-40.
BOGGS, C. L., AND L. E. GILBERT. 1979. Male contribution to egg production
in butterflies: Evidence for transfer of nutrients at mating. Science
206: 83-84.
BORGIA, G. 1981. Mate selection in the fly Scatophaga stercoraria: Female
choice in a male-controlled system. Anim. Behav. 29: 71-80.
CHAYKIN, S., AND J. G. WATSON. 1983. Reproduction in mice: Spermatozoa
as factors in the development and implantation of embryos. Gamete
Res. 7: 63-74.
COOK, R., AND A. COOK. 1975. The attractiveness to males of female
Drosophila melanogaster: Effects of mating, age and diet. Anim.
Behav. 23: 521-26.
DARWIN, C. 1871. The descent of man and selection in relation to sex. John
Murray, London.
DAWKINS, R. 1976. The selfish gene. Oxford University Press, Oxford.
DEWSBURY, D. A. 1972. Patterns of copulatory behavior in male mammals.

86 Florida Entomologist 68(1) March, 1985

Quart. Rev. Biol. 47: 1-33.
1974. Copulatory behavior of white-throated wood rats (Neotoma
albigula) and golden mice (Ochrotomys nuttalli). Anim. Behav. 22:
601-10.
1975. Diversity and adaptation in rodent copulatory behavior.
Science 190: 947-54.
1981a. On the function of the multiple-intromission, multiple-
ejaculation copulatory patterns of rodents. Bull. Psychonomic Soc.
18: 221-23.
1981b. Social dominance, copulatory behavior, and differential re-
production in deer mice (Peromyscus maniculatus). J. Comp. Physiol.
Psychol. 95: 880-95.
1982a. Dominance rank, copulatory behavior, and differential re-
production. Quart. Rev. Biol. 57: 135-59.
- 1982b. Ejaculate cost and male choice. Am. Nat. 119: 601-10.
1984. Sperm competition in muroid rodents. Pages 547-571 in R. L.
Smith ed. Sperm competition and the evolution of animal mating
systems. Academic Press, New York.
AND D. L. BAUMGARDNER. 1981. Studies of sperm competition in
two species of muroid rodents. Behav. Ecol. Sociobiol. 9: 121-33.
---, AND D. K. SAWREY. 1984. Male capacity as related to sperm pro-
duction, pregnancy initiation, and sperm competition in deer mice
(Peromyscus maniculatus). Behav. Ecol. Sociobiol. 16: 37-47.
FUENTES. 9. M., AND D. A. DEWSBURY. 1984. Copulatory behavior of voles
(Microtus montanus and M. ochrogaster) in multiple-female test
situations. J. Comp. Psychol. 98: 45-53.
GEORGE, J. A., AND M. G. HOWARD. 1968. Insemination without spermato-
phores in the oriental fruit moth, Grapholitha molesta (Lepidoptera:
Tortricidae). Canadian Ent. 100: 190-92.
GERBER, G. H., N. S. CHURCH, AND J. G. REMPEL. 1971. The structure,
formation, histochemistry, fate, and functions of the spermatophore
of Lytta nuttalli Say (Coleoptera: Meloidae). Canadian J. Zool. 49:
1595-1610.
-- AND 1972. The anatomy, histology, and physiology
of the reproductive systems of Lytta nuttalli Say (Coleoptera:
Meloidae). II. The abdomen and external genitalia. Canadian J. Zool.
50: 649-60.
GETZ, L. L., AND C. S. CARTER. 1980. Social organization in Microtus
ochrogaster populations. Biologist 62: 56-69.
GRAUR, D., AND D. WOOL. 1982. Dynamics and genetics of mating behavior
in Tribolium castaneum (Coleoptera: Tenebrionidae). Behav. Genet.
12: 116-79.
GWYNNE, D. T. 1981. Sexual difference theory: Mormon crickets show role
reversal in mate choice. Science 213: 779-80.
GWYNNE, D. T. 1984. Male mating effort, confidence of paternity and insect
sperm competition. Pages 000-000 in R. L. Smith ed. Sperm competi-
tion and the evolution of animal mating systems. Academic Press,
New York, in press.
HALLIDAY, T., AND A. HOUSTON. 1978. The newt as an honest salesman.
Anim. Behav. 26: 1273-74.
HARTUNG, T. G., AND D. A. DEWSBURY. 1978. A comparative analysis of
copulatory plugs in muroid rodents and their relationship to copula-
tory behavior. J. Mammal. 59: 717-23.
HOLMES. H. B. 1974. Patterns of sperm competition in Nasonia vitripennis.
Canadian J. Genet. Cytol. 16: 789-95.
HOOPER, E. T., AND G. G. MUSSER. 1964. The glans penis in neotropical

Insect Behavioral Ecology-'84 Dewsbury 87

cricetines (Family Muridae) with comments on classification of
muroid rodents. Misc. Publ. Mus. Zool. Univ. Michigan 123: 1-57.
HOWELL, J. F., R. B. HUTT, AND W. B. HILL. 1978. Codling moth: Mating
behavior in the laboratory. Ann. Ent. Soc. America 71: 891-95.
JONES, J. C. 1973. A study on the fecundity of male Aedes aegypti. J. Insect
Physiol. 19: 435-39.
LEOPOLD, R. A. 1976. The role of male accessory glands in insect reproduc-
tion. Ann. Rev. Ent. 21: 199-221.
LEVINE, L. 1967. Sexual selection in mice IV. Experimental demonstration
of selective fertilization. Am. Nat. 101: 289-94.
LINLEY, J. R. 1975. Sperm supply and its utilization in doubly inseminated
flies, Culicoides melleus. J. Insect Physiol. 21: 1785-88.
LLOYD, J. E. 1979. Mating behavior and natural selection. Florida Ent. 62:
19-34.
LOHER, W., AND B. RENCE. 1978. The mating behavior of Teleogryllus
commodus (Walker) and its central and peripheral control. Z. Tier-
psychol. 46: 225-59.
MAJERUS, M. E. N., P. O'DONALD, AND J. WEIR. 1982. Female mating
preference is genetic. Nature 300: 521-23.
MARKOW, T. A., M. QUAID, AND C. KERR. 1978. Male mating experience and
competitive courtship success in Drosophila melanogaster. Nature 276:
821-22.
MCLAIN, D. K. 1980. Female choice and the adaptive significance of pro-
longed copulation in Nezara viridula (Hemiptera: Pentatomidae).
Psyche 87: 325-36.
MICHENER, C. D. 1974. The social behavior of bees. Harvard University
Press, Cambridge, Massachusetts.
OGLESBY, J. M., D. L. LANIER, AND D. A. DEWSBURY. 1981. The role of pro-
longed copulatory behavior in facilitating reproductive success in
male Syrian golden hamsters (Mesocricetus auratus) in a competitive
mating situation. Behav. Ecol. Sociobiol. 8: 47-54.
O'HARA, E., A. PRUZAN, AND L. EHRMAN. 1976. Ethological isolation and
mating experience in Drosophila paulistorum. Proc. Nat. Acad. Sci.
USA. 73: 975-76.
OUTRAM, I. 1971. Aspects of mating in the spruce budworm, Choristoneura
fumiferana (Leoidoptera: Tortricidae). Canadian Ent. 103: 1121-28.
PARKER, G. A. 1970. Sperm competition and its evolutionary consequences
in the insects. Biol. Rev. 45: 525-67.
PARTRIDGE, L. 1980. Mate choice increases a component of offspring fitness
in fruit flies. Nature 283: 290-91.
PRUZAN, A. 1976. Effects of age, rearing and mating experiences on fre-
quency dependent selection in Drosophila pseudoobscura. Evolution
30: 130-45.
RUTOWSKI, R. L. 1979. The butterfly as an honest salesman. Anim. Behav.
27: 1269-70.
-- 1982a. Mate choice and lepidopteran mating behavior. Florida Ent.
65: 72-82.
.1982b. Epigamic selection by males as evidenced by courtship
partner preferences in the checkered white butterfly (Pieris proto-
dice). Anim. Behav. 30: 108-12.
SMITH, R. L. 1980. Evolution of exclusive postcopulatory paternal care in
the insects. Florida Ent. 63: 65-78.
THORNHILL, R. 1976. Reproductive behavior of the lovebug, Plecia nearctica
(Diptera: Bibionidae). Ann. Ent. Soc. America 69: 843-47.
1980a. Competitive, charming males and choosy females: Was
Darwin correct? Florida Ent. 63: 5-30.