Title: Florida Entomologist
ALL VOLUMES CITATION DOWNLOADS THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098813/00070
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 1990
Copyright Date: 1917
 Subjects
Subject: Florida Entomological Society
Entomology -- Periodicals
Insects -- Florida
Insects -- Florida -- Periodicals
Insects -- Periodicals
 Notes
General Note: Eigenfactor: Florida Entomologist: http://www.bioone.org/doi/full/10.1653/024.092.0401
 Record Information
Bibliographic ID: UF00098813
Volume ID: VID00070
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Downloads
Full Text






(ISSN 0015-4040)


FLORIDA ENTOMOLOGIST

(An International Journal for the Americas)

Volume 73, No. 1 March, 1990

TABLE OF CONTENTS


Announcement 73rd Annual Meeting ....................................... ............. i
Announcement Forum Section ............................................................ i

SYMPOSIUM: INSECT BEHAVIORAL ECOLOGY-'89

FRANK, J. H., AND E. D. McCoY-Introduction to Attack and Defense: Be-
havioral Ecology of Predators and Their Prey. Endemics and Epidemics
of Shibboleths and Other Things Causing Chaos ................................. 1
LINLEY, J. R.-The Predatory Behavior of Toxorhynchites amboinensis and Tx.
brevipalpis Larvae (Diptera: Culicidae) in Response to Subsurface Prey 9
LLOYD, J. E.-Firefly Semiosystematics and Predation: A History .............. 51
PEARSON, D. L.-The Evolution of Multi Anti-Predator Characteristics as Illus-
trated by Tiger Beetles (Coleoptera: Cicindelidae) .............................. 67
WITZ, B. W.-Antipredator Mechanisms in Arthropods: A Twenty Year Litera-
ture Survey .................................................... ......................... 71
ALLEN, J. C.-Chao and Phase-Locking in Predator-Prey Models in Relation to
the Functional Response ............................................ .............. 100


Research Reports
DIAWARA, M. M., B. R. WISEMAN, D. J. ISENHOUR, AND G. R. LOVELL-Re-
sistance to Fall Armyworm in Converted Sorghums ........................... 111
STEIN, M. B., H. G. THORVILSON, AND J. W. JOHNSON-Seasonal Changes in
Bait Preference by Red Imported Fire Ant, Solenopsis invicta (Hymenop-
tera: Form icidae) ........................................................................... 117
SIVINSKI J.-Colored Spherical Traps for Capture of Caribbean Fruit Fly, Anas-
trepha suspense .............................................................................. 123
SCHROEDER, W. J.-Water Absorbent Starch Polymer: Survival Aid to
Nematodes for Control of Diaprepes abbreviatus (Coleoptera: Cur-
culionidae) in Citrus ....................................................................... 129
VAN DRIESCHE, R. G., A. C. BELLOTTI, J. CASTILLO, AND C. J. HERRERA-
Estimating Total Losses from Parasitoids for a Field Population of a Con-
tinuously Breeding Insect, Cassava Mealybug, Phenacoccus herreni,
(Homoptera: Pentatomidae) in Columbia, S. A. ................................ 133
MCLAUGHLIN, R. E.-Predation Rate of Larval Corethrella brakeleyi (Diptera:
Chaoboridae) on Mosquito Larvae .................................................... 143
RossI, A. M., AND D. R. STRONG, JR.-Natural History of the Leafhopper Car-
neocephala floridana (Homoptera: Cicadellidae) in a North Florida Salt
M arsh ........................................................................................... 147
BROWN, J. W.-New Species and First U. S. Record ofAuratonota (Lepidoptera:
Tortricidae .................................................................................... 153
WIRTH, W. W., AND J. L. CASTNER-New Neotropical Species of "Stick-tick"
(Diptera: Ceratopogonidae) from Katydids ........................................ 157


Continued on Back Cover

Published by The Florida Entomological Society










FLORIDA ENTOMOLOGICAL SOCIETY
OFFICERS FOR 1989-90
President .................................................................................. J. E Eger
President-Elect ..................... .. ...... ................... J. F. Price
Vice-President ................................... ..... ................... J. L. Knapp
Secretary .................. ... .. .. .............................. J. A. Coffelt
Treasurer .................... .. .......... ..................... A. C. Knapp
Other Members of the Executive Committee
R. S. Patterson J. E. Pefia F. D Bennett
M. Camara R. Coler
PUBLICATIONS COMMITTEE
J. R. McLaughlin, USDA/ARS, Gainesville, FL ...................................... Editor
Associate Editors
Agricultural, Extension, & Regulatory Entomology
Ronald H. Cherry-Everglades Research & Education Center, Belle Glade, FL
Michael G. Waldvogel-North Carolina State University, Raleigh, NC
Apiculture
Stephen B. Bambara-North Carolina State University, Releigh, NC
Biological Control & Pathology
Ronald M. Weseloh-Connecticut Agricultural Experiment Sta., New Haven, CT
Book Reviews
J. Howard Frank-University of Florida, Gainesville
Chemical Ecology, Physiology, Biochemistry
Louis B. Bjostad-Colorado State University, Fort Collins, CO
Ecology & Behavior
John H. Brower-Stored Product Insects Research Laboratory, Savannah GA
Theodore E. Burk-Dept. of Biology, Creighton University, Omaha, NE
Forum & Symposia
Carl S. Barfield-University of Florida, Gainesville
Genetics & Molecular Biology
Sudhir K. Narang-Bioscience Research Laboratory, Fargo, ND
Medical & Veterinary Entomology
Arshad Ali-Central Florida Research & Education Center, Sanford, FL
Resumen
Omelio Sosa, Jr.-USDA Sugar Cane Laboratory, Canal Point, FL
Systematics, Morphology, and Evolution
Michael D. Hubbard-Florida A&M University, Tallahassee
Howard V. Weems, Jr.-Florida State Collection of Arthropods, Gainesville
Willis W. Wirth-Florida State Collection of Arthropods
Business M manager ....................................................................... A. C. Knapp
FLORIDA ENTOMOLOGIST is issued quarterly-March, June, September, and De-
cember. Subscription price to non-members is $30 per year in advance, $7.50 per copy;
institutional rate is $50 per year. Membership in the Florida Entomological Society,
including subscription to Florida Entomologist, is $25 per year for regular membership
and $10 per year for students.
Inquiries regarding membership and subscriptions should be addressed to the Busi-
ness 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 in Winter Haven, FL.
Manuscripts from all areas of the discipline of entomology are accepted for consider-
ation. At least one author must be a member of the Florida Entomological Society.
Please consult "Instructions to Authors" on the inside back cover.
This issue mailed March 30, 1990













ANNOUNCEMENT 73RD ANNUAL MEETING
FLORIDA ENTOMOLOGICAL SOCIETY

The 73rd annual meeting of the Florida Entomological Society will be held August
5-9, 1990 at the Camino Real Hotel in Cancun, Mexico. Travel and hotel arrangements
are being handled through Holbrook Travel, 3540 N.W. 13th Street, Gainesville, FL
32609 (Phone 1-800-345-7111), Attn: Ms. Joyce Rickard. Registration forms and addi-
tional information will be mailed to members in the Newsletter.

SUBMISSION OF PAPERS

The deadline for submission of papers and posters for the 73rd annual meeting of
the Florida Entomological Society will be May 15, 1990. The meeting format will contain
seven symposia so there will be concurrent sessions. Submitted papers will be eight
minutes allocated for the oral presentation with two minutes for discussion. A separate
Poster Exhibit Session is planned. There will be student paper and poster sessions with
awards as in previous years. Students participating in these judged sessions must be
members of the Society and registered at the meeting.

For additional information contact:
Joseph L. Knapp, Chairman
Program Committee. FES
University of Florida
Citrus Research and Education Center
700 Experiment Station Road
Lake Alfred, FL 33850
813-956-1151)




FORUM
A New Type of Article for our Authors and Subscribers

We are proud to announce that scientists may submit articles for publication in a
FORUM section of Florida Entomologist. FORUM articles (1-2 per issue) will appear
at the beginning of each issue in a section marked FORUM. If available, the first
FORUM articles will appear in the June 1990 issue.
Articles for the FORUM section must follow the general style guidelines for all
other articles submitted to Florida Entomologist. FORUM articles must be of high
scientific quality, demonstrate acceptable experimental design and analysis, and cite
appropriate sources to support findings. FORUM articles will include "cutting edge"
science, scientifically meritorious but controversial subjects, new methodologies (de-
signed and tested), experimentally-based designs and tests of pedagogical methods, and
documented challenges to existing entomological techniques, philosophies or experimen-
tal paradigms.
Submitted articles should include "Submitted to Florida Entomologist: FORUM"
on the title page. Three or more peer reviews will be acquired by the Associate Editor
for FORUM publications.
We feel the addition of a FORUM section will expand the scope of Florida En-
tomologist and allow readers and publishing scientists an additional creative outlet that
will complement our symposia, research articles, and notes.


















C)
U I4





aW cP W
c o0 -
ro o E




DO *C





4-
co. j

















4o a




o o
a I-
I ^ ^
M~a ^ ^2












I~ I, f ^* e^_____










Insect Behavioral Ecology-'89 Frank & McCoy


Introduction To
ATTACK AND DEFENSE: BEHAVIORAL ECOLOGY OF
PREDATORS AND THEIR PREY

ENDEMICS AND EPIDEMICS OF SHIBBOLETHS
AND OTHER THINGS CAUSING CHAOS

J. H. FRANK
Entomology and Nematology Department, 3103 McCarty Hall
University of Florida, Gainesville, FL 32611

E. D. McCoY
Department of Biology, University of South Florida
Tampa, FL 33620

"When I use a word," Humpty Dumpty said, in rather a scornful
tone, "it means just what I choose it to mean-neither more nor
less."

"The question is," said Alice, "whether you can make words mean
so many different things."

"The question is," said Humpty Dumpty, "which is to be master-
that's all."
Lewis Carroll (1872)

Classical biological control is a set of activities in which non-native predators,
parasites, parasitoids, competitors or pathogens, are introduced into an area for the
control of pests, especially immigrant pests. The language of classical biological control
is that of ecology, which uses words of classical (Greek and Latin) origin because these
can be assigned precise meanings. The precision of the meanings allows for conciseness
and accuracy of expression, or such is the intention. Unfortunately, the meanings of a
group of words describing the origins of predators etc. and pests alike are anything but
precise. Their definitions are chaotic in several dictionaries. Therefore we take an
etymological voyage to the origins of the words precinctive, autochthonous, indigen-
ous, epidemic, endemic, adventive, introduced, and immigrant to see how they have
been used, and we illustrate this voyage by citations from the literature. Then we show
the relationships of the five words that we deem most useful in describing origins, and
we redefine the words adventive, immigrant, and introduced.

AN ETYMOLOGICAL VOYAGE

Precinctive (from Latin praecinctus, present participle ofpraecingere, to gird, encircle)

David Sharp rose to eminence as a describer of insects, many of them predatory, in
the final quarter of the last century. His works include contributions on the faunas of
Britain, the Amazon basin, New Zealand, Central America, Japan, and Hawaii. In the
last of these works (Sharp 1900: 91), faced with need of a word under which to group
organisms restricted in distribution to Hawaii, he chose and defined the word precinc-
tive: "I use the word precinctive . in the sense of 'confined to the area under
discussion'. .. 'Precinctive forms' means therefore forms that are confined to the area











2 Florida Entomologist 73(1) March, 1990

specified." The word was adopted by Bequaert (1940: 266) and others: "Of a total of 42
recorded species [of Antillean Tabanidae], 33 (or nearly 80%) are precinctive and 26 (or
60 per cent) are restricted to a single island".

Autochthonous (from Greek autochthon = sprung from the land itself)

Browne (1646: 274) "There was never any Autochthon, or man arising from the
earth but Adam." Gardiner (1804) "If the English have this great predilection for au-
toohthonous bread and butter". Anon. (1860) "Most of them [the Red Indians], of
course, believe themselves to be autochthonous; but the Chippewas and one or two
others retain, or till lately retained, the tradition of a migration from over the sea."
Torre-Bueno (1937) autochthonouss, native or aboriginal; used for those species which
are considered to have arisen as a part of the native or aboriginal fauna or flora, as
contrasted with those which are considered to have immigrated from outside regions
([after] Tillyard [1926])." Mackerras (1970: 191) "An autochthonous group is one that
evolved within the country".

Indigenous (from Latin indigenous = native)

Browne (in Pseudoxia epidemic 1646: 325) "and although in many parts thereof it
be confessed there bee at present swarmes of Negroes serving under the Spaniard, yet
were they all transported from Africa, since the discovery of Columbus, and are not
indigenous or proper natives of America." Williams (1809: 85) "GINSENG was formerly
esteemed a plant indigenous only to China and Tartary." Buckle (1857: 118) "Indeed,
of those cruel diseases now existing in Europe, scarcely one is indigenous; and the
worst of them were imported from tropical countries in and after the first century of
the Christian era." Lyell (1875: 419) "The insects of Madeira, the Salvages, and the
Canaries, unlike the birds, exhibit a large proportion of indigenous species".

Epidemic (from Greek epi = on and demos = population, through Late Latin epidemic
and French 6pidemique)

Lodge (1603: B2b) "Popular and epidemic have one and the same signification; that
is to say, a sicknesse common unto all people, or to the most part of them." Malthus
(1803: 330) "The endemick and epidemic diseases in Scotland, fall chiefly, as is usual,
on the poor." Southwood & Comins (1976: 963) "The model is used to describe field-data
on endemic and epidemic populations of aphids and eucalyptus psyllids and to correlate
experience in biological control situations." Epidemic has a long history of use as a
descriptor of populations and is used to describe population explosions, in contrast to
the endemic condition.

Endemic (from Greek en = in and demos = population, through French end4mique)

Larousse (1972) lists 2 French synonyms (chronique, permanent) and 4 antonyms
(cyclique, dpidemique, momentand, and passage) of endemique, and defines the French
word endgmicit4 (endemicity) as the endemic state of a disease. Lodqe (1603: B2) "And
such sicknesses as are these, are called Endemiques, provintiall or regional infirmities,
yet for all that they are not to be accounted pestilentiall or contagious." Hickeringill
(1705: 42) "For which I need not beg credit, since there is no country disease (as at
Virginia and Surinam) endemically raging throughout the Isle". Malthus (1803, above,
under epidemick. Buckle (1857: 118) "For evidence of the extra-European origin of
European diseases, some of which, such as the small-pox, have passed from epidemics











Insect Behavioral Ecology-'89 Frank & McCoy 3

into endemics." Smith (1869: 77) "Hence famines are periodical or endemic in Hindos-
tan." Cameron (1870: 149) "The endemical disorder passing rapidly into epidemical."
Southwood & Comins (1976) "The predator can reduce all populations below the release
point to extinction, so there is no lower equilibrium or endemic level." OAD (1980)
"Yellow fever was endemic in parts of South America. The last American epidemic of
this disease occurred in New Orleans in 1905." Price (1984) "When the effect of enemies
is disrupted or environmental conditions become particularly favorable for reproduction,
the population escapes the stabilizing influence of enemies ... and increases to epidemic
proportions. .. However, shortage of food and disease may lead to massive mortality
and low natality causing the population to crash to ... endemic levels." Church (1989)
"In a country where problems are endemic". Endemic has a long history of use as a
descriptor of populations and is the antonym of epidemic. It implies nothing about
native origin [yellow fever is endemic to parts of South America but is believed to be
indigenous to Africa].

Second and third meanings of endemic

Darwin (1872: 178) "Although in oceanic islands the species are few in number, the
proportion of endemic kinds (i.e. those found nowhere else in the world) is often ex-
tremely large." Dallas (1872: 311) "ENDEMIC.-Peculiar to a given locality". By coin-
cidence, Carroll's "Through the looking-glass. ." also was published in 1872. Mackerras
(1970: 192) "An endemic (or precinctive) group".
However, Lyell (1875) and Darwin (1876), who were contemporaries and colleagues,
used the word endemic elsewhere, without definition. Lyell (1875: 413) "General infer-
ences to be deduced from the endemic, and other species of animals and plants in the
Atlantic Islands." Darwin (1876: 415) "Bees . visit many exotic flowers as readily as
the endemic kinds". These uses of endemic apparently were interpreted by some as
"Of plants or animals; Having their ordinary habitat in a certain country; opposed to
exotic" [= indigenous] (OED 1971), and "belonging or native to a particular region or
country" [= indigenous] (Webster 1986), and Ehrlich & Roughgarden (1987: 619) de-
fined endemic as "Native to a particular region" [= indigenous].
Only two meanings were defined by Allaby (1977), the ecological meaning, "Of pests
or disease-producing species. The normal population level of a species which occurs
continuously in a given area", and a biogeographical meaning, "General. Confined to a
given region and having originated there."

A fourth meaning of endemic

Wallner (1987) "on the other hand, endemic or rare species approximate stable
equilibrium. In this review 'endemic' refers to insects that are either rare or uncom-
monly abundant and therefore seldom, if ever, occur at densities sufficient for them to
be considered pests." The expression endemic species is used, but such species are
so-named according to their population level, regardless of whether they are indigenous
or adventive.

Adventive (from Latin advenire = to arrive)

Bacon (1605: 137) "Upon the first of these, the considerations of the origin of the
soul, whether it be native or adventive". Pemberton (1964: 695) "Since a few hundred
more have been added to the endemic [= precinctive] list we can assume that the
adventive species, including harmful, beneficial and indifferent species number well
over 2,000."











4 Florida Entomologist 73(1) March, 1990

A related word, adventitious, is used in biology, especially botany, to describe struc-
tures occurring in other than their customary position, e.g., the adventitious roots
which arise from branches of some Ficus species.

Immigrant (from Latin immigrantem, present participle ofimmigrare, one who or that
which migrates into a country as a settler)

Belknap (1792: 6) "There is another deviation from the strict letter of the English
dictionaries; which is found extremely convenient in our discourses on population ...
the verb IMMIGRATE and the nouns IMMIGRANT and IMMIGRATION are used
without scruple in some parts of this volume." Kendall (1809: 252) "Immigrant is
perhaps the only new word, of which the circumstances of the United States has [sic]
in any degree demanded the addition to the English language."

Introduced (species or other taxa introduced deliberately by man)
Sailer (1978) used the expressions introduced, exotic and immigrant virtually inter-
changably, but here we restrict the expression introduced, following Zimmerman (1948:
64) "the word introduced should be reserved for those species which have been pur-
posely imported".

A RATIONALIZATION

The words indigenous and autochthonous are synonyms and mean native. Indigen-
ous has been used more widely in biology, so for that reason is preferable to au-
tochthonous. We prefer indigenous to native as a biological term because the latter
has subsidiary meanings in English. The third meaning of endemic is as a synonym of
indigenous, but we can find no excuse for this misconception.
Epidemic and endemic were employed as antonyms almost 400 years ago and they
have retained this sense in epidemiology and ecology to mean outbreak and non-out-
break population levels respectively. The biogeographio meaning of "endemic", appar-
ently first employed or at least popularized by Darwin (1872), is unfortunate. Brodie's
(1856: 26) words are pertinent: "There are epidemics of opinion as well as of disease,
and they prevail at least as much among the well-educated as among the uneducated
classes of society." The opinion of persons concerned with distribution of species seems
to be to continue Darwin's usage despite existence of a more valid alternative (precinc-
tive) and despite prior and continuous usage of endemic as an antonym of epidemic by
persons concerned with populations of organisms. Sharp's (1900) statements are explicit:
"I use the word precinctive in preference to endemic or peculiar-both of which are in
common use-in the sense of 'confined to the area under discussion.' The word endemic
has been objected to on the grounds that its definition does not indicate geographical
restriction, and that it is actually used in medicine to signify constant, but not necessar-
ily exclusive, presence in a locality." We conceive precinctive to be a subclass of indi-
genous, and we prefer not to use endemic in this sense.
Wallner's (1987) use of the word endemic is a logical extension of the epidemiological/
ecological use. We doubt that it will be popularized until Darwin's (1872) sense of en-
demic is replaced widely by precinctive.
It is clear that adventive has the broadest sense among the words used to denote
non-indigenous organisms. Our definitions are: Adventive species are those which have
arrived in a previously-unoccupied area, whether of their own volition or through the
inadvertent or deliberate agency of man. They include the 2 subclasses immigrant and
introduced. Immigrant species are adventive species which arrived without the delib-
erate agency of man, even though they may have been transported accidentally by man.










Insect Behavioral Ecology-'89 Frank & McCoy


The word suggests to us an active movement which is the complement of the passive
movement implied by the word introduced. Introduced species are adventive species
which have been introduced by the deliberate agency of man. Among them are to be
included those introduced for biological control purposes, together with those introduced
for other purposes (e.g., crop plants, farm animals, and ornamental plants). We exclude
those organisms which Sailer (1978) called "accidentally introduced" because we con-
sider them immigrants.
More is known about current and former distributions of Homo sapiens L. than
about any other species, so we can employ these as examples to demonstrate use of the
vocabulary. Man as a species is not indigenous to the Americas, but is adventive and,
more specifically, is an immigrant. At an infra-subspecific level, we recognize several
waves of immigration. Some immigrants (Amerindians [including Chippewas] and Es-
kimos) are considered by sociologists to be indigenous to North America because immi-
gration occurred >10,000 yrs BP, whereas no other group is considered indigenous.
Eskimos populated Arctic areas of Asia and North America, so were not and are not
precinctive to Arctic America. However, Amerindians evolved in and thus became
precinctive to the Americas (it is arguable whether they can still be considered precinc-
tive because a few have migrated to other continents). The time frame of these immigra-
tions spans >10,000 years. Descendants of even the earliest European immigrants still
are considered immigrants.
Ideally, we should apply the same criteria to all species. The criteria are formulated
below as a dichotomous key.
For any species (or other taxon) of organism occurring in a specified area:
1 It achieved its current taxonomic identity elsewhere (it was formerly absent
but is now present) ................................................................... adventive 2
1' It achieved its current taxonomic status here (and has been present virtually
continuously since then) ......................................................... indigenous 3
For an adventive species (or other taxon):
2 It was introduced deliberately by man ........................................ introduced
2' It was not introduced deliberately by man .................................... immigrant
For an indigenous species (or other taxon):
3 It is known from no other area .................................................. precinctive
3' It is known from other areas ........................ indigenous but not precinctive
It is reasonably simple to distinguish between alternatives in couplet 2 (introduced
vs immigrant) and in couplet 3 (precinctive vs indigenous but not precinctive). How-
ever, the parenthetic time frame in couplet 1 adventivee vs indigenous) is uncertain.
The ability to distinguish very recent immigrants is of practical value in economic
entomology, because some of these immigrants are, or are likely to become, pests.
Whitehead & Wheeler (1990) suggested that indigenous species should be distinguished
from immigrant species through records of former absence and current presence of the
putative immigrants in the area of interest (this criterion is placed in parentheses in
couplet 1 of the key above) Most records for invertebrates are associated with presence
or absence of preserved specimens in museums. For the most part such specimens can
at best help us decide whether an organism immigrated or was present continuously
during the last few decades of human history.
There are two avenues to extend the record further into the past. The fossil record,
although scanty, has demonstrated the presence of some invertebrate species in the
distant past though it is better adapted for demonstrating presence rather than absence.
Probable immigration can be inferred from cladistic studies.
To take examples from Florida mosquitoes, Aedes albopictus (Skuse) is an immi-
grant, and a recent one, with its arrival well-documented. Aedes aegypti (L.) likewise
is an immigrant, believed to have originated in Africa, and probably has been present











6 Florida Entomologist 73(1) March, 1990

in Florida for some hundreds of years (at or soon after Spanish settlement), long before
humans began to catalog the mosquito fauna. Toxorhynchites amboinensis (Doleschall)
was introduced into Florida for biological control purposes. All 3 species are adventive.
The two Wyeomyia species whose larvae develop in the leaf axils of bromeliads of the
genus Tillandsia in southern Florida cause more thought. Both occur also in the Greater
Antilles (and one in eastern Mexico), so they are not precinctive to Florida (and neither
are Tillandsia utriculata L. and Tillandsia fasciculata Swartz, the 2 plants providing
principal habitat for their larvae). It is likely that both Wyeomyia were immigrant to
Florida in prehistoric times, and perhaps immigration by their conspecifics from the
Greater Antilles still occurs in hurricane winds. However, their evolutionary biogeog-
raphy is unstudied, so by default we consider them indigenous at least for the present.
We do not know of any mosquito species precinctive to Florida.

THIS SYMPOSIUM

This tenth Behavioral Ecology Symposium is entitled "Attack and Defense: Be-
havioral Ecology of Predators and Their Prey." It continues the theme of the ninth
symposium, which was entitled "Attack and Defense: Behavioral Ecology of Parasites
and Parasitoids and Their Hosts." In our introduction to the ninth symposium (Frank
& McCoy 1989), we asked how the contributions fit in with predictions that have been
made about the immediate future courses of behavioral ecological research in general
(Krebs 1985) and of behavioral ecological research on insects in particular (Burk 1988).
The answer clearly was that the contributions fit in well, and it seems appropriate to
ask the same questions for the contributions to this thematically-similar tenth sym-
posium.
Before we can answer the question, however, we will need to reiterate the predic-
tions. Krebs (1985) predicted five paths of behavioral ecological research: (1) life history
and population dynamics in relation to behavioral ecology, (2) mating systems, (3) para-
sites and sexual selection, (4) learning, and (5) the genetic basis of behavior. Likewise,
Burk (1988) predicted five paths for research specifically on insects: (1) sexual selection,
(2) resource competition among females, (3) learning, (4) orientation and movement,
and (5) communication. The salient difference between the two lists is the emphasis
upon studies of orientation, movement, and communication in insects.
It is interesting to note, then, that four of the five contributions to the tenth sym-
posium deal in some manner with orientation, movement, and/or communication. John
Linley's contribution details the movements employed by species of Toxorhynchites to
effect the capture of subsurface prey. His exhaustive analysis reveals an unimagined
melange of movements necessary for successful prey capture, rivaling in complexity
those documented for many larger, much more visible predators. Jim Lloyd's contribu-
tion suggests that the sometimes puzzling "sexual signals" of fireflies may be hard to
interpret because they have evolved under constraints beyond simple efficacy of attrac-
tion of potential mates. He posits that these signals also reflect strong selection imposed
by predation and, therefore, that they are not solely species-isolating mechanisms.
Multiplicity of selective pressures also is central to Dave Pearson's contribution. He
questions the common assumption that prey have only single anti-predator characters,
and presents six theories, based upon his work with tiger beetles, to explain the evolu-
tion of multiple anti-predator characters. He shows how various types of movements,
orientations, signals, and other mechanisms work in concert to deter predation upon
tiger beetles. Finally, Brian Witz catalogues some of the recent literature on predator-
prey interactions, and places the studies into categories of taxonomy of participants and
type of interaction. Many of the categories of interaction he erects are based upon
directed movements of prey species toward predators, or upon some sort of signalling











Insect Behavioral Ecology-'89 Frank & McCoy


by prey species to predators. He suggests that the frequency with which taxa and types
of interactions are studied often is not the same as their proportional representations
in nature.
Jon Allen's contribution is quite distinct from the other four. He explores the "phase-
locked," "quasiperiodic," and "chaotic" behaviors of predator-prey models in relation to
the functional response, and finds unexpected and complex switching among these be-
haviors. He also finds that some types of functional response produce more complexity
than do others. He notes that his models ignore genetics, arrangements of individuals
in space, and other real complications, but suggests that inclusion of these complicating
factors is not likely to reduce the complexity he has uncovered. Allen discussed some
contributions of genetics and arrangement of individuals in space to parasitoid-host
models in the ninth Behavioral Ecology Symposium.

ENDNOTES

We thank (at the University of Florida) Dale Habeck (Entomology & Nematology
Dept.) and Smith Kirkpatrick (English Dept.) for critical reviews of drafts of this man-
uscript, Sandy Fairchild (Entomology & Nematology Dept.) for bibliographic sugges-
tions, and Tomas Zoebisch (Entomology & Nematology Dept.) for translating abstracts
into Spanish. We also thank several anonymous reviewers of manuscripts of the papers
contributed to this symposium, and David Doerrer (Library, University of West
Florida) for examining a microfilm of The Saturday Review. . This is University of
Florida, Institute of Food & Agricultural Sciences, journal series no. R-00483.

REFERENCES CITED

ALLABY, M. 1977. A dictionary of the environment. Van Nostrand Reinhold, New
York.
ANON. 1860. The deserts of North America [a book review]. Saturday Rev. Politics
Lit. Sci. Art 10: 148-49.
BACON, F. 1605. The two books of Francis Bacon, of the proficience and advancement
of learning, divine and humane. Henrie Tomes; London, [1 +] 45 + 118 [+ 1]
p. [edited by T. Case 1951 with title The advancement of learning and New
Atlantis. Oxford Univ. Press, London, xxiii + 298 p. (see p. 136-137)].
BELKNAP, J. 1792. The history of New-Hampshire. Belknap & Young; Boston, vol.
3, A geographical description of the state; with sketches of the natural history,
productions, improvements, present state of society and manners, laws and gov-
ernment, 480 p.
BEQUAERT, J. 1940. The Tabanidae of the Antilles (Dipt.). Rev. Ent., Rio de Janeiro
11: 253-369.
BRODIE, B. C. 1856. Psychological inquiries: In a series of essays, intended to illus-
trate the mutual relations of the physical organization and the mental faculties.
Longman, Brown, Green & Longman; London, 3rd edn., xii + 276 p.
BROWNE, T. 1646. Pseudoxia epidemic: Or, enquiries into very many received te-
nents, and commonly presumed truths. Edward Dod; London, [18 + ] 386 p.
BUCKLE, H. T. 1857. History of civilization in England. J. W. Parker & Son; London,
vol. 1, xxxi + 854 p.
BURK, T. 1988. Insect behavioral ecology: Some future paths. Annu. Rev. Ent. 33:
319-36.
CAMERON, J. 1870. Phases of thought. Simpkin; London, [4 + ] 208 p.
CARROLL, L. 1872. Through the looking-glass, and what Alice found there. Macmillan;
London, [6 + ] 224 p.
CHURCH, G. J. 1989. Look who's feeling picked on. Time (25 Sept. 1989): 36-37.
DALLAS, W. S. 1872. Glossary. p. 307-322 in C. Darwin. The origin of species ..
[see below].










8 Florida Entomologist 73(1) March, 1990

DARWIN, C. 1872. The origin of species by means of natural selection or the preserva-
tion of favored races in the struggle for life. With additions and corrections from
sixth and last English edition [of 1872, but this US edn. publ. 1927, D. Appleton;
New York, xxvi + 338 p.].
DARWIN, C. 1876. The effects of cross and self fertilisation in the vegetable kingdom.
John Murray; London, viii + 482 p.
EHRLICH, P. R., AND J. ROUGHGARDEN. 1987. The science of ecology. Macmillan,
New York.
FRANK, J. H., AND E. D. McCoY. 1989. Behavioral ecology: From fabulous past to
chaotic future. Florida Ent. 72: 1-6.
GARDINER, J. 1804. Cursory observations on the act for ascertaining the bounties,
and for regulating the exportation of corn. Annu. Rev. Hist. Lit. 3: 306-10.
HICKERINGILL, E. 1705. Jamaica viewed: with all the ports, harbours, and their
several soundings, towns, and settlements thereunto belonging. Together with
the nature of its climate, fruitfulness of the soil, and its suitableness to English
complexions. With several other collateral observations upon the island. B.
Bragg; London, 3rd edn. [8 + ] 44 p. + 1 pl. + 1 map. [the 1st edn. of 1661 may
contain the same words but has not been examined by us].
KENDALL, E. A. 1809. Travels throughout the northern parts of the United States,
in the year 1807 and 1808. I. Riley; New York, vol. 1, vi + 312 p.
KREBS, J. 1985. Sociobiology ten years on. New Sci. 108(1): 40-43.
LAROUSSE. 1972. Grand Larousse de la langue frangaise en six volumes. Librairie
Larousse; Paris, vol. 2, p. 737-1,728.
LODGE, T. 1603. A treatise of the plague: Containing the nature, signes, and accidents
of the same, with the certain and absolute cure of the fevers, botches and
carbuncles that raigne in these times: And above all things most singular exper-
iments and preservatives in the same, gathered by the observation of divers
worthy travailers, and selected out of the writings of the best learned phisitians
in this age. Edward White and N. L.; London, A-L ff [reprinted 1979, Theatrum
Orbis Terrarum; Amsterdam].
LYELL, C. 1875. Principles of geology or the modern changes of the earth and its
inhabitants considered as illustrative of geology. John Murray; London, 12th
edn., vol. 2, xviii + 652 p. + 3 pl.
MACKERRAS, I. M. 1970. Composition and distribution of the fauna. p. 187-203, in
The insects of Australia. Melbourne Univ. Press, Carlton, Victoria.
MALTHUS, T. R. 1803. An essay on the principle of population; Or, a view of its past
and present effects on human happiness; With an inquiry into our prospects
respecting the future removal or mitigation of the evils which it occasions. J.
Johnson; London, xi + 610 p.
OAD. 1980. Oxford American dictionary. Avon Books (for Oxford Univ. Press), New
York.
OED. 1971. The compact edition of the Oxford English Dictionary. Oxford Univ.
Press; Glasgow, [ca. 16,640 p. reproduced micrographically in] 4,116 p.
PEMBERTON, C. E. 1964. Highlights on the history of entomology in Hawaii 1778-
1963. Pacific Ins. 6: 689-729.
PRICE, P. W. 1984. Insect ecology. John Wiley, New York.
SAILER, R. I. 1978. Our immigrant insect fauna. Bull. Ent. Soc. America 24(1): 3-11.
SHARP, D. 1900. Coleoptera. I. Coleoptera Phytophaga. p. 91-116, in D. Sharp (ed.),
Fauna Hawaiiensis, being the land-fauna of the Hawaiian Islands. Cambridge
Univ. Press; Cambridge, vol. 2, 700 [ + 21] + 46 p. + 21 pl.
SMITH, A. 1869. An inquiry into the nature and causes of the wealth of nations.
Oxford; Clarendon Press, Vol. 1, lii ... 423 p. [the book was published first in 1776,
but we have not seen an earlier edition with the footnote cited].
SOUTHWOOD, T. R. E., AND H. N. COMINS. 1976. A synoptic population model. J.
Anim. Ecol. 45: 949-65.
TILLYARD, R. J. 1926. The insects of Australia and New Zealand. Angus &
Robertson; Sydney, xi + 560 p. + 44 pl.
TORRE-BUENO, J. R. DE LA. 1937. A glossary of entomology. Smith's "An explanation











Insect Behavioral Ecology-'89 Linley


of terms used in entomology" completely revised and rewritten. Brooklyn Ent.
Soc., New York, ix + 330 p. + ix pl.
WALLNER, W. E. 1987. Factors affecting insect population dynamics: Differences
between outbreak and non-outbreak species. Annu. Rev. Ent. 32: 317-40.
WEBSTER. 1986. Webster's Ninth New Collegiate Dictionary. Merriam-Webster,
Springfield, MA.
WHITEHEAD, D. R., AND A. G. WHEELER. 1990. What is an immigrant arthropod?
Ann. Ent. Soc. America 83: 10-14.
WILLIAMS, S. 1809. The natural and civil history of Vermont. Samuel Mills, Bur-
lington, vol. 1, 515 p.
ZIMMERMAN, E. C. 1948. Insects of Hawaii. Univ. Hawaii Press, Honolulu, vol. 1.





THE PREDATORY BEHAVIOR OF TOXORHYNCHITES
AMBOINENSIS AND TX. BREVIPALPIS LARVAE
(DIPTERA: CULICIDAE) IN RESPONSE TO
SUBSURFACE PREY

JOHN R. LINLEY
Florida Medical Entomology Laboratory, IFAS,
University of Florida, 200 9th Street S.E.,
Vero Beach, Florida 32962

ABSTRACT

Time lapse video recordings and high speed cinematography were used to provide
a description and analysis of the predatory behavior of Toxorhynchites amboinensis and
Tx. brevipalpis larvae. Only behavior in response to subsurface prey was examined
with emphasis on the bending response, in which larvae turn towards approaching prey,
and the strike, which effects prey capture. Bending was a very common response and
occurred when prey was positioned in any direction relative to the larva's body. To-
xorhynchites brevipalpis was more responsive than Tx. amboinensis; Tx. brevipalpis
larvae bent more rapidly, towards more distant prey, and through angles representing
larger proportions of the prey angle. Bend angle increased with increasing prey angle,
but as a proportion of prey angle, bend angle increased as prey angle decreased. Bend
angle was little affected by prey distance. Movement during being was smooth and
continuous. Each bend consisted of a brief accelerative and longer decelerative phase,
with average bending rates varying greatly depending on prey angle and distance.
Average bending rate increased with decreasing prey distance, the rate of increase
being especially rapid as prey approached close to the body. Prey capture during strikes
was accomplished in 0.012-0.024 s, and the entire strike completed in 0.060-0.076 s. Only
the lateral palatal brushes were used to capture prey. Immediately after capture, prey
was seized by the mandibles and released by the palatal brushes, which played no
further role in holding or manipulating food.
For descriptive convenience, 3 types of strike were recognized, frontal with head
extension, lateral with head extension, and lateral without head extension. These 3
form part of a continuous series. Frontal strikes involved little or no lateral turning
towards prey and involved dramatic forward extension of the head, accompanied by
opening and closing of the palatal brushes. Head extension was accomplished by sudden
increase in the larva's internal pressure resulting from rapid contraction of circular
muscles primarily in abdominal segments 1 and 2. Lateral strikes always involved some
degree of turning towards prey, and also some degree of head extension when prey was










10 Florida Entomologist 73(1) March, 1990

positioned at small angles to either side of the head, but not when prey was at larger
angles. Strikes made to larger angles also were accompanied by rotation of the head
through approximately 90', and partial rotation of the thorax. In lateral strikes the
degree of angular change during turning was usually greatest between head and thorax,
less within the abdomen, and least between thorax and abdomen. Angular velocity and
head speed (at the front margin of the head) during the strike varied, in 5 examples
filmed, from about 5,000/s and 210 mm/s, to 12,000/sec and 600 mm/s. Strikes were
made only when prey approached very close to the larva's head. Overall, the proportion
of successful strikes for Tx. brevipalpis (71.5%) was significantly higher than for Tx.
amboinensis (54.7%). In Tx. amboinensis, but not Tx. brevipalpis, the proportion of
successful versus unsuccessful strikes was significantly greater for prey located at smal-
ler angles relative to the head. In both species there was a linear relationship between
prey angle and strike angle for both successful and unsuccessful strikes, with the success-
ful strike angle averaging about 81% of prey angle in Tx. amboinensis and about 73%
in Tx. brevipalpis.


RESUME

Se usaron grabaciones de video magnetosc6pico continue y de cinematografia a alta
velocidad para proveer una descripci6n y un analysis del comportamiento depredador de
larvas de Toxorhynchites amboinensis y T. brevipalpis. Solo se examine el compor-
tamiento como respuesta a la superficie baja de la presa, con las larvas orientindose
hacia la presa que se aproxima, y el ataque que efectu6 la capture de la presa. Doblarse
fue una acci6n com6n que ocurri6 cuando la presa estaba en cualquier posici6n relative
al cuerpo de la larva.
La capture de la presa durante el ataque se llev6 a cabo en 0.12-0.024 segundos, y
todo el ataque se complete en 0.60-0.076 segundos. Solo usaron los cepillos laterales
palatales para capturar la presa. Inmediatamente despues de la capture, la presa fue
sujetada por las mandibulas y soltada por los cepillos palatales, los cuales no tuvieron
mas funci6n en aguantar o manipular la comida.
Por conveniencia descriptive, se reconocieron 3 tipos de ataque, frontal con extension
de la cabeza, lateral con extension de la cabeza, y lateral sin extender la cabeza. Estas
tres forman parte de una series continue. Ataques frontales involucran poco o ningfn
movimiento lateral hacia la presa y causa una dramatic extension de la cabeza hacia
adelante, acompafado del abrir y cerrar de los cepillos palatales. La extension de la
cabeza se llev6 a cabo por un aumento repentino de la presi6n internal de la larva,
causada por rapidas contracciones de los mdsculos circulares, principalmente de los
segments abdominales 1 y 2. Los ataques laterales siempre involucraron algtn grado
de orientaci6n hacia la presa, y tambien un poco de extension de la cabeza cuando la
presa estaba en posici6n de angulos pequefos a cualquier lado de la cabeza, pero no
cuando la press estaba en posici6n de anglos mayores. Ataques hechos hacia angulos
mayores tambien estuvieron acompafados por rotaci6n de la cabeza hasta ap-
roximadamente de 900, y una rotaci6n parcial del t6rax. En ataques laterales, el grado
del cambio del angulo durante la orientaci6n fue usualmente mayor entire la cabeza y el
torax, menos en el abd6men, y menor entire el t6rax y el abd6men. La velocidad angular
y de la cabeza (al margen frontal de la cabeza) durante el ataque vari6, en 5 muestras
filmadas, de cerca de 5,000/is a 210mm/seg, a 12,0000/seg y 600 mm/seg. Los ataques
se hicieron solo cuand6 la presa se aproxim6 much a la cabeza de la larva. Teniendo
todo en cuenta, la proporci6n de ataques exitosos por Tx. brevipalpis (71.5%) fue sig-
nificativamente mayor que por Tx. amboinensis (54.7%). En Tx. amboinensis, pero no
en Tx. brevipalpis, la proporci6n de ataques exitosos y no exitosos fue sig-
nificativamente con press localizadas a angulos pequefios relatives a la cabeza. En
ambas species hubo una relaci6n linear entire el angulo de la presa y el angulo de ataque
en ataques con 6xito y sin exito, con el angulo de ataque exitoso promediando de 81%
del angulo de la presa en Tx. amboinensis y aproximadamente un 73% en Tx. brevipal-
pis.











Insect Behavioral Ecology-'89 Linley


The predatory habits of Toxorhynchites larvae first attracted attention from en-
tomologists many years ago (Green 1905, Banks 1908, College 1911). Modern studies
have enhanced understanding of the behavior, but have generally tended to emphasize
the behavior's quantitative impact on both predator and prey populations, rather than
description and understanding of the behavior itself. Consequently, much is known
about the numbers of prey consumed in the laboratory (Padgett & Focks 1980, Frank
et al. 1984), and integrated laboratory and field observations have revealed interactions
between predator and prey in nature (Trpis 1973, Lounibos 1979, Frank et al.
1984,Lounibos et al. 1987).
In contrast, accounts of the behavior of Toxorhynchites larvae on the approach of
prey and during its capture are given only in very general terms (e.g. Breland 1949,
Muspratt 1951, Goma 1964, Crans & Slaff 1977, Furumizo & Rudnick 1978), with little
or no quantitative information. Some accounts have mentioned that Toxorhynchites
larvae are capable of detecting potential surface prey from short distances if the water
surface is disturbed. Detection is said to be by means of large thoracic setae (Paine
1934, Rubio et al. 1980), and may cause the larva to approach the prey either by inching
forward using the siphon as a pivot (Crans & Slaff 1977), or by swimming (Paine 1934,
Breland 1949). Neither of these behaviors has been closely studied and Steffan &
Evenhuis (1981) concluded from the literature that larvae only rarely swim towards
prey. Another easily observed reaction towards approaching prey is that in which larvae
bend their bodies to being the head closer to the prospective victim. This behavior,
referred to here as bending, has rarely even been mentioned (Russo 1986), despite its
probably importance in increasing the success rate during attempted prey capture.
Movements that accompany the final strike, which effects seizure of the prey, are
extremely rapid. Consequently, the strike behavior has been described with considera-
ble uncertainty in the literature, especially with regard to the mouthparts involved.
Breland (1949) writes that "it has been stated frequently that the larvae use the brushes
(the lateral palatal brushes of Harbach & Knight 1980) to seize and hold the prey", but
then admits that he did not determine if these statements resulted from observation or
conjecture. His own observations, however, led him to conclude that the mandibles and
not the palatal brushes were used to capture prey, with the possibility that proper
mandibular function was in some way dependent on the brushes. Involvement of the
brushes was suggested by the failure of two larvae with damaged brushes to capture
prey. Likewise, Furumizo & Rudnick (1978) concluded that the brushes played some
role, but at the same time stated that the mandibles primarily were employed to seize
and hold prey. Going a step further, Russo (1986) stated that "the mandibles opened as
the head reached its maximal extension", a claim one may reasonably question in view
of the fact that prey capture requires only 1/50-1/75 s (see results). The fact is that such
brief events simply cannot be resolved by unaided stereomicroscopic observation. Re-
liance on this alone has led not only to incorrect conclusions, because the palatal brushes
and not the mandibles are used to seize prey (see results), but has provided no informa-
tion about the physical actions used to effect prey capture, their timing, or how they
are accomplished.
In this paper, I have provided a more complete and quantitative analysis of certain
aspects of predatory behavior in Toxorhynchites larvae, using Tx. amboinensis (Doles-
chall) and Tx. brevipalpis (Theobald) as experimental animals. Emphasis was entirely
on behavior connected with response to approaching prey, namely the bending reaction,
and with the strike sequence itself. The reaction whereby larvae turn and swim towards
the source of surface disturbances (Paine 1934, Breland 1949) is not considered here
because it is primarily a response to surface prey, it is probably much more common
and important than is suggested by the literature, and is sufficiently complex to warrant
separate analysis. When the mosquito larvae used as prey in the present study occasion-










Florida Entomologist 73(1)


ally broke the water surface with their siphons, the Toxorhynchites larvae frequently
swam towards the source of disturbance.

MATERIALS AND METHODS

Insects

Tx. amboinensis and Tx. brevipalpis larvae were taken from laboratory colonies
reared at 27C under a light regimen of LD 12:12. Aedes aegypti L. larvae were provided
in excess as prey throughout development. Toxorhynchites larvae used for observation
were removed from the colony pans on their 2nd day in the 4th instar, placed individu-
ally in small wells containing tap water and starved for 72 h at 25-270C, which was also
the temperature at which all subsequent observations were done.

Methods of recording behavior

Video tapes were used for study of some aspects of both bends and strikes; high-
speed cinematography were used only for recording movement during strikes. Video
records were made as follows. The starved larvae were placed individually in plastic
petri dishes, 8.5 cm diameter, 1.4 cm deep, containing tap water about 1 cm deep. Each
larva was left undisturbed for 5 min, before addition of a single large 4th instar Ae .
aegypti larva to serve as prey. Behavior was then recorded for 5 min or until the prey
larva was captured. Recording was done with a Panasonic NV 8050 time-lapse video
recorder and General Electric TE-44BSA camera fitted with a 7.6 cm (focal length) lens
mounted in front of a small extension tube. Individual Toxorhynchites larvae were
numbered sequentially by means of numerals set close to one side of the dish within the
recorded image field. Records were obtained from 112 Tx. amboinensis larvae and 130
Tx. brevipalpis. Of these, 94 Tx. amboinensis and 114 Tx. brevipalpis were used for
analysis, the remainder being considered superfluous.
For cinematography of strikes, only Tx. brevipalpis larvae were used. They were
treated exactly as for video records, except that they were placed prior to filming in
dishes 3.5 cm in diameter and 0.9 cm deep, filled to about 0.5 cm deep. Smaller dishes
allowed greater magnification on film. and the shallower water forced the larva's body
to the desired position, almost perpendicular to the optical axis. Larvae were illumi-
nated during filming with 3 lights totalling about 1000 watts, allowing the necessary
effective shutter speed of 1/1500 s, using the camera's variable shutter and Kodak 4-X
16 mm reversal film (ASA 320). The camera was a Locam model 51002, capable of
filming speeds up to 500 frames/s over a continuously variable range. At the high filming
rates used (350 and 500 frames/s,) much film was wasted during "false starts", when
the camera was started in anticipation of a strike that eventually did not occur. A few
prey larvae were therefore presented by holding them in fine forceps and slowly advanc-
ing them towards the predator. On analysis, the initial filming rate of 350 frames/s
proved too slow and only 1 strike filmed at this speed, because it illustrated one type
of strike particularly well, was used eventually for analysis. In addition, 5 strikes filmed
at 500 frames/s were analyzed.

Methods of analysis

Review of the video records showed that where measurements of changes in the
angle of the predator's body were needed, only certain bends and strikes could be used.
However, the sector (see below) in which prey was positioned was determined for every
bend or strike recorded. When bends occurred, the body of the Toxorhynchites larva


March, 1990











Insect Behavioral Ecology-'89 Linley


was usually in a straight attitude and all such bends were measured. On those occasions
when a bend in one direction was followed immediately by an opposite one (i.e. initiated
from a bent position), only the first bend was analyzed. More rarely, when a larva
formed a limited angle in response to rather distant prey and then, after a pause, bent
further in the same direction as the prey approached more closely, only the initial bend
was measured and only one bend was scored. In the case of strikes, which often were
initiated from a bent position, all were measured, except in relatively rare instances
where surface reflections interfered with the image, or two or three strikes occurred
in rapid succession, when only the first was measured.
The video image on the monitor screen was about 2.2x life size. The record for each
larva was played back until either a bend or a strike occurred. The video "fields" (each
corresponding to 1/60 s) were then reversed at slow speed and stopped at the field
immediately preceding that in which motion started. A machinist's caliper was then
used to measure the distance of the nearest part of the prey to either (for bends) the
middle of the predator's body (middle of the 3rd abdominal segment), or (for strikes),
the middle of the head. The angle of the nearest part of the prey to the head (or to the
middle of the body for bends) was subsequently measured with a transparent overlay
made from polar coordinate graph paper. The overlay was positioned for head measure-
ments with its central point in the middle of the head (or 3rd abdominal segment for
bends) and with the 0 line aligned (for strikes) along the mid-line passing through the
head to the middle of the thorax and, for bends, along the longitudinal axis of the body.
Such measurements were made for plotting the position of prey, but it was also neces-
sary to determine angular change during the bend or strike. To do this, for both bends
and strikes, the central point of the overlay was moved to the middle of the thorax,
judged to be the approximate pivot point of the body, and the angle to prey measured
again. With the overlay kept in position, the video fields were then advanced until the
bend or strike motion had reached maximum amplitude, whereupon the angle of dis-
placement was recorded. It was often necessary to re-position the overlay's central
point slightly to compensate for small shifts in the position of the thorax, but care was
taken to move the overlay without rotation from its original alignment.
A similar method was used to estimate changes in the rate of bending throughout a
few selected bends (see results). In this case, points showing the starting alignment of
the predator's head, mid-thorax and siphon were marked on a transparent overlay, then
the image was advanced uniformly by 4 to 30 field increments, as necessary to resolve
movement for different bending rates, and the position of the middle of the head marked
at each increment. Angles relative to the starting head/mid-thorax alignment were then
measured from enlarged copies of each overlay.
Data on strikes were obtained only from filmed sequences. Each 16 mm image was
first re-photographed at 5.7x magnification onto 35 mm Kodak Panatomic-X film using
a bellows and Zeiss 63 mm Luminar lens. Resulting negatives were then printed at 6. 1x
magnification, making the final image of the larva about 80 mm long, or approximately
8x life size. All measurements of head movements and spatial changes in the abdomen
(see results) were made from these enlarged prints, using a machinists's caliper, and
were then converted to actual distances. To measure changes in angle between head
and thorax (Hd-Th), thorax and abdomen (Th-Ab), and within the abdomen (Ab-Ab),
high contrast photocopies were first made of these prints. White lines were drawn to
mark the longitudinal axes of the head, thorax, and abdominal segments 1 and 2 and 7
and 8 (Fig. la), and the required angles (Fig. la) measured with a protactor.
In addition to angular changes between body parts, it was necessary for most strikes
to determine the actual arc of motion of the head and its speed and angular velocity.
For this, each 16 mm frame was projected (magnification about 16x life size) to exact
registration on white paper taped to a screen (small background spots in the image








March, 1990


Florida Entomologist 73(1)






ft.


-AAb
PIP


a


4


Fig. 1. a, angles measured to record angular changes in body during strike; b, a
typical bend towards prey; c, a frontal strike with head extension from bent position;
d, closure arc of lateral palatal brushes and angle measured to record brush movement.

served as fiducial markers). The position of a central point on the anterior margin of
the head in each frame was then marked and the appropriate angles (relative to mid-
thorax) and distance later measured.


b
J-"3










Insect Behavioral Ecology-'89 Linley


Several aspects of the results are presented by data grouped according to which 300
sector of a circle (Table 1) the prey occupied when response occurred. The circle's center
was in the middle of the head for strikes and in the middle of the body (abdominal
segment 3) for bends.

Scanning electron microscopy

Scanning electron micrographs were made of mouthparts of 4th instar Tx. amboinen-
sis and Tx. brevipalpis larvae reared in clean water to prevent contamination. The
animals were fixed in 30% ethanol, dehydrated in ethanol, and dried finally by the
critical point method. Specimens were coated with gold and examined in a Hitachi S-510
scanning electron microscope.

Limitations of the study

The study was limited to movements that were lateral with respect to the larva.
Larvae occasionally bent downwards in response to prey passing beneath, but these
movements were of small amplitude and were uncommon in the dishes used here. Bend-
ing towards surface prey is common, as will be described in a later paper, but the
behavior is essentially the same as described here. In the case of strikes, larvae occa-
sionally struck at prey slightly beneath the head, and they will do so at surface prey
(Breland 1949). Again, however, the important elements of the behavior are rep-
resented by present data.

RESULTS

Confirmation of bending as a subsurface behavior

It was known (unpublished personal observations) that surface prey would elicit the
bending response. To be sure that bending occurred in the absence of surface effects,
which were sometimes caused by the Ae. aegypti larvae, 10 Tx. amboinensis and 10
Tx. brevipalpis larvae were isolated as usual, but a large 4th instar Culicoides variipen-
nis (Coq.) larva was added as prey. The Culicoides larvae swam always beneath the
surface, but to be certain that surface contacts were not made (that might be undetect-
able from the video image), the dishes were watched until the midge larva approached
closely enough to elicit a bending response. All the Toxorhynchites larvae except one
amboinensis, which did nothing within 5 min, eventually responded without any surface
disturbance that could be seen under close scrutiny by reflected light. Subsequently, 2
larvae of each of the predator species, ones that had not captured the midge larva, were
placed in a deeper container, in water about 2 cm deep, and induced to dive by gentle
probing. A midge larva was again added and, within a few minutes, bending was ob-
served in 3 of the submerged Toxorhynchites larvae.

THE BENDING RESPONSE

Frequency and appearance

Of the 94 Tx. amboinensis and 114 Tx. brevipalpis larvae observed, 87.7% and
93.4%, respectively, bent towards prey. Several animals in each case bent 5 or 6 times
as the Ae. aegypti larvae moved about them. Of the few larvae that did not bend, most
were approached frontally, so that bending did not occur before prey capture. Bending
was a very distinct, clear reaction in which the Toxorhynchites larva turned towards
the prey (Fig. Ib), often when the latter made a sudden, vigorous movement.











Florida Entomologist 73(1)


Position and distance of prey

In both predators, bends were observed towards prey approaching from or moving
in any sector (Fig. 2, Table 1). Fewer bends were recorded for prey located at small


Tx. amboinensis


Tx. brevipalpis


I


. -C


20mm


Fig. 2. Spatial distributions of prey (nearest part) when bends initiated.


March, 1990











Insect Behavioral Ecology-'89 Linley


TABLE 1. PREY DISTRIBUTION (BY SECTOR) AT BEND INITIATION AND MEAN DIS-
TANCES OF PREY FROM MIDDLE OF BODY.

Tx. amboinensis Tx. brevipalpis
Sector' No. in Mean distance2 No.in Mean distance2
(0) sector % (mm) sector % (mm)

0-15 2 1.1 7.0 10 4.6 22.4
16-45 27 14.7 9.9 43 19.6 15.5
46-75 40 21.7 7.8 51 23.3 15.1
76-105 51 27.7 7.4 42 19.2 16.2
106-135 34 18.5 7.6 26 11.9 16.6
136-165 23 12.8 8.1 40 18.3 11.7
166-180 7 3.8 5.8 7 3.1 15.2

'Relative to middle of body (abdominal segment 3)
'Measured to middle of body


angles to either side of the front of the predator's head, or behind the tail, especially
in the case of Tx. amboinensis (Fig. 2, Table 1). Bends of very small amplitude some-
times occurred in response to frontally positioned prey, even, in the case of Tx. brevipal-
pis, if the Ae. aegypti larvae was some distance away (Fig. 2). Otherwise, the positions
of prey were distributed symmetrically about the body, with most (total 67.9% for Tx.
amboinensis, 54.4% for Tx. brevipalpis) in the 3 sectors from 46-1350 (Table 1). Prey
positions were not clustered with an obvious focus around any particular part of the
body, but rather evenly surrounded each predator (Fig. 2), suggesting that the sense
organs involved in bending are present along much of the predator's body.
Toxorhynchites brevipalpis larvae were more responsive than Tx. amboinensis and
bent when prey was considerably further away (Fig. 2). The mean distances at which
prey elicited bends were quite uniform between sectors for both species (Table 1), with
Tx. brevipalpis distances about twice those for Tx. amboinensis. The maximum distance
recorded for Tx. brevipalpis was 49.6 mm, compared to 24.6 mm for Tx. amboinensis,
while the minimum distances were 1.9 mm and 2.4 mm, respectively. Most prey were
distributed in a diffuse belt around the predator (Fig. 2), with density falling off with
increasing distance. Since the distributions were similar between sectors, all were com-
bined in each case to reveal the strongly skewed overall patterns (Fig. 3), especially in
the case of Tx. brevipalpis. The main range of response for both species, with the prey
used, was between 3 and 12 mm (Fig. 3), about 1/3 to 1 body length away.

Nature and rate of bending

During bending, most angular change within the body took place between the ab-
dominal segments and less between the abdomen and thorax and thorax and head (Fig.
Ib). To examine changes in bending rate, 5 selected bends of different bend angles and
prey distances were analyzed for each species (Fig. 4). As seen most clearly in Tx.
amboinensis, bending consisted of a short accelerative phase and a longer decelerative
one (Fig. 4). With one exception, the time increments used were too great in the case
of Tx. brevipalpis to resolve the brief accelerative phase. In Tx. amboinensis, acceler-
ation of the head was complete in from 0.14-0.60s, whereas it was over in less than 0.14s
in all the Tx. brevipalpis bends except one rather slow one (open circle, Fig. 4).
Maximum bending rate varied considerably in both species, from about 300/s to over
450/s, with the rates in Tx. brevipalpis tending to be somewhat higher. Bends to larger
angles were executed at higher bending rates in both species, as measured by calculat-









Florida Entomologist 73(1)


30-

24-

18-


12-


40-

32-

24-


16-


"1


Tx. amboinensis


Tx. brevipalpis


0 F44VZZAVIZZA II
0 10 20 30 40 50

Prey distance (mm)
Fig. 3. Distance distributions of prey (all sectors combined) when bends initiated.




ing the regressions of bend rate on bend angle for prey distances of 5-10 mm. Both
relationships were linear (not shown), with highly significant regression coefficients
(1.044, P < 0.01, for Tx. amboinensis, 1.468, P < 0.01 for Tx. brevipalpis), implying,
respectively, increases of 10.4 and 14.7/sec for each 100 increment in the bend angle.


M lIi A n


C Y"Y"/ "'UY" If/Mf / f l


March, 1990


Ifn' ri F










Insect Behavioral Ecology-'89 Linley


Tx. amboinensis
Prey distance (mm) Bend angle (0)
A 1.8 72
2.5 98
A 4.2 56
0 5.6 61
S 5.8 35


400-


320-


240-


160-


80-


0-


Tx. brevipalpis


Prey dista
A 1.
\* 3.
A 6.
0 26.
.\A o 35.


A0 0

S o
+-r ^-a,* :--i-n.


nce (mm)
0
2
1
4
2


Bend angle (0)
90
134
93
82
74


0.00 0.28 0.56 0.84 1.12 1.40 1.68 1.96 2.24 2.52

Time (sec)
Fig. 4. Bending rates throughout 5 selected bends completed to different bend angles
for prey at different distances.

Prey angle, bend angle and prey distance

Both predators showed similar relationships between prey angle and bend angle
(Fig. 5). The data could be described empirically in each case by positive exponential
regression (Fig. 5), but considerable variation in bend angle was observed for prey in
any particular position, and bend angle rarely exceeded prey angle. Variation also
increased as prey angle increased. Thus, although the Toxorhynchites larvae turned
towards prey, the degree of turning was very varible and only rarely was a bend made
completely to or beyond the prey angle.
The relationship between degree of bending and prey angle (Fig. 6), showed that
bend angle as a proportion of prey angle was least for the largest prey angles. It then
increased with decreasing prey angle relatively slowly, but then progressively much


I I I I I I I I I T 1


0
a)
0)

C)
-2
oV

cm


I I I I I I I


0O 500-
C
'L0
V
C 400-
-


300-


200-


100-

n










Florida Entomologist 73(1)



y = 14.05(2.720.009)





*


150-




120-




90-




60


March, 1990


Tx. amboinensis







00 0 0
* *
*- '-*: r*


.0 0
.0 *
0 0.0 ~
*


"0
:.l


0
0


y = 20.33(2.720.007x)


\


0t
0


I i I I I


Tx. brevipalpis


0

0 0


00
0. 0
0 00 :
0 0 *
0 0000
0 0cd
.0


* *0
.* *
.0 0


0


$0
0


Oj .0


* 0.*


108


144


Prey angle (0)


Fig. 5. Relationship between prey angle and bend angle.


0 0
~~7 0


30-




0


150-


00

a)





C
0)

4)
In


120-


90-




60-




30-




0i


*
* * ** ** 4 .
..* *
* 0 0. 1
.00 00 L


*


36 72


180


-










Insect Behavioral Ecology-'89 Linley


Tx. amboinensis

y= 116.37 15.85 logx

120


*
96



72 .
72- .

C *
3 48 0-: T.. *: .
, ^ %." ."

a *
24. *" *." "
0.0 .
.. * *









2 240-



0 180 I
3 0
C%













120- .

300. Tx. br vip
.* ,.* 1* :
0 ) *2, 40*
.* .* >*** .y *
0,











0 36 72 108 144 180

Prey angle (o)

Fig. 6. Relationship between prey angle and bend angle as a percent of prey angle.
Fig. 6. Relationship between prey angle and bend angle as a percent of prey angle.










22 Florida Entomologist 73(1) March, 1990

more rapidly as prey angle diminished below about 400. By analysis of variance, the
data were best described by negative logarithmic regressions (Fig. 6). From these, it
was apparent that bend angle as a percent of prey angle was, for prey angles of 25, 50
to 1700, about 65, 54, and 35% for Tx. amboinensis and 103, 78 and 33% for Tx. brevipal-
pis. Again, Tx. brevipalpis appeared more responsive, and turned towards the prey by
a greater proportion of the prey angle. Additional evidence for this came from a simple
comparison of the two species in terms of the average degree of bending, by sector
(Table 2).
In contrast to prey angle, the distance of prey affected the degree of bending very
little in either species (Fig. 7). For any prey distance there was a great deal of variation
in the proportion of the prey angle to which either predator would bend. Calculation of
linear regressions (Fig. 7) gave negative coefficients for both species (-0.365 for Tx.
ambionensis, -0.310 for Tx. brevipalpis, but only the value for Tx. brevipalpis was
significant (P < 0.02), because very few values for Tx. amboinensis were obtained at
longer prey distances (Fig. 7). The very limited effect of prey distance was well illus-
trated by the data for Tx. brevipalpis, where bend angle increased only 3.1% of prey
angle for every 10 mm that the prey more closely approached the predator.
Although prey distance had little effect on the degree of bending, its effect on the
average bending rate was substantial (Fig. 8), as determined by slow motion timing of
complete bends of known angle with a stop watch. Data for the two Toxorhynchites
species were quite similar and were fitted by negative power regression equations (Fig.
8). Until prey had approached quite closely, bending rate increased only slowly as prey
distance decreased. At about 4 mm in Tx. amboinensis and 10 mm in Tx. brevipalpis,
however, the rates began to increase rapidly (Fig. 8), indicating much more rapidly
incrementing turning rates as prey moved closer within a narrow range very near the
body. Average bending rates did not often exceed 1500/s in either species until prey
was closer than 4 or 10 mm, but then a number of bends, particularly in Tx. brevipalpis,
considerably exceeded 2000/s (Fig. 8).

THE STRIKE

General observations

Execution of the strike movement in Toxorhynchites larvae is extremely rapid and
its component movements cannot be seen under a stereomicroscope. In this study, only
strikes of Tx. brevipalpis were recorded by high speed cinematography, but the be-
havior of Tx. amboinensis and other species can be assumed to be very similar.
Basically two movements of the body may be involved during a strike; head exten-
sion (and retraction), and lateral turning of the body (contraction and relaxation). These
may or may not be combined, depending on prey angle, to produce 3 basic types of
strike, (i) frontal with head extension, (ii) lateral with head extension and (iii) lateral
without head extension. The distinction between the 3 types is made for descriptive


TABLE 2. MEANS, BY SECTOR, OF BEND ANGLE AS A PERCENT OF PREY ANGLE.

Sector Tx. amboinensis Tx. brevipalpis
(0) Mean % Mean %

0-45 64.2 107.2
46-75 49.8 57.3
76-105 39.2 47.7
106-135 35.3 45.5
136-180 41.1 44.3











Insect Behavioral Ecology-'89 Linley


Tx. amboinensis


y = 43.91 0.37x


* *
*


* 0


g. 0


0 *


* **
0 0g
% ooe


5.6


11.2 16.8 22.4 28.0


Tx. brevipalpis


. y = 55.97 0.310x


0)


Cu
0
c-



0)
0





C0

(Q


0)

(0
*0

0)
(D


*; /. .*
S ,. *
00 *





0. *. * 0
*. ,,. .* . 0
*. o .*


S 10
10


I I I I I I I I
20 30 '40 51


Prey distance (mm)
Fig. 7. Relationship between prey distance and bend angle as a percent of prey
angle.


120-



96-



72-


0
0 0
~0
S
~ ~ 0


48



24



0--
0.0


120-



96


72-


48-



24-



0-
(


_









Florida Entomologist 73(1)


* y = 263.45x-0.80


*
* 3.
0t


. -


0 0~ O S


4 8 12 16 20
4 8 12 16 20


Tx. brevipalpis


y = 525.62x-0.93


* 0


*


* :t
*


I I I 3
3 10 20 30 40 5(


Prey distance (mm)
Fig. 8. Relationship between prey distance and bending rate.


Tx. amboinensis


700-


600-


500-


400-


300-

200-


I *


0
0
(0
0

4.

CI



0
m


100


0-
0


500-



400-


300-



200-



100-



0-


March, 1990


4 . * .











Insect Behavioral Ecology-'89 Linley


convenience only, as it will be shown in due course that the three form part of a
continuous series.

Frontal strike with head extension

Strikes of this type occur when the prey is directly in front of the Toxorhynchites
larva. Head extension and retraction are involved, but very little or no bodily turning,
although the larva's body may already be acutely bent when such a strike is made (Fig.
Ic). One of the strikes filmed at 350 frames/s showed this type of strike particularly
well (Fig. 9).
The remarkable feature of this strike was that it was accompanied by extraordinarily
rapid forward extension of the head. The entire strike sequence consisted of extension
and retraction phases (Fig. 10a), with extension being about half the duration of retrac-
tion. Forward head movement increased progressively at first and then diminished
towards the end of the extension phase (Fig. 10a). Thus, the extension phase in ex-
panded time scale (Fig. 10b) revealed that head speed at first accelerated to a maximum
of about 145 mm/s, then diminished as the head approached full extension, which was
completed in about 0.014 sec (Fig. 10b). About 60% of the head's extension took place
in the period from about 0.006 to 0.011 s. Concurrently with movement of the head, the
lateral palatal brushes, which clearly were the appendages used to capture prey (Fig.
9), opened and subsequently closed on the victim. Opening of the brushes (arrow, Fig.
9) occurred very rapidly (in < 0.003 s), momentarily after head extension began. As
measured by the angle of the brushes (Figure Id), the rate of closure was then relatively
constant (Fig. 10b) at about 10,0000/s. Thus, important events in the action of the palatal
brushes took place at the following times after strike initiation; opening, 0.006 sec; first
contact with prey (frame 6, Fig. 9), 0.014 s; completion of prey capture (black dot, Fig.
9), 0.02 s.
Front (Fig. lla) and side views of the head (Fig. l b) show that when the Toxorhyn-
chites larva is at rest the folded palatal brushes are directed ventrally, just inside the
protruding antennae. For the brushes to open (Fig. 11c) at strike initiation, they must
either lift sideways through 90, momentarily pushing aside the antennae, or swing
forwards and upwards before spreading to the side. However opening is accomplished,
the brushes are obviously much better adapted for snaring prey than the mandibles
(Fig llc). They are able to reach further in front of the head, where they close through
overlapping arcs, each of approximately 0.46 mm radius (Fig. ld). Each brush,
moreover, is curved inwards apically (Fig. 12a), which prevents escape of prey and pulls
it towards the mouth, while terminal hooks on almost all the brush elements (Fig. 12b)
provide added security. It is noteworthy that each brush element is distinctly flattened
in the plane of the closure arc (Fig. 12b), probably to enhance strength and rigidity and
to reduce resistance.
Although the palatal brushes were used to capture prey, all of many Tx. brevipalpis
larvae examined under a stereomicroscope immediately after prey capture were holding
prey with the mandibles. In no instance of many larvae of both species watched through-
out the entire consumption of prey, were the palatal brushes used to hold or manipulate
food. Prey is captured with the brushes, but is apparently transferred to the mandibles
as soon as it is drawn close to the head.
After prey capture, the head retracted back towards the thorax. This phase, plotted
in expanded time scale (Fig. 10c) showed rapid acceleration in head speed at first, to a
maximum of about 55 mm/s, before speed gradually diminishing as the head returned
to its starting position (Fig. 9). The entire strike sequence lasted about 0.05 s (Fig. 10a).











Florida Entomologist 73(1)


March, 1990


v" ~

Cp : tJ


. 'V'


Fig. 9. Frontal strike with head extension (Tx. brevipalpis); frames at 1/350 (0.0029)
s intervals. Sequence reads top to bottom, left to right. Arrow indicates opening of
palatal brushes, dot marks completion of prey capture.


~khd~*~
;r


~i~L~i~
.;,.t - x.- :.` c


''


")t
C;F i~i~
i .Ir


'-- m.










Insect Behavioral Ecology-'89 Linley


extension


retraction


0
E


> (
o


E

0





0100
C


60-

, 40- 0

a) 20[-

CL O


1.22

).97

).73

).48

).24

.00


a 50
E
E 40
0 30

20
0
0


0
I

d


II











)/


0

0
O5


a


0
0\
0\
*S0.. **


I a


(D M r CV) 0)
o o o C
o0 0 0 0


Time (sec)

Fig. 10. Analysis of events during a frontal strike with head extension (Fig. 9). a,
movement of head during complete strike (extension and retraction); b, head speed,
head extension, and palatal brush angle during head extension; c, head speed during
retraction.


--


I I I I I i









28 Florida Entomologist 73(1) March, 1990










Insect Behavioral Ecology-'89 Linley


Fig. 11. Scanning electron micrographs showing, a, front view of head (Tx. am-
boinensis), dorsal side uppermost, with palatal brushes at rest position; b, same as (a),
partially oblique lateral view; c, front view of head (Tx. brevipalpis), with palatal
brushes in open position. Scale = 1 mm.


Mechanism of head extension

Toxorhynchites larvae do not possess powerful circular muscles in the neck such as
would be necessary to extend the head. In the photographs, however, there were
visible changes in the spacing and dimensions of the abdominal segments during the
strike (Fig. 9). A machinist's caliper was used to measure the intersegmental diameter
between the back of the thorax and abdominal segment 1, and also between successive
abdominal segments up to segment 5. The longitudinal distances separating these parts
were also recorded.
Over the period of head extension, intersegmental diameter increased between the
thorax and abdominal segment 1 and much more so between segments 3-4 and 4-5 (Fig.
13). Especially between segments 1-2 and also between 2-3, however, the diameter
decreased very rapidly, as visible in the photographs (Fig. 9). Corresponding changes,
but of opposite trend, took place in the distance separating segments (Fig. 14). During
retraction, both dimensions returned to the initial state as the head returned to its
normal position (Fig. 13, 14).
These measurements indicated that the head was extended by a hydrostatic
mechanism powered by very rapid contraction of the abdominal muscles, particularly
in the first and second abdominal segments. Contraction produced a sudden increase in
internal pressure, forcing the neck to extend rapidly and carry the head forward a
distance of somewhat over 1 mm (Fig. 10a).










Florida Entomologist 73(1)


Fig 12. Scanning electron micrographs (Tx. brevipalpis) showing, a, enlarged view
of single palatal brush; b, detail of tips of individual brush elements. Scale = 50 pm.


March, 1990










Insect Behavioral Ecology-'89 Linley


Th Ab 1


1.69

1.58

1.47

1.36

1.25

1.69

1.58

1.47


- Ab 1 Ab 2

-0

S0
-A 2 *-*Ab



Ab 2- Ab 3


* l 0
S* 0


1.36 H


1.251

1.80

1.69

1.58

1.47
1.80

1.69

1.58

1.47,


I I I


I I I


*. Ab 3 Ab 4





Ab4-Ab

r Ab 4 Ab 5


1.25-


d d


Time (sec)


Fig. 13. Intersegmental diameters
strike with head extension (Fig. 9).


between indicated body parts during a frontal











Florida Entomologist 73(1)


March, 1990


0.42

0.36

0.30

0.24
1.00

0.94

0.88

0.82


0.70 V
E 0.94

C 0.88

0.82

r. 0.76
4)
- 0.70
E 0.90
0)
E 0.84
0)
CI) 0.78

0.72

0.66

0.60


0.78

0.72

0.66


Th Ab 1


\NO 0s
.0 **
0 ,
`-,~ ~0
- I I


S Ab 1 -Ab2


F


I I I I I I


Ab 2-Ab 3
<*'" -'* .--~-..,
> *-**** 0
^- 0
0 6O 0
*


AI I3 Ab 4
Ab 3 Ab 4


0 *
0 *~~-
S
,,~0
0
0 0


I I I I I
Ab 4 Ab5


0.60 I I I I i ,
o O Ol 1O 4 7-


Time (sec)
Fig. 14. Segmental separation between indicated body parts during a frontal strike
with head extension (Fig. 9).


I


I


I


I










Insect Behavioral Ecology-'89 Linley


Lateral strike with head extension

If the prey was slightly to one side of the head, the larva sometimes extended its
head during the strike and also turned its body a limited degree (Fig. 15). As illustrated
below, however, when the prey angle became too great, the head was not extended,
probably because muscular actions necessary to induce extension could not be executed
at the same time as those required to turn the body to any appreciable degree. In
addition, drag effects would probably make it impossible to have the head extending
forward during such rapid turning. The video records were re-examined to determine
the approximate maximum limit of turning that still permitted head extension, but it
proved impossible to discern whether many strikes had contained an extension compo-
nent. It was considered unlikely, however, that head extension occurred for prey angles
greater than about 30.
In the example filmed (Fig. 15), head extension was not measured as it seemed
generally consistent with the example already described. Also consistent were the ac-
tions of the lateral palatal brushes. They opened (arrow, Fig. 15) 0.004 s after the strike
was initiated, and completed closure momentarily after full extension of the head, thus
effecting prey capture (dot, Fig. 15) in about 0.018 s. Pronounced contraction between
abdominal segments 1 and 2 was again visible. For reference, this larva was denoted
as larva 1 and its turning movements were analyzed together with those of 4 other
larvae that made strikes without head extension (see below).

Lateral strike without head extension

Strikes of this type were made when prey was positioned more to the side of or
behind the head. Four examples were filmed (Larvae 2, 3, 4, 5), of which the one that
showed the most extreme degree of turning (larva 4) was selected for illustration (Fig.
16). This was an unsuccessful strike at prey held in tweezers, but the strike movements
were stereotyped and were not affected by whether prey was ultimately captured.
There were two phases in such strikes, contraction (body turned towards prey with
associated movements of mouthparts, head and thorax), and relaxation (body returned
to rest position). Several important differences distinguished these strikes from ones
incorporating head extension. Firstly, the head clearly was not extended (Fig. 16).
Early in the strike, from about 0.004 to 0.008 s (frames 2, 3, 4, Fig. 16), the head began
to rotate so that as the body turned to its maximum extent (full contraction), the head
was oriented with its dorsal and ventral surfaces in the vertical plane (frames 10-13).
Simultaneously, but to a lesser degree, the thorax was rotated (Fig. 16) and was about
half turned at full contraction. The palatal brushes did not open as soon in this particular
strike as when the head was extended. The head was at least half rotated (abopt 0.010
s after strike commencement) when the brushes were first discernible (arrow, Fig. 16).
Their subsequent closing could not be seen, but was assumed to be complete by the
time the body reached full contraction (dot, Fig. 16). Once contraction was completed,
relaxation to the rest position took place more slowly (Fig. 16). Summary of the 4
examples filmed (Table 3) showed that opening of the palatal brushes was usually very
soon after strike initiation, but was rather slow in larva 4. Head rotation was completed
in the first half of the period required to reach full contraction, which also closely
approximated the time to prey capture (0.012-0.024 s. Table 3). Relaxation lasted about
twice as long as contraction and the entire strike (both phases) was concluded in from
0.058-0.076 s.
Angular change between body parts during the strike (Fig. 17, Table 4) indicated
that amplitude of change was usually greatest between the head and thorax, although
in larva 4 the angle of change was greatest between the abdominal segments (Fig. 17).
Change in angle was usually least between thorax and abdomen (Table 4). As regards
timing, angular change during contraction occurred most quickly between head and





Florida Entomologist 73(1)


r
rPg
Nj~


Cr
N,9


rfc


r


Fig. 15. Lateral strike with head extension (Tx. brevipalpis); frames at 1/500 (0.0020)
s intervals. Sequence reads top to bottom, left to right. Arrow indicates opening of
palatal brushes, dot marks completion of prey capture.


March, 1990







Insect Behavioral Ecology-'89 Linley


4A



001 -dO


js o

3 rA
F1


*Sol


.

YI~C
''*'
o

:C"

.~
'
1 ~"


r


-Ik
'! r _opt
^' IB

l 40"11 1 (K11Popp
Oft -tl T


e*


Fig. 16. Lateral strike without head extension (Tx. brevipalpis); frames at 1/500
(0.0020) s intervals. Sequence reads top to bottom, left to right. Arrow indicates opening
of palatal brushes, dot marks completion of prey capture.


-CL


~:


'MIN










36 Florida Entomologist 73(1) March, 1990

TABLE 3. TIMING OF EVENTS IN LATERAL STRIKES WITHOUT HEAD EXTENSION IN
4 Tx. BREVIPALPIS LARVAE.

Time (s) from strike initiation to
Opening of Complete Completion Completion
Larva palatal rotation of contraction of entire
No. brushes ofhead phase strike

2 0.002 0.012 0.020 0.058
2 0.004 0.012 0.022 0.064
4 0.010 0.012 0.024 0.076
5 0.004 0.008 0.012 0.076


thorax (Fig. 17). Thus, in the initial stages of the strike, the head led in amplitude the
degree of turning towards prey. An exception was larva 4, in which turning between
all body parts occurred at approximately the same time. During relaxation, because
contraction between head and thorax was usually greatest, the rate of return to the
rest position was usually higher between these parts than for the other body parts in
the early stages of relaxation (Fig. 17). The rates were quite similar in the later stages,
however.
Combining all the angle changes (Fig. 18) gave a clearer impression of the entire
strike as a coordinated action of the body and showed the uniformity of the movement
despite differences in the total angle of change. For reasons given below, it was evident
also that the total angular change undergone overall was considerably greater than
achieved by the head relative to the prey.
The small time increments during filmed strikes did not effectively convey how rapid
the body movements were and the extraordinarily high angular velocities achieved,
particularly during contraction. Larvae 4 and 5 serve to illustrate this point (Fig. 19).
Plotted separately for purposes of clarity, the data showed a peak head-thorax angular
velocity of about 3,8000/s for larva 4, and a remarkable level of almost 11,0000/s for larva
5 (Fig. 19). A very high velocity was attained also within the abdomen in larva 4, as
expected from the large angle of change that occurred in this individual (Table 4).
Somewhat lower peak velocities (4,000-5,000/s), recorded for changes between thorax
and abdomen, reflected the smaller angles of change between these parts (Table 4).
Although the measurements discussed above provided information about angle
changes between body parts, they did not describe the actual arc or rate of movement
of the striking head. This was because for a Toxorhynchites larva suspended freely in
water, only a portion of the turning angle generated between body parts is transferred
to the head. The position of the body at full contraction (dotted outline, Fig.~,0) com-

TABLE 4. TOTAL ANGULAR CHANGE WITHIN ENTIRE BODY, AND MAXIMUM ANGU-
LAR CHANGE BETWEEN BODY PARTS OF 5 TX. BREVIPALPIS LARVAE
MEASURED AT REST AND THEN AT FULL CONTRACTION IN STRIKE.

Total angular Maximum angular change between body parts
Larva change (0)
No. (0) Hd-Th % Th-Ab % Ab-Ab %

1 102 42 41.2 20 19.6 41 39.2
2 147 73 49.7 32 21.8 42 28.5
3 133 50 37.6 41 30.8 42 31.6
4 199 62 31.2 53 26.6 84 42.2
5 133 69 51.9 42 31.6 22 16.5










Insect Behavioral Ecology-'89 Linley


180

160

140

120

100
180

160

140

120

) 100
1t
1 180

> 160
0
.0 140
C
4 120
S180
a)
M 160
a)
) 140

120

100

80
180,


Larva 1


Larva 2








Larva 3


Larva 4


Larva 5


+00-*f\7. + ..l .+++.= +++++++ +
100 + +

80 I + I i I i
0 0 0 0 0 0 0 0 0O
O O O CV) l


Time (sec)

Fig. 17. Angles between body parts during complete strike sequences of 5 Tx.
brevipalpis larvae filmed at 500 frames/s. Larva 1 is depicted in Fig. 9, larva 4 in Fig. 16.










Florida Entomologist 73(1)


0

-40

-80

-120
0

-40


-120

-160
0

-40


-80

-120

-160
0


-40

-80


-120

-160

-200
0

-40

-80

-120


j2


March, 1990


Larva 1


- Larva 2


,
\ t


I I I i I i I


Larva 3

\







.-""
- x ****
0\ Larva 4
- '
\ *


op

- Larva 5

-\ o-*** -
,,


0 T- N M tO to r- t-
o 0 0 0 0 0 R o o

Time (sec)
Fig. 18. Total angle change within entire body during complete strike sequences of
5 Tx. brevipalpis larvae filmed at 500 frames/s. Larva 1 is depicted in Fig. 9, larva 4
in Fig. 16.












Insect Behavioral Ecology-'89 Linley


5000
3000
1000
-1000
-3000
-5000
-7000
-9000
-11000L
c
C


Larva 4

*eooo
r--------.--__- --


*-*= Hd-Th
o-o = Th -Ab
+-+= Ab-Ab

I 0 I I I I







0


0 0 0 0 0 0 00
dd dd d ddd


5000
3000
1000
-1000
-3000
-5000
-7000
-9000
-11000

5000
3000
1000
-1000
-3000
-5000
-7000
-9000


0 0 0 0 0 0 0
R R A q R
S0 0 0 0 0 0


Time (sec)

Fig. 19. Angular velocities between body parts during complete strike sequences of
2 Tx. brevipalpis larvae. Larva 4 is depicted in Fig. 16.


pared to the starting position (continuous outline), showed that most of the anterior
portion of the abdomen was displaced away from the direction of the strike, while the
tail was pulled towards it. The pivot point varied somewhat from strike to strike, but
was usually near the middle of the thorax (Fig. 20). Consequently, the striking head
travelled through an arc of only limited radius, with prey accessible within small lateral
distances.
Enlarged plots of the actual arcs (points at 0.002 s increments) showed the paths
followed by a central point on the head's front margin (Fig. 20). There was an apparent
difference between the paths for 2 strikes of more limited angle (larvae 1 and 2) com-
pared to larger angle strikes (larvae 3, 4, and 5). Forward movement of the head was
expected in larva 1 (Fig. 15), but the head of larva 2 also moved (Fig. 20). This move-
ment occurred early in the contractive phase (Fig. 20), suggesting that it resulted from
the usual hydrostatic mechanism. It is probable, therefore, that muscular actions effect-
ing head extension are initiated in every strike, but that their effects are offset in
strikes involving appreciable turning by other changes in body position that affect inter-
nal pressure. The three types of strike probably form parts of a continuous series in
which the degree of head extension diminishes with increasing lateral turning.


Larva 5












--0







+' I I I
\ P/



i/


11nnn


+

-

- +


I
I


i








Florida Entomologist 73(1)

Larva 1 f\


March, 1990





Sb


[ *.. 5mm


2mm b


b *' a


Fig. 20. a, Relative body position at initiation of strike (continuous line) and at fully
contracted position (dotted line) for 5 Tx. brevipalpis larvae; b, enlarged plots of each
true strike are, relative to the middle of the thorax, of a point on the mid-front margin
of the head (points at 0.0020 s intervals). Larva 1 is depicted in Fig. 9, larva 4 in Fig. 16.










Insect Behavioral Ecology-'89 Linley


From measurements of head position and angle in the strike arcs (Fig. 20), it was
possible to estimate the actual head speeds and angular velocities relative to the middle
of the thorax (Fig. 21). Generally, despite variation between individuals, increase and
decrease in head speed and angular velocity were fairly equally distributed in time
within the contraction phase. Head extension affected the rates in larvae 1 and 2 (Fig.
21) and in larva 5 the accelerative stage was of shorter duration than the decelerative.
Angular velocities and speeds attained by the head (measured at the middle of the
anterior margin) were remarkably high. Even in larva 1, which extended its head (Fig.
15), angular velocity reached nearly 5,000/s and the head speed almost 30 cm/s, while
the fastest rates observed, in larva 5, were 12,000/s and 70 cm/s.

Spatial distribution of prey

Points were plotted to show, at the instant of strike initiation, the closest part of
the prey's body in relation to the predator. Separate plots differentiated successful and
unsuccessful strikes (Fig. 22, 23). The distributions of these points with respect to
distance from the head (within each sector) were skewed, but could be rectified by
logarithmic transformation, allowing the mean prey distance and 95 percent confidence
regions to be superimposed (Fig. 22, 23).
With the capability of capturing prey only within a limited distance of the head, it
was not surprising to find that larvae initiated strikes only when prey approached very
closely. Even for unsuccessful strikes, the outer limit of the 95 percent confidence
region was no more than about a body length away from the head in every sector (Fig.
22, 23). Larvae of both species occasionally initiated strikes at more distant prey, as
shown by the few outlying points, one of which, in the case of Tx. brevipalpis, was 31.4
mm away (Fig. 23). Rather obviously (Fig. 22, 23), mean prey distance from the head
increased with increasing sector angle for all strike categories (Table 5), and was consid-
erably greater for unsuccessful strikes compared to successful. The confidence regions
showed that both species were successful in capturing prey only within limited cresent-
shaped regions centered on the head, and extending some distance along each side of
the abdomen (Fig. 23, 23). The region was more limited in Tx. brevipalpis (Fig. 23) and
extended posteriorly only to about the level of abdominal segment 1. Toxorhynchites
brevipalpis' capture area was particularly small, probably because this species tended
to reduce the prey angle by bending proportionately further towards its intended victims.
On the basis of all strikes made (Table 6), rather than just those that were measured
and plotted, the two species did not quite differ significantly with respect to the propor-
tions of strikes made in each sector (X2 = 9.424, d.f. = 5, P << 0.1). Nonetheless, data
for the first three sectors (0-75) combined and compared to the last three (76-180)
showed that a smaller portion (135:176 = 40.2%) of the Tx. amboinensis strikes was to
prey at the lower angles, as compared to Tx. brevipalpis (107:100 = 51.7%). This differ-
ence was not unexpected in view of Tx. brevipalpis' tendency to bend more towards
prey, and it is probably important even though the difference did not quite attain the
5 percent level of significance, because the two species did differ with respect to the
proportion of successful strikes in each sector. In Tx. amboinensis, the proportions of
successful as opposed to unsuccessful strikes (Table 6) were significantly different be-
tween sectors (X2 = 12.597, d.f. = 5, P < 0.05), with higher success percentages in the
three lower angle sectors, whereas the proportions were not significantly different in
Tx. brevipalpis (x2 = 3.277, d.f. = 5, n.s.). However, the success levels for Tx. brevipal-
pis were distinctly higher than for Tx. amboinensis in all sectors except one (Table 6)
and, on the basis of the data for all strikes combined, Tx. brevipalpis was significantly
more successful (X2 = 15.132, d.f. = 1, P < 0.001). Of Tx. brevipalpis' strikes, 148/207
(71.5%) resulted in prey capture as compared to 170/311 (54.7%) in Tx. amboinensis.









Florida Entomologist 73(1)


5000 r Larva 1 300

2500 / 150

0 I I I 0I I I I
12500r Larva 2 -750

10000 = av 600

7500 / \ o= hs 450

5000- 300

2500 -150

0 I 0
1 7500 _" Larva3 450
5000- o-0;0 -300

2500-0 150

0 0
> 12500 Larva 4 750 .



C 7500 0 -450 )
0-o-/ I
5000 / 300

2500 / -150

0 0
12500 Larva 5 -750

10000- 0 -600

7500- -450

5000- -300

2500 150

0 1 1 d 0

dd ddddddooodd
Time (sec)
Fig. 21. Angular velocity and head speed during the true strike arc (see Fig. 20) of
5 Tx. brevipalpis larvae. Larva 1 is depicted in Fig. 9, larva 4 in Fig. 16.


March, 1990













Insect Behavioral Ecology-'89 Linley


Tx. amboinensis




. . . .. ................. ........ ....
*. *** .





:.-
**o *'1 *". "o *







**








successful








5mm


unsuccessful


Fig. 22. Spatial distributions of prey (nearest part), mean prey distance (line with
dots widely spaced), and 95% confidence regions (line with dots closely spaced), for
successful and unsuccessful strikes of Tx. amboinensis.


""""""""'"'
"` ""''~
''" ""
'''
'''


i.












Florida Entomologist 73(1)


Tx. brevipalpis


successful


5mm


.-31.4


16.2 .


unsuccessful

Fig. 23. Spatial distributions of prey (nearest part), mean prey distance (line with
dots widely spaced), and 95% confidence regions (line with dots closely spaced), for
successful and unsuccessful strikes of Tx. brevipalpis.


..
~
..





.




~

,...


March, 1990











Insect Behavioral Ecology-'89 Linley


TABLE 5. MEAN DISTANCES OF PREY FROM HEAD, BY SECTOR, FOR SUCCESSFUL
(S) AND UNSUCCESSFUL (U) STRIKES, AND ALL STRIKES (A) COMBINED.

Tx. amboinensis Tx. breviopalpis
Sector Mean distance (mm) Mean distance (mm)
() s u A s u A

0-15 1.36 2.71 1.75 1.49 2.38 1.72
16-45 1.63 3.57 2.68 1.64 2.94 2.07
46-75 2.29 3.10 2.70 1.90 2.87 2.26
76-105 2.19 3.65 2.84 2.11 4.06 2.64
106-135 3.10 4.19 3.63 2.84 4.74 3.51
136-1801 4.38 4.18 4.26 2 6.50 6.08

'Sectors 136-165 and 165-180 combined
2One observation only



Prey angle and strike angle

Regressions of prey angle on strike angle for successful strikes were linear (Fig.
24), indicating that both predators pivoted through strike angles that were a fairly
constant proportion of prey angle for all prey angles. Strikes were not made at propor-
tionately greater angles as prey angle diminished. The proportions are given by the
regression coefficients, so that Tx. amboinensis pivoted about 81% and Tx. brevipalpis
about 73% of the prey angle. The regression coefficients did not different significantly.
These results indicated that the Toxorhynchites larvae were able to estimate prey
angle with considerable accuracy and match the strike angle accordingly. Some degree
of error was implied in the scatter of points about the regression line in each case.
However, the coefficients of determination (r2) showed that prey angle accounted for
74.4% of the variation in strike angle for Tx. amboinensis, and 70.7% for Tx. brevipal-
pis. Moreover, the prey angle/strike angle relationships for unsuccessful strikes were
very similar (Fig. 25), suggesting that accuracy with respect to angle was not due to
curtailment of the strike arc when the predators's head or mouthparts made sensory
contact with the prey. Some contact may have occurred in unsuccessful strikes, but
probably not in a large portion. The regression coefficients were lower in both cases
(but not significantly so) and the scatter of points somewhat greater (Fig. 25), with r2
values of 62.7% for Tx. amboinensis and 51.9% for Tx. brevipalpis. Underexecution
and inaccuracy in the strike angle may have reduced success in capturing prey to a
limited extent.


TABLE 6. NUMBERS OF SUCCESSFUL (S) AND UNSUCCESSFUL (U) STRIKES, TOTAL
(T) STRIKES, AND PERCENT SUCCESSFUL STRIKES (%S) IN EACH SECTOR.

Sector Tx. amboinensis Tx. brevipalpis
(0) s u T %s s u T %s

0-15 39 17 56 69.6 24 11 35 68.6
16-45 20 18 38 52.6 34 8 42 81.0
46-75 38 13 41 68.3 22 8 30 73.3
76-205 36 38 74 48.6 29 16 45 64.4
106-135 31 37 68 45.6 21 9 30 70.0
136-180 16 18 34 47.1 18 7 25 72.0









Florida Entomologist 73(1)


y = 3.63 + 0.81x


180-

IS-

144-



108-



72-


March, 1990

Tx. amboinensis

*


IIt# *


STx brevipalpis
Tx. brevipalpis


y = 13.57 + 0.73x


0


0
* 0


* .


. *


r;*O 0
8, 0


I I I I I I I I I I


0 36 72 108 144 180

Prey angle (0)

Fig. 24. Relationship between prey angle and strike angle for successful strikes.


* *Oy



i0
** *
6 *


.e
0=
0




0)

0)
lid


36-



n.


180-



144-


108-



72-



36-


Al


1


W











Insect Behavioral Ecology-'89 Linley


Tx. amboinensis


y = 15.51 + 0.79x


180-




144-




108-



72-




36-




0-


180-




144-




108-




72-


0O
0(
0
<>0
.3


* 0
0 00~4~~0 0
0


1 ITx brevipalpis
Tx. brevipalpis


y = 30.40 + 0.61x


0 0


0
0


0 0


* 00.

%
.ol 01--


10 0
J*0* 0
i I I I I I
0


O
*


0 36 72 108 144 180

Prey angle (0)

Fig. 25. Relationship between prey angle and strike angle for unsuccessful strikes.


* 0


0



0
0 S


0

a0

C/)

0)


36-



n-


v









48 Florida Entomologist 73(1) March, 1990


DISCUSSION

It has been stated many times in the literature (e.g. Rubio et al. 1983, Steffan &
Evenhuis 1981, Russo 1986) that Toxorhynchites larvae are passive ambush predators
that wait for prey rather than actively seek it. This characterization is only partially
accurate. The larvae certainly do not move constantly through their habitat seeking
prey, but the degree to which larvae actively respond to prey depends on sensory cues
indicating the prey's location, whether beneath or on the surface. In container habitats,
favored by these insects, potential prey organisms also living in the water are usually
few in number and occupy the same relatively limited volume. These are prey that
probably will remain in the container for some time and,. because of their own activity,
will probably encounter the yToxorhynchites larvae. For such prey it is apparently
more energy efficient for the Toxorhynchites to remain inactice until the prey ap-
proaches closely, and then respond so as to optimize the chance of capture. Logic sup-
ports this tactic because a container of water provides a 3-dimensional universe in which
considerable search energy would be required for each prey encounter, especially when
prey are active and limited in number. Even with sensory mechanisms capable of fixing
the position of prey, considerable probability of error remains because the prey can
move or escape in any direction. When potential prey is on the surface, however, the
situation is different in two important ways. Firstly, the prey may have fallen onto the
surface and may soon escape, or it may be there temporarily for its own purposes. This
food, therefore, is much less likely to move itself within range by chance, and it will
probably be lost unless the predator acts. Secondly, surface prey is moving in a 2-di-
mensional sense, where sensory mechanisms for localizing the prey's position and
locomotory ones for approaching it must deal with reduced potential for erro. Thus, as
reported by Paine (1934) and Breland (1949), and as I shall describe quantitatively in a
later paper, hungry Toxorhynchites larvae will almost always turn and swim towards
surface prey, provided the latter creates surface disturbances. The fact that Toxorhyn-
chites larvae in the laboratory have been maintained and studied almost exclusively
using other mosquito larvae as food, has overemphasized the perception of them as
sessil, almost totally inactive predators.
Given that the Toxorhynchites larva must rely on the movement of subsurface prey
and chance to bring a potential victim close enough to attack, then it is very important
to optimize the probability of successful capture, especially with such a limited strike
range. Bending towards prey is a device for increasing the probability that the prey
will enter the capture arc. It may immediately cause prey to come within strike range.
If not, it at least reduces the distance of the head from the prey, which in turn causes
the angle subtended at the prey by the strike arc to be increased, so that the prey is
less likely to pass outside capture range. A less obvious factor that leads to the same
result involves the shapes of the capture regions (Fig. 22, 23), which are such that more
of each region is presented to the prey when a bend is executed. For Tx. amboinensis,
bending also is important for increasing the likelihood that the prey will enter the strike
arc at a relatively small angle, where more of the strikes are successful. This consider-
ation presumably would not be important for Tx. brevipalpis, which, on the basis of the
present data, showed no difference in the proportion of successful strikes between
sectors. However, the overall level of success in Tx. brevipalpis was significantly higher
than in Tx. amboinensis (71.5% c.f. 54.7%) and, although the difference was not quite
significant, Tx. brevipalpis made proportionately more strikes to lower angles (0-75)
than Tx. amboinensis (51.7% c.f. 40.2%). In view of the evidence that Tx. brevipalpis
bends towards prey to a greater degree, as well as more rapidly, it seems by this
activity to have improved its overall success rate for strikes and eliminated differences
in proportional success rate between sectors.











Insect Behavioral Ecology-'89 Linley


While there are advantages to bending, there are also possible disadvantages. One
involves the generation of disturbances in the water that might precipitate avoidance
reactions by prey. It is perhaps for this reason that bending takes place, particularly
in Tx. brevipalpis, before prey approaches extremely closely, and is executed smoothly
and slowly, rather than suddenly. Bending may also be important in bringing sense
organs involved in strike initiation, which probably include mechanoreceptors on the
antennae (Jez & McIver 1980), to a more frontal position, from which the prey's position
may be estimated more accurately.
Confusion in the past over which appendages are used to seize prey has arisen for
two reasons. Firstly, the extreme rapidity of the strike sequence (prey is seized in about
0.02 s) renders it impossible to resolve events with the naked eye or under a stereomic-
roscope. Secondly, and perhaps the factor that has contributed most of the confusion,
the appendages used to handle prey are switched as soon as prey is captured. The
palatal brushes are certainly the capture instruments, but prey is then immediately
engaged by the mandibles. The palatal brushes release their hold and return to their
resting position, and are never seen holding or manipulating food as it is consumed.
Striking at prey is a vigorous action and for its brief duration is undoubtedly energy
expensive. Furthermore, individual prey may only rarely come within reach. These
factors exert strong selective pressure in favor of strike behavior that optimizes high
success/failure ratio. Ensuring success is one reason why strikes are initiated only when
prey approaches very closely, almost always within the region where capture probabil-
ity is high. For Tx. amboinensis, for example, 85.5% of all strikes were at prey within
the 95% confidence zone for successful strikes, with a comparable figure of 77.8% for
Tx. brevipalpis. The confidence zones are themselves limited because prey can only be
reached at small distances and captured only within the overlapping closure arcs of the
palatal brushes as they move through the strike arc of the head. In addition to any
direct energy consequences of failed strikes, the sudden movements they involve almost
certainly have the effect of warning some prey organisms of the predator's presence
and precipitating escape reactions.
The need to offset escape reflexes in the prey is undoubtedly the major reason why
the strike is so fast. Very rapid movement in water, however, because of friction and
drag, imposes limitations that doubtless have been important in shaping the strike
appendages and behavior in Toxorhynchites larvae. A feature of the palatal brushes
that seems initially surprising is that they are so small. Longer brushes spread to a
greater angle would seem more effective for securing prey. Such brushes, however,
would probably cause excessive drag, especially during large angle lateral strikes,
where head speed may reach 70 cm/s. Even with the brushes in their observed form,
frictional forces on them and on the head itself have probably influenced behavior.
During lateral strikes, for example, rotation of the head takes place very early in the
strike, within 0.012 sec, and probably permits the head to bend to a more acute angle
relative to the thorax and, since the head's frontal aspect now leads along the path of
the strike arc, offer less resistance. Also, it is in the rotated head position that the
narrower aspect of each element of the palatal brushes is aligned in the plane of the
strike arc.
Although the majority of strikes are successful, some are missed, probably for the
following reasons. In many instances, strikes are made before the prey is close enough
and the larva simply cannot reach its intended victim. There is some evidence that
unsuccessful strikes are executed to a somewhat smaller portion of prey angle than
successful ones. Thus, "understriking" may be a factor, as well as greater inaccuracy
in matching strike angle to prey angle. It is important to remember, also, that the
palatal brushes can capture only if part of the prey intersects the arc and plane of the
brushes as they close. Capture is much more likely if the prey, or part of it, is aligned










Florida Entomologist 73(1)


at right angles to the closure plane of the brushes. It is quite often observed that when
a larva misses a strike, it will immediately strike again if the prey is still close, some-
times several times. Repeat strikes are often successful and function as a behavioral
means of compensating for the limitations of the capture system.

ACKNOWLEDGMENTS

I thank D. Duzak for her help in maintaining laboratory colonies and for some
assistance during the analysis. Bonnie Pattok prepared the illustrations and Tomas
Zoebisch translated the abstract. This paper is Institute for Food and Agricultural
Sciences, University of Florida Experiment Stations Journal Series no. R-00417.

REFERENCES CITED

BANKS, C. S. 1908. Biology of Philippine Culicidae. Philippines J. Sci. Sect. A. 3:
235-258.
BRELAND, 0. P. 1949. The biology and the immature stages of the mosquito,
Megarhinus septentrionalis Dyar & Knab. Ann. Ent. Soc. Am. 42: 38-47.
CRANS, W. J., AND M. E. SLAFF. 1977. Growth and behavior of colonized Toxorhyn-
chites rutilus septentrionalis. Mosq. News. 37: 207-211.
COLLEGE, W. R. 1911. Notes on a brush-tongued mosquito. Proc. R. Soc. Queensland
23: 121-130.
FRANK, J. H., G. A. CURTIS, AND G. F. O'MEARA. 1984. On the bionomics of
bromeliad-inhabiting mosquitoes. X. Toxorhynchites r. rutilus as a predator of
Wyeomyia vanduzeei (Diptera: Culicidae). J. Med. Eng. 21: 149-158.
FURUMIZO, R. T., AND A. RUDNICK. 1978. Laboratory studies of Toxorhynchites
splendens (Diptera: Culicidae): biological observations. Ann. Ent. Soc. Am. 71:
670-673.
GOMA, L. K. H. 1964. Laboratory observations on the influence of illumination on the
predatory habits of Toxorhynchites larvae. Ann. Trop. Med. Parasitol. 58: 350-
354.
GREEN, E. E. 1905. On Toxorhynchites immisericors (Walker), the elephant mos-
quito. Spolia zeylan. 2: 159-164.
HARBACH, R. E., AND K. L. KNIGHT. 1980. Taxonomists' glossary of mosquito
anatomy. Plexus Publishing, Marlton, New Jersey. 415 p.
JEZ, D. H., AND S. B. MCIVER. 1980. Fine structure of antennal sensilla of larval
Toxorhynchites brevipalpis Theobald (Diptera: Culicidae). J. Insect Morphol. &
Embroyl. 9: 147-159.
LOUNIBOS, L. P. 1979. Temporal and spatial distribution, growth and predatory be-
haviour of Toxorhynchites brevipalpis (Diptera: Culicidae) on the Kenya coast.
J. Anim. Ecol. 48: 213-236.
LouNIBOS, L. P., J. H. FRANK, C. E. MACHADO-ALLISON, P. OCANTO, AND J. C.
NAVARRO. 1987. Survival, development and predatory effects of mosquito lar-
vae in Venezuelan phytotelmata. J. Trop. Ecol. 3: 221-370.
MUSPRATT, J. 1951. The bionomics of an African Megarhinus (Dipt., Culicidae) and
its possible use in biological control. Bull. Ent. Res. 42: 355-370.
PADGETT, P. D., AND D. A. FOCKS. 1980. Laboratory observations on the predation
of Toxorhynchites rutilus rutilus on Aedes aegypti (Diptera: Culicidae). J. Med.
Ent. 17: 466-472.
PAINE, R. W. 1934. The introduction of Megarhinus into Fiji. Bull. Ent. Res. 25:1-32.
RUBIO, Y., D. RODRIGUEZ, C. E. MACHADO-ALLISON, AND J. A. LEON. 1980.
Algunos aspects del comportamiento de Toxorhynchites theobaldi (Diptera:
Culicidae). Acta Cient. Venezolana 31: 245-351.
Russo, R. 1986. Comparison of predatory behavior in five species of Toxorhynchites
(Diptera: Culicidae). Ann. Ent. Soc. A. 79: 715-722.


March, 1990










Insect Behavioral Ecology-'89 Lloyd


STEFFAN, W. A., AND N. L. EVENHUIS. 1981. Biology of Toxorhynchites. Annu.
Rev. Ent. 26: 159-181.
TRPIS, M. 1973. Interaction between the predator Toxorhynchites brevipalpis and its
prey Aedes aegypti. Bull. Wld. Hlth. Org. 49: 359-365.






FIREFLY SEMIOSYSTEMATICS AND PREDATION:
A HISTORY1

JAMES E. LLOYD2
Department of Entomology and Nematology
University of Florida
Gainesville, Florida 32601-0143

ABSTRACT

Modern firefly systematics began when it was realized that the luminescent patterns
emitted by flying, mate-seeking males provided an invaluable clue for recognizing
species in nature. An approach in systematics giving special attention to these signals
("semios") could be called semiosystematics, the focus thus being on the coded patterns
that the insects are using for "interbreeding" (gene flow), and/or non-interbreeding
(isolation).
In North America, Photuris spp. females prey upon signaling males of other species
by mimicking the mating signals of their preys' females, and by directing aerial attacks
toward the males' signals. Countermeasures against these attacks that have evolved in
the signals of prey species, and aspects of the signals of the predator species that have
evolved as a result of the mimicry predation, may be dominant features in the signaling
of many of these fireflies. The flash signals that firefly systematists have used, and in
particular those that have given difficulty for decades, may often be (or historically have
been) closely connected with this strong ecological force, predation.

RESUME

La sistemAtica modern de las luci6rnagas comenz6 cuando se dieron cuenta que los
patrons luminiscentes emitidos por machos en vuelo en busca de la hembra, dan una
pista valiosa de c6mo reconocer las species en la naturaleza. Un aproche en sistemAtica
dandole una atenci6n especial a estas sefiales ("semios") se pudiera llamar semiosis-
temdtico, el foco siendo entonces el c6digo de patrons que los insects usan para "inter-
cruzarse" (flujo de genes), y o el no-intercruze (aislamiento).
En Norteambrica, hembras de Photuris spp. depredan los machos de otras species
que hacen seiias, imitando la serial de las hembras press, y dirigiendo ataques a6reos
contra la serial de los machos. Contramedidas contra esos ataques que han evolucionado
en la serial de la especie depredada, y los aspects de las sefiales de la especie depredad-
ora que ha evolucionado como el resultado de la mimica de depradci6n, pudieran ser
razgos dominantes en las sefiales de muchas de estas luci6rnagas. Las sefales de destello
que los sistematicos han usado, y en particular aquellas que han dado dificultad por
d6cadas, pudieran haber sido (o historicamente lo han sido) cercamente asociadas con
esta poderosa fuerza ecol6gica, la depredaci6n.









52 Florida Entomologist 73(1) March, 1990

An insect systematist or naturalist functioning as one, who has worked on a group
of organisms for a while, after first inheriting and learning the standard techniques,
procedures, and conceptual beliefs for the group begins to modify and develop them,
and discover some new ones. With luck he (= he or she) may discover or identify some
key technique or ecological element that seems at the heart of taxonomic resolution or
a new understanding of the taxon. If luckier, he may be correct. I cannot imagine a
cricket taxonomist worth a chirp being without a taperecorder and audiospectrograph
since about 1950 (I have heard that a group of Michigan orthopterists once discussed
the possibility each jocularly encouraging the next to a higher level of seeming
impracticality that each cricket specimen should have its song preserved with it,
perhaps in a tiny case on the insect pin below the corpse, maybe with a string that could
be pulled to hear the song the insect sang in life, and certainly this should be required
for all Holotype specimens). With time, perhaps decades, a systematist may even de-
velop new and idiosyncratic notions of speciation and species concepts for "his" animals.
In my long hunt of North American fireflies, I inherited, improved, and added some
techniques, and developed a feeling for the continuity and epigenesis of discovery and
contribution in the field. I think I have identified a key factor for understanding the
systematics of these insects, one that has been responsible for much confusion.

Firefly Semiosystematics: Beginning and Theory

Modern study of firefly systematics began with Frank A. McDermott, a nonprofes-
sional (i.e., unpaid) lampyridologist who spent most of his working life as a chemist. He
published his first observations on firefly flashing behavior in 1910. Among his last
works, i.e., published in the mid-1960s, were a catalogue of the nearly 2,000 described
species of Lampyridae of the world (the notes for which fill a large carton of fileboxes),
and the Taxonomy of the Lampyridae (1964), which have a worldwide taxonomic synop-
sis and overview of the family.
In the 1910s McDermott reported and urged that if you watched fireflies closely you
would see that the flashes were sexual signals. He found how the basic signal system
works in flashing American fireflies, and in particular in the common Photinus pyralis
(L.). The fact that Osten-Sacken (1861) had observed and connected mating and flashing
in P. pyralis decades earlier, or that the significance of luminescence in sexual interac-
tions in European glowworm species had been appreciated centuries earlier (early
1600s?), cannot diminish the significance of McDermott's independent discovery; he
rediscovered the sexuality of pyralis' luminescent signals, but only started there. He
found that the student could locate and identify male fireflies of different species in the
field by the species-specific flashing patterns they emitted as they flew about their
habitat seeking mates (Fig. 1). He found that he could simulate the flashes of male and
female fireflies with matches and electric lights and experimentally determine signifi-
cant elements. And, he discussed how different patterns and timings could prevent the
intermixing of species.
McDermott's firefly taxonomy work thus began with locating, recognizing, dis-
criminating and experimenting with mating signals, and he thought about species and
"reproductive isolating mechanisms." He put his conclusions into practical application
by making what appears to be the first formal taxonomic decision based on mating
signals; he took Photinus castus LeConte out of synonomy because he believed that its
mating signals were different from those of the species it had been lumped with (McDer-
mott 1912). (Though castus is now again synonomized, the biological significance of the
castus "morph" is yet to be understood; see Green 1956: 575). When you read McDer-
mott's 1912 paper you realize that he had a feeling for what amounted to biological
species and sibling species, long before biology at large did, though another solitary









Insect Behavioral Ecology-'89 Lloyd


Cif


Sec.















ud'r~ ~ ~







ZZd LZZ


Photinus
pyralis.





Photinus
consa eus.
n. sp.
barberi"



Photinus
cactus.






Photinus
Seintillans.
C




Photinus
marginellus.


P raoctomena


Leco a
lu era.


Fig. 1. First known firefly flash pattern chart, from McDermott (1914).
Nomenclatural changes in keeping with current understanding (Green 1956, 1957; Lloyd
unpub.) are indicated. Horizontal axis is time, as indicated; vertical axis, intensity.










Florida Entomologist 73(1)


entomologist, Benjamin Dann Walsh, had tracked this swampy ground nearly 50 years
earlier (1864); and, Samuel Scudder had carefully listened to and scored (musically
speaking) the songs of several Orthoptera species [e.g., 1874; note also B. B. Fulton in
Alexander (1964), and the latter's discussion of the use of cricket songs in their classifi-
cation].
Such a point of departure and orientation in the literal and figurative pursuit of
firefly species, in nature and thought, and the tendency to be more than attentive to
and even favor the evidence of mating signals when it conflicts with that from morphol-
ogy, is what I mean by semiosystematics (semio from the Greek word for sign or signal).
It is not the last word, but a first word; a tentative and hypothetical word; a cautionary
word and watch word, in a study of firefly species. At the end of the successful chase,
its etymology and entomology will be understood, or so a semiosystematist hopes. The
rationale for such an approach is obvious, recalling that mating signals actuatee" identifi-
cation, copulation, and insemination, hence flowing of genes and defining of biological
species: which definition may not be preferred or best, but elements and the error of it
must be part of finding a satisfactory understanding of what firefly "species" are in
nature.
To summarize, McDermott showed that one pursues and watches firefly flashes to
pursue and understand species. At night in the dark, as an avocation, with primitive
technology (including transportation to study sites), with many yet unknown and mostly
poorly or inadequately described, nominate species, McDermott saw through a confu-
sion of flashing patterns in the field, and put together the working foundation for
semiosystematics.

Photuris Semiosystematics: Application and Technique

McDermott recognized that different Photinus and Pyractomena species emit differ-
ent flash patterns, and applied this in an example case of conflicting evidence, but he
never acted on his theory in the case of Photuris pennsylvanica (DeGeer). For many
decades it had been assumed that the genus Photuris had limited representation in the
North American fauna. America's Dean of Coleoptera John L. LeConte (1881:37,
1883:208) had concluded that except for Photuris frontalis LeConte and another from
the Near West, all of the Photuris fireflies in the United States and Canada were P.
pennsylvanica.3 McDermott observed that the presumptive Pennsylvania Firefly emit-
ted distinctively different flash patterns, but he left it without further comment. It
should be noted that a little earlier Henry W. Wentzel (1896) of Philadelphia, "One of
the best collectors of Coleoptera in the country" (Smith 1899:730), had made the
generalization that each firefly species has a different way of flashing (:294), and had
distinguished two species of Photuris frontalis (sic), on the basis of unspecified flashing
differences (:296).
In the 1920s Herbert S. Barber, a beetle taxonomist at the National Museum, and
long-time correspondent of McDermott, chased fireflies seriously as a (job-related)
hobby. He used the semiosystematic approach on LeConte's ubiquitous P. pennsyl-
vanica, and recognized several unnamed species (Barber, in Barber and McDermott
1951). His chart of firefly flash patterns has since been often used to illustrate how a
study of mating signals can disclose the presence of sibling (=cryptic) species (Fig. 2)4.
From his experience with Photuris, Barber particularly stressed the importance of
repeating observations on flashing patterns, and the preservation of flash (pattern)-
voucher specimens. He found that with carefully collected series of such vouchers,
specimens that had emitted various discrete and distinctive patterns could be distin-
guished morphologically of course this was not critical, but it was what was needed
for building confidence in such an approach. Barber gave some of his species scientific


March, 1990











Insect Behavioral Ecology-'89 Lloyd 55

12C1E3E SUOM 13.

I 2tWC.ES 5 9 0 2 3

II I I I I I CESI I I I




_____T 3 1,.I I I i II
IHio. figi
VERSICMLO& IPLI |












Fig. 2. Flash pattern chart from Barber's (1951) classic, often- cited study. As in
Figure 1, axes indicate time and intensity.


names descriptive of their distinctive flash patterns: Photuis tremulans, P. lucicres-

Thus, Barber began to pick at and understand the Photuris species problem that
LeConte could not see in dead specimens, and apparently had not looked for on his
several trips and expeditions, and that McDermott had observed but silently put aside.
Barber began to understand, and to "split up" P. pennsylvanica as his mentor and
fellow coleopterist E. A. Schwarz (Fig. 3) much earlier had predicted someone would
do (McDermott, in B. and M. 1951: iv). His work was certainly more exciting and better
empirical documentation and application of what McDermott The Theorist had thought
about. When it was ultimately published, posthumously and nearly three decades after
his major field work was done, it was current, but not prescient as it could have been,
for the New Synthesis in evolutionary and systematic biology was in full charge by that
time (Mayr 1982).
Of equal importance, a posteriori, because it led to the next step in the development
and understanding of Photuris systematics, was Barber's observation that males of
IQ LUCIKRESCEJS





























some PhotLuis species actually do emit more than one flash pattern (Fig. 2, nos. 10 &
11; Barber, in B. and M. 1951: 6-8). This must have been of as much concern to Barber














as apparent multiple patterns in P. pennsylvanica may have been to McDermott, but
neither appears to have left us his inner suspicions or thoughts on the matter, at least
4* TNEMUILAN Ij













Fig.to my knowledge none was published except for Barber's br(1951) classic, often- cited study. As in
Figure 1, axes indicate time and intensity.












musings" on the legitimacy of usr listing the flash patterns to distinguish species (Barber,
cens.





















in B. and M. 1951: 8).
Thus, Barber begattern to changing was and is a significantPhotuis species problem. Aye, now there's
LeContea rub, for if species-ee inspecific flash patterns are "only appareproductive isolating mechanis
several trips was believed expedifor a long time, insuring that McDermott had observe uncomplicated but small-brained put aside.
Barber began to understand, and to "split up" P. pennsylvanica as his mentor and
fellow coleopterist E. A. Schwarz (Fig. 3) much earlier had predicted someone would










do (McDermott, in B. and M. 1951: iv). Hises, what could be the utility of switching among multi-
ple, distinctive pdocumentatterns? Why should a species need of what McDermore than one pattern? It had thought
about. When it was ultimately published, posthumously and nearly three decades after
his major field work was done, it was current, but not prescient as it could have been,
for the New Synthesis in evolutionary and systematic biology was in full charge by that
time (Mayr 1982).
Of equal importance, a posterior, because it led to the next step in the development
and understanding of Photuris systematics, was Barber's observation that males of
some Photuris species actually do emit more than one flash pattern (Fig. 2, nos. 10 &
11; Barber, in B. and M. 1951: 6-8). This must have been of as much concern to Barber
as apparent multiple patterns in P. pennsylvanica may have been to McDermott, but
neither appears to have left us his inner suspicions or thoughts on the matter, at least
to my knowledge none was published except for Barber's brief "concerned, rhetorical
musings" on the legitimacy of using the flashing patterns to distinguish species (Barber,
in B. and M. 1951: 8).
Photuris flash pattern changing was and is a significant problem. Aye, now there's
a rub, for if species-specific flash patterns are "only reproductive isolating mechanisms"
as was believed for a long time, insuring that their uncomplicated, small-brained posses-
sors get mates of the same species, what could be the utility of switching among multi-
ple, distinctive patterns? Why should a species need more than one pattern? It was here










Florida Entomologist 73(1)


Fig. 3. H. S. Barber (left) and his mentor E. A. Schwarz in early 1900s, dining while
on a collection trip to Plummer's Island in the Potomac River, at Washington, D.C.
Photo from the USNM Archives, courtesy of T. Spilman. Barber was a photographer
and may have taken the photo; there does not appear to be a third place-setting, and
Barber's pose suggests that he may have tripped the shutter timer, rushed to his chair
and "froze."

that the Father and Great Uncle of firefly semiosystematics left the problem to future
lightningbug chasers.

Predation By Photuris

Fireflies Barber passed along a critical anecdote on the phenomenon that I am
suggesting may be the key to understanding the signaling behavior of American fire-
flies. He observed female Photuris fireflies flashing correct "sexual" answers to the
flash patterns of males of other species (Barber, in B. and M. 1951: 9-10). Since he and
others had found female Photuris eating males (only) of other luminous species (Fig.
4), Barber was cautiously suspicious that the females preyed on the males by luring
them via signal mimicry [see Lloyd (1984a) for review, including early history of obser-
vations of Photuris predation, and references, analysis, and tabulation of observations
of aggressive mimicry]. Predation by Photuris females is an insidious affront to those
who wish to see beauty and serenity in the silent, twilight sparkling of these gentle,
leather-winged beetles. It is aimed at the signals that the fireflies would use for mate
finding, mate recognizing, and gene flow: the signals that a naive semiosystematist
believes will enable him to resolve firefly species problems "simply."


March, 1990










Insect Behavioral Ecology-'89 Lloyd 57


Fig. 4. Photuris species "B" female eating a male of Pyractomena angulata, on
campus of U. of Florida, Gainesville. Her males mimic the flash pattern of male P.
angulata (Lloyd 1980).










Florida Entomologist 73(1)


The Photuris female, again collectively speaking, though it is technically improper
to use a generic name in such fashion (Blackwelder 1968: 51), attacks signaling males
in two main ways. In Barber's tactic, she pretends to be a female of the targeted male's
species and with false signals lures him to where she is perched (McDermott was also
long suspicious that this occurred, pers. comm. ca. 1965, see also McDermott 1917:60).
In this predation the predators have gotten inside the coded signal programs of their
prey, and must cause eternal pressure for arms races as they track changes that prey
species must evolve in their signaling behavior to avoid them.
In the second tactic, which the females sometimes use in conjunction with their
aggressive mimicry, they launch aerial, hawking attacks, guided by the light-emissions
of flying males (Figs. 5-7; Lloyd & Wing 1983). Obviously, flashed signal patterns that
let males transmit their message and then hide in the dark should be safer than long
patterns that offer easy targets, and allow attackers to approach from a distance. How
could luminescent signals not be affected by the actions of these females, or at least,
how could a semiosystematist not theorize that they were?
On Technique: Predation by semiosystematists. The physical pursuit of firefly
species by semiosystematists is predation, and because it is fundamental to semiosys-
tematic practice it is worth mentioning. The semiosystematist scouts, locates, identifies,
and tracks; lures, nets, traps, and trolls for his quarry in ways dependent upon precise
and detailed knowledge of the most intimate sexual habits and entreaties of his prey.
While in the dark, he quantifies the behavior (see Barber, in B. & M. 1951: 10 for early
technology), even borrows techniques of the stalking sportsman with gun (photo-multip-
lier) and fishing rod (with light-emitting diode lures),for the pursuit of individually
observed and significant trophies called behavior- or semiosystematic-voucher speci-
mens. (When in restricted areas such as National Parks, having not planned ahead for
collecting permits, it is possible that he circumspectly poaches.) Vouchers are carefully
curated with genitalia extruded, colorfully labelled, identified and displayed in "voucher
cabinets": their behavior and capture are described in detail in field-journals which,
with recordings and specimens, are cross-referenced and indexed.


Consequences Of Predation For Firefly Systematics

Given: 1) that firefly signals are easily simulated by "foreigners" with light organs
(and LED's), and 2) that most luminescent fireflies in the dark have little information
for decision-making other than light flashes, THEN, their luminescent behavior must be
expected to have changed, as they evolved escapes. Through comparative studies there
should be little difficulty in finding illustrative examples, and in experimentally demon-
strating their feasibility. Tests with LED's on fishpoles, proved the "obvious," that
continuous glows make better targets than intermittent, short flashes, and that com-
plete darkness is safest (Lloyd & Wing 1983, see Figs. 4-6)5.
Here are some possible predation-systematic connections, keeping in mind that the
interests of systematists range from the location, identification and capture of specimens
mentioned above, through the understanding of the evolution of higher taxonomic
categories, to the natures and evidences of evolutionary tempo and mode (with driving
mechanisms).



Species of North American Photinus for the most part fit comfortably into apparent
phylogenetic groups, each of which is defined by a characteristic male genitalic form
[Green (1956: 561-562); note that beetle taxonomy in McDermott's early time did not
know of this character, and without it many nearctic Photinus and Pyractomena are
sibling species4. Characters in addition to genitalia are often useful, but many are


March, 1990










Insect Behavioral Ecology-'89 Lloyd 59

















































Fig. 5. Photuris species "D" female, with feet stuck on decoy (light-emitting bead)
she has attacked. Wires to fishpole and electronic controls are above.

subtle and require initial separation by genitalia for appreciation. Today an experienced
Photinus pin-pusher and connoisseur is not likely to misidentify this taxon even as it is
now, somewhat broadly constituted. Flash patterns within Photinus (genitalic) groups
also show relationship (Green 1956: 563; Lloyd 1966: 77, 1984b).









Florida Entomologist 73(1)


Fig. 6. Photuris "D" female eating a prey male that had been attached to an LED
target, indicating that females that are attracted to "flying" LEDs are in predacious
mode (or switch to it after striking the light).


March, 1990










Insect Behavioral Ecology-'89 Lloyd 61

















































Fig. 7. Two Photuris "D" females simultaneously attacked the target LED with
impaled male. One attacker has nearly severed the head from the other (above). The
head of the impaled prey male is seen below.










62 Florida Entomologist 73(1) March, 1990

Species in the Photinus consanguineus group (Fig. 8, n=6 +) are most interesting:
P. indictus shares the group characters, except that it has no light organ. The eyes of
males are smaller than typical, and indictus uses pheromones for sexual communication
(Lloyd 1973. LeConte (1881) described and named indictus (before the value of using
genitalia had been discovered), and placed it in a genus with unrelated, nonluminescent
fireflies, noting that (p. 33): "This insect has a deceptive resemblance to Photinus
consanguineus and other species of that genus." LeConte thus set aside the unmistak-
able resemblance of habitus detail and gave absolute weight to a character that "merely"
revealed a somewhat recent change in the mode of sexual signaling. Some-day, -millen-
nium, descendent species of indictus may comprise/warrant a different genus; maybe
this is what happened to modern genera with diurnal adults, such as Pyropyga,
Lucidota, and Ellychnia, which account for 18 species in North America-the species
of two of these genera were considered congeneric with indictus by LeConte.
Why would indictus give up flashing? What better hypothesis than predation by
Photuris? Might not some prey species, now flash-signaling in the dark with no other
information available, use chemical clues to distinguish predacious from conspecific re-
spondents? -a hitherto unsuspected bridge for the presumptive gap between the lumin-
ous ancestors of indictus and total reliance upon pheromones. We cannot criticize
LeConte as an erring semiosystematist, since there is no indication that he knew why
fireflies flashed. Nonetheless, his taxonomy may well have been influenced by an unap-
preciated predation connection.
Certain other species in this group have 2-flash patterns, with differences among
species being in the interval between the (homologous) two flashes (see Lloyd 1984b for
illustrations). In one species the interval is often 2 sec or more, depending upon temper-
ature, allowing plenty of time for an attacker to fly toward the first flash and be in
position to strike the second. Three other species in this group could have evolved their
shorter, species-specific intervals, and a fourth dropped the second flash of its pattern,
because of this. Could such signal changes coupled with geographic isolation lead to
rapid speciation? Notably, the 2-sec species (P. macdermotti) has an extensive geo-
graphic range, and the ranges of the others are peripherally sympatric. There are other
predator-related explanations for the noted changes in flash patterns in this species
group (Lloyd 1984c).
Not only could the flash patterns have been influenced by predator action, but so
also the space in which sexual activity takes place, and whether 3-dimensional, dark
airspace must be used for hiding between flashes. It is doubtful that sedentary flashing
systems as known from Asia and New Guinea could occur with Photuris present. Cicero
(1984) studied an Arizona member of the consanguineus group, P. knulli, outside the
range of Photuris, and found that it has a sedentary signaling system! One could venture
that this species is not likely to extend its range into eastern North America without
changing its sexual manners!
In a predator vacuum new and quite different evolutionary pathways must open up,
as Cicero's firefly suggests. Implications for taxonomy?: In the swarming Asian and
New Guinea species mentioned, where high density has resulted in keen male sexual
competition, males have evolved clamps to hold their females (see especially Wing et
al. 1984, Lloyd & Wing 1982, Lloyd et al. 1989). This has apparently happened indepen-
dently two or more times (L. Ballantyne, pers. comm.). The bent elytral tip, forming
the upper jaw of the clamp was the primary character that taxonomists used to separate
these species from Luciola, and put them in their own genus Pteroptyx (which because
of the aforementioned convergence is to be divided).
I will now summarily blame historical and present confusions of Photuris taxonomy
on the predations by females, and leave the details for later. Photuris genitalia are
largely the same throughout the genus and so far offer little help; nor are there any










Insect Behavioral Ecology-'89 Lloyd


Fig. 8. Habitus of Photinus macdermotti Lloyd. Without observing the male flash
pattern in the field, identification of this species is very difficult or impossible. From
the dorsal view this could as well be the nonluminescent P. indictus (LeConte). (Carbon
dust by L. Reep.)


apparent morphological characters to permit a satisfactory separation into species, the
hundreds of drab specimens now in museum trays. Hence the century-long existence
of "Photuris pennsylvanica." In the field Photuris males often emit flash patterns that
to be distinguished require electronic analysis. Males of several presumptive species
switch from one discrete pattern to another (Fig. 9), and some patterns they switch
among are copies of the patterns of species flying with them. To demonstrate this with









Florida Entomologist 73(1)


1.0 2.0 3.0 4.0 5.0 11:30 pm
Crep Units
Fig. 9. Barber (1951) noted that both P. lucicrescens and P. tremulans emitted two
different flash patterns: the former varying usage among sites, the latter spontaneously
switching back and forth with apparent "contagion." The switching of two species from
northern U.S. shows temporal consistency. (9A) During early evening males of P.
pennsylvanica (sensu Barber 1951) emit a single flash, an apparent copy of those emit-
ted by twilight Photinus spp. (1980), but gradually all switch to the species' "own"
pattern (Fig. 2, pattern 12). The 3 samples shown are from the same site in two different
months and years (4 July 1984; 15, 16 June 1987), the only samples that were made.
Fig. (9B) On two consecutive nights the proportion of Photuris "LIV males emitting
the mimicry pattern (a Pyractomena angulata- like flicker, see Lloyd 1980) rapidly
increased to more than 60% and then gradually fell off [with an at-the-time unnoted


March, 1990


64


1.0


0 0.8
E


0.0 t
0.7


1.2 1.7 2.2
Crep Units











Insect Behavioral Ecology-'89 Lloyd 65

confidence requires electronic analysis of patterns AND careful immediate collection of
PM-recorded specimens.
In many cases the flash patterns that Photuris males mimic are those of known prey
of their own firefly-hunting females. Thus, there is reason to suspect that problems
relating to multiple patterns are connected to female predation (Lloyd 1980). Male
pattern mimicry is recognized because models belong to other genera. But, Photuris
females also hunt Photuris; if Photuris males of one species mimic the pattern of males
of another and look-alike Photuris ....
The individual patterns used by males of a species are potential sources of reproduc-
tive isolation should members of isolated demes "lock onto" one, excluding others from
their repertoires6, because, say, of local predation ecology. For example, males using
one of the species' patterns might be less subject to aerial or mimicry predation from
other (or their own?) Photuris females. There are many models possible.
These facts, 1) prey and predator species vary in their geographic and seasonal
distributions and densities, 2) some Photuris may (through predation) dominate others
under certain circumstances, and 3) Photuris have multiple prey and probably switch
among them depending upon a variety of ecological circumstances, suggest that local
conditions must be extremely variable for prey and predator.
In the past thousands of years glaciers have closed out, opened up, and isolated
habitats and ranges, and shuffled species combinations and isolations considerably. With
a kaleidoscopic historical and environmental background, with shifting mosaic distribu-
tions, ecologies, and strong and sharply-focused predation, and given Photuris' certain
intrinsic potentials such as signal (neural) "plasticity," I temporarily feel comfortable
suggesting that recent (current) and rapid speciation is a common phenomenon in
Photuris, and invoke it as the blame for the cabinet taxonomic problems that the genus
has given. It is time to add technical developments such as electrophoresis to the
semiosystematic backpack. Surely there must be a knowable explanation for it all.

ENDNOTES

1. Extracted and abstracted from a monograph in preparation, "Fireflies of North
America."
2. I thank the editors of this symposium for inviting me to participate; Jon Allen, Paul
Choate, Howard Frank, and Tom Walker for helpful comments on the manuscript;
Jack Schuster for translating the "Resumen"; Steve Lasley for technical assistance
in computer analysis of data in Fig. 9; and Barbara Hollien for preparing the manu-
script. Florida Agric. Exp. Sta. J. Series No. R-00410.
3. Author of this species Karl DeGeer originally spelled the epithet with one "n," an
apparent, obvious misspelling of Pennsylvania (territory), and though both spellings
are seen (see Fig. 2, pattern 12), I believe pennsylvanicaa" is a legitimate emenda-
tion.
4. The term "sibling species" is used here for species that were not distinguished by
previously used methods, and does not necessarily imply relatedness or evolutionary
affinity.
5. Such studies are easily within the reach of student and class field-trip, given a little
instruction and familiarity with local species, and can integrate insect natural history,
basic electronics and computer driven and analyzed experimental protocols (see
Lloyd & Wing 1983).


sharp rise at the last observation(?)]. Two other samples made at a different site had
much smaller percentages of mimicking males, and the (a) profile (curve) is not appar-
ent. (Vertical axis, proportion of males emitting mimicry pattern; horizontal axis, time,
measured in Civil Twilight durations (creps) for geographic locality and date.)










Florida Entomologist 73(1)


6. Some evidence suggests that the possibility of polymorphism or other complex
mechanisms should not be overlooked (Lloyd 1983: 152).

REFERENCES CITED

ALEXANDER, R. D. 1964. The role of behavioral study in cricket classification. Sys-
tem. Zool. 11: 53-72.
BARBER, H. S., AND F. A. MCDERMOTT. 1951. North American fireflies of the genus
Photuris. Smithson. Inst. Misc. Coll. 117: 1-58.
BLACKWELDER, R. E. 1968. Some common taxonomic errors. Turtox News 46: 50-
53.
CICERO, J. 1984. Lek assembly and flash synchrony in Photinus knulli Green. Coleop.
Bull. 37: 318-342.
GREEN, J. W. 1956. Revision of the nearctic species of Photinus (Lampyridae: Col-
eoptera). Proc. California Acad. Sci. 28: 561-613.
1957.1 Revision of the Nearctic species of Pyractomena (Coleoptera: Lam-
pyridae). Wasman J. Biol. 15: 237-284.
LECONTE, J. L. 1881. Synopsis of the Lampyridae of the United States. Trans. Amer-
ican Entomol. Soc. 9: 15-72.
S_1883. Classification of the Coleoptera of North America. Smithson. Inst. Misc.
Coll. 38: 1-567.
LLOYD, J. E. 1966. Studies on the flash communication system in Photinus fireflies.
Univ. Michigan Mus. Zool. Misc. Pub. 130: 1-93.
S1973. Chemical communication in fireflies. Environ. Entomol. 1: 265-266.
S1980. Photuris fireflies mimic signals of their females' prey. Science 210:
669-671.
1983. Bioluminescnce and communication in insects. Annu. Rev. Entomol. 28:
131-60.
1984a. Occurrence of aggressive mimicry in fireflies. Florida Entomol. 67:
368-376.
1984b. Evolution of a firefly flash code. Florida Entomol. 67: 228-239.
1984c. On deception, a way of all flesh, and firefly signaling and systematics.
Oxford Surveys Evol. Biol. 1: 48-84.
LLOYD, J. E., AND S. R. WING. 1981. Florida Entomol. (photo story on Pteroptyx
valida copulation clamp). Florida Entomol. 64: 459.
1983. Nocturnal aerial predation of fireflies by light-seeking fireflies. Science
222: 634-635.
LLOYD, J. E., S. R. WING, AND T. HONGTRAKUL. 1989. Ecology, flashes and be-
havior of congregating Thai fireflies. Biotropica 21: 373-376.
MCDERMOTT, F. A. 1912. A note on Photinus castus. Canadian Entomol. 44: 312.
1914. The ecologic relations of the photogenic function among insects. Z. Wiss.
Insektenbiol. 10: 303-307.
1917. Observations on the light-emission of American Lampyridae. Canadian
Entomol. 49: 53-61.
S_1964. The taxonomy of the Lampyridae. Trans. American Entomol. Soc. 90:
1-72.
MAYR, E. 1982. The growth of biological thought. Harvard Univ. Press, Cambridge,
MA. 974 p.
OSTEN-SACKEN, C. R. VON. 1861. Die amerikanischen Leuchtkfer. Stettiner En-
tomol. Ztg. 22: 54-56.
SCUDDER, S. H. 1874. The distribution of insects in New Hampshire, pp. 331-384 in
Geology of New Hampshire, Cambridge.
SMITH, J. B. 1899. Insects of New Jersey. Annu. Rep. St. Bd. A. No. 27, Trenton.
755 p.
WALSH, B. D. 1864. On phytophagic varieties and phytophagic species. Proc. En-
tomol. Soc. Philadelphia 3: 403-429.
WENTZEL, H. W. 1896. Notes on Lampyridae, with the description of a female and
larva. Entomol. News 7: 294-297.


March, 1990










Insect Behavioral Ecology-'89 Pearson


THE EVOLUTION OF MULTI ANTI-PREDATOR
CHARACTERISTICS AS ILLUSTRATED BY TIGER BEETLES
(COLEOPTERA: CICINDELIDAE)

DAVID L. PEARSON
Department of Zoology,
Arizona State University
Tempe, AZ 85287, USA

ABSTRACT

Tiger beetles as a family show a broad spectrum of morphological, behavioral
and physiological mechanisms by which their enemies are deterred. This phenomenon
of multiple anti-predator mechanisms is also evident within species and individuals.
Although multiple anti-predator mechanisms have been widely recognized among most
if not all insects groups, general models and broad theoretical studies of predator-prey
interactions have largely ignored this confounding pattern. Based on experiments and
observations of tiger beetles, six theories are presented that explain the evolution of
multiple anti-predator characters: 1) several characters must operate in concert to
minimize predation, 2) each anti-predator character is largely or uniquely targeted
against one of several distinct foraging phases used by the predator, 3) increasingly
potent lines of defense may be used as a predator overcomes the primary ones, 4)
separate anti-predator characters are directed at each of several different types of
predator, 5) an individual prey is the result of a phylogenetic or ontogenetic accumula-
tion of anti-predator characters, and 6) competing or counterselective forces may over-
ride or supplement the effectiveness of some anti-predator characters.


RESUME

Las cicindelas, como familiar, demuestran amplios espectros morfol6gicos, de compor-
tamiento y de mecanismos fisiol6gicos por los cuales sus enemigos son disuadidos. Este
fen6meno de mecanismo multiples anti-depredador, tambi6n es evidence entire las es-
pecies y entire individuos. Aunque mecanismos multiples anti-depredadores han sido
reconocidos ampliamente entire la mayoria si no de todos los grupos de insects, models
generals y amplios studios teor6ticos sobre la interacci6n entire depredadores y su
presa, han practicamente ignorado estos patrons desconcertantes. Basado en ex-
perimentos y observaciones de cincindelas, se presentan seis teorias que explican la
evoluci6n de caracteres multiples anti-depredadores: 1) various caracteres deben de
operar al mismo tiempo para disminuir depredaci6n, 2) cada caracter anti-depredador
es mayormente, o es el blanco unico, contra una de las distintas fases forageras usadas
por el depredador, 3) aumentando la potencia de las lines de defense pudieran ser
usadas a media que el depredador sobrelleva las primaries, 4) caracteres separados
anti-depredadores son divigidos hacia cada uno de los diferentes tipos de depredadoves,
5) una presa individual es el resultado de una acomulaci6n filogen4tica o ontogen6tica
de caracteres anti-depredadores, y 6) fuerzas competitivas o contra-selectivas pudieran
sobrellevar o suplementar la efectividad de algunos caracteres anti-depredadores.



It is likely that the vast majority of prey species exhibit multiple anti-predator
characters (Pearson 1985, Endler 1988). However, most general models and mathemat-
ical theories of predator-prey interactions assume, at least implicitly, that prey have
only single anti-predator characters. This single character assumption has great poten-
tial for misleading and invalid results.










Florida Entomologist 73(1)


At least six theories predict the evolution of multiple anti-predator characteristics
within a single individual prey: 1) Some characters mayfunction in concert to minimize
predation. For instance, aposematic coloration and distasteful compounds are frequently
associated. A complication with this category is that each of these characters may not
effectively deter predation by itself, and if they only or usually function in combination,
they may technically be considered one character. Tiger beetles use body size,
brightly-colored abdomens exposed in flight, and defense chemicals against robber fly
predators. The per cent deterrence by these characters is greatest for larger beetles
with bright orange abdomens and benzaldehyde released from their defense glands
(Pearson 1985). As each of these characters is eliminated from models presented to wild
robber flies, the deterrent effect is reduced. Some characters such as large body size
are more important by themselves than other single characters, but the greatest protec-
tion is derived from a combination of all three together. Tropical forest tiger beetles
use nocturnal communal roosts that apparently rely on gregariousness to enhance de-
fense compound potency (Pearson and Anderson 1985).
2) Some anti-predator characters may be largely or uniquely targeted against each
of the distinct foraging phases of a predator (Endler 1986). Predator behavior can be
divided into distinct stages such as search, pursuit, capture and processing (Holling
1966). Among tiger beetles anti-predator characters like crypsis (Willis 1967) are
primarily effective against the searching phase, rapid flight against the pursuit phase
(Pearson 1985), chemical defense such as benzoyl cyanide (Pearson et al. 1988) against
the capture phase, and sharp mandibles together with enzymes in extradigestive juices
against the processing phase. Most individual tiger beetles have all these characters,
and together they provide protection from either a single predator or several different
predators through all these foraging phases.
3) Increasingly potent lines of defense may be used as a predator overcomes the
initial ones. The primary lines of defense function regardless of whether the predator
has been perceived by the prey, and they are likely to be energetically cheap (crypsis).
The secondary lines are initiated by an encounter with a predator and are generally
more energetically expensive (chemicals) (Robinson 1969, Rotheray 1986). Tiger beetle
colors that may serve as camouflage (Schultz 1986) or in mimicry of stinging hymenopte-
rans (Acorn 1988) are always present and take no extra energy to protect the beetle.
Flight and pygidial chemicals are only used when the primary defenses have been
breached.
4) Prey encountering several different types of predators may need a separate anti-
predator character targeted at each predator (Downes 1987). This phenomenon com-
pounds problems interpreting results from over-simplified experimental design as well
as makes models of frequency-dependent predation extremely complex (Endler 1988).
Tiger beetle adults use flight as an important mechanism to escape predation from
insectivorous lizards. However the instant they fly up from the ground, they become
susceptible to predation by robber flies, most of which can only take flying prey items
(Lavigne 1972. Pearson 1985). Specific defense chemicals like benzaldehyde are targeted
at robber flies. In addition, tiger beetles are the only beetles known to possess tympana,
and many species are thus apparently able to hear and respond to ultrasounds produced
by the wings of attacking robber flies (Spangler 1988). Tiger beetles respond to these
sounds with an instant contraction of abdominal muscles that in flight disrupts
aerodynamics and causes the beetle to fall to the ground where it is again susceptible
to the reptilian ground predators.
Tiger beetle body size is also related to reduction of predation from various predator
types. Small tiger beetles (<8 mm are more readily taken by insectivorous lizards and
spiders than are larger beetles. Large tiger beetles (>15 mm), however, are more likely
to be taken by insectivorous birds. Intermediate size tiger beetles are taken by robber


March, 1990











Insect Behavioral Ecology-'89 Pearson 69

flies (Pearson 1985). As with flight, body size that provides protection against one type
of predator automatically makes the prey more susceptible to another type of predator.
5) Selective forces on one stage of the life cycle or direct ancestors can be carried
over into other stages or descendants (Endler 1986, Downes 1987), and the resultant
individual may be a composite of anti-predator characters derived from predators on
all life cycle stages and ancestors. Larval tiger beetles have several highly specialized
parasitoid enemies (Pearson 1988). Large sized larvae need considerably more time
than do smaller congeners to sequester sufficient food to advance through their three
instars and pupal stage (Pearson & Knisley 1985). They are thus exposed longer to
mortality from parasitoids than are small larvae and are at an adaptive disadvantage.
However, selection for large size in the larval stage is probably at least partially the
result of adaptive advantages for large individuals as adults (Pearson & Knisley 1985).
Closely associated with this theory of an ontogenetic composite of multiple anti-pre-
dator characters is a theory that explains a composite individual based on historical
factors (see Edwards & Reddy 1986), evolutionary lag times, and differential genetic
liability of various characters. Among tiger beetle species, it is apparent that body color
for camouflage and mimicry can evolve relatively quickly (Schultz 1986). Chemical de-
fenses, however, evidently are extremely conservative and evolve very slowly (Pearson
et al. 1988). These differential evolutionary rates could result in an accumulation of
slowly evolving characters adapted for different predators over time. Some of these
characters may have served as pre-adaptations against subsequent predator(s), and
others may have taken on other functions such as in thermoregulation, competition or
courtship.
6) Additional factors may supplement selection for anti-predator characters or even
be counter-selective (Pearson 1988). This multiplicative or synergistic potential tre-
mendously complicates a simple understanding of the function and evolution of multiple
anti-predator characters and this complicating factor can impinge on all the preceding
theories. Thermoregulation, competition, courtship and a host of additional factors can
supplement the function of anti-predator characters or be counter selective (Endler
1987, Kingsolver & Wiernasz 1987). Among larval tiger beetles, for instance, the longer
it takes to capture sufficient food to molt into the subsequent instars and finally imagoes,
the greater the probability of mortality to parasitoids (Pearson & Knisley 1985).
Body size among endothermic and ectothermic insects can have a significant effect
on heating and cooling rates (May 1976, Heinrich 1981). Body size may also help
minimize interspecific competition (Pearson & Mury 1979, Pearson and Stemberger
1980, Pearson 1980, Pearson & Lederhouse 1987). If small body size to maximize ther-
moregulation and minimize competition overrides the disadvantages accrued by suscep-
tibility to predation, small body size is likely to be selected regardless of the increased
costs to predation. The ambiguities of character function involving several types of
adaptation besides predation is likely a common phenomenon.
Considerations for investigating the role of multiple anti-predator characters:
1) Determine all the potential predators and the relative risks to lifetime fitness by each
predator (Endler 1988).
2) Identify the rate of susceptibility and mortality in all life cycle stages.
3) Validate the target of each anti-predator character or suite of characters.
4) Establish the reduction of successful predation produced by each character.
5) Determine alternative functions of the anti-predator characters such as courtship,
thermoregulation, competition, etc.
ENDNOTES

John S. Edwards and John A. Endler critically reviewed early drafts of this article.
Thomas Zoebisch provided the Spanish translation of the abstract.










Florida Entomologist 73(1)


REFERENCES CITED

ACORN, J. H. 1988. Mimetic tiger beetles and the puzzle of cicindelid coloration (Col-
eoptera: Cicindelidae) Coleop. Bull. 42: 28-33.
DOWNES, W. L. 1987. The impact of vertebrate predators on early arthropod evolu-
tion. Proc. Ent. Soc. Washington 89: 389-406.
EDWARDS, J. S. AND G. R. REDDY. 1986. Mechanosensory appendages and giant
interneurons in the firebrat (Thermobia domestic, Thysanura:) a prototype sys-
tem for terrestrial predator evasion. J. Comp. Neur. 243: 535-546.
ENDLER, J. A. 1986. Defense against predators, p. 109-134, in M. E. Feder and G.
V. Lauder (eds.) Predator-prey relationships: perspectives and approaches from
the study of lower vertebrates. Univ. Chicago Press, Chicago, IL.
1987. Predation, light intensity and courtship behaviour in Poecilia reticulata
(Pisces: Poeciliidae). Anim. Behav. 35: 1376-1385.
1988. Frequency-dependent predation, crypsis, and aposematic colouration.
Phil. Trans. Roy. Soc. London B 319: 505-523.
HEINRICH, B. (ed.) 1981. Insect thermoregulation. Wiley-Interscience, NY.
HOLLING, C. S. 1966. The functional response of invertebrate predators to prey den-
sity. Mem. Ent. Soc. Canada 48: 1-86.
KINGSOLVER, J. G. AND D. C. WIERNASZ. 1987. Dissecting correlated characters:
adaptive aspects of phenotypic covariation in melanization pattern of Pieris but-
terflies. Evolution 41: 491-503.
LAVIGNE, R. J. 1972. Cicindelids as prey of robber flies (Diptera: Asilidae). Cicindela
4: 1-7.
MAY, M. L. 1976. Warming rates as a function of body size in periodic endotherms.
J. Comp. Physiol. B 111: 55-70.
PEARSON, D. L. 1980. Patterns of limiting similarity in tropical forest tiger beetles
(Coleoptera: Cicindelidae). Biotropica 12: 195-204.
1985. The function of multiple anti-predator mechanisms in adult tiger beetles
(Coleoptera: Cicindelidae). Ecol. Ent. 10: 65-72.
1988. Biology of tiger beetles. Annu. Rev. Ent. 33: 123-147.
and J. J. ANDERSON. 1985. Perching heights and nocturnal communal roosts
of some tiger beetles (Coleoptera: Cicindelidae) in southeastern Peru. Biotropica
17: 126-129.
M. S. BLUM, T. H. JONES, H. M. FALES, E. GONDA AND B. R. WITTE.
1988. Historical perspective and the interpretation of ecological patterns: defen-
sive compounds of tiger beetles (Coleoptera: Cicindelidae). American Nat. 132:
404-416.
AND C. B. KNISLEY. 1985. Evidence for food as a limiting resource in the life
cycle of tiger beetles (Coleoptera: Cicindelidae). Oikos 45: 161-168.
SAND R. C. LEDERHOUSE. 1987. Thermal ecology and the structure of an
assemblage of adult tiger beetle species (Cicindelidae). Oikos 50: 247-255.
SAND E. J. MURY. 1979. Character divergence and convergence among tiger
beetles (Coleoptera: Cicindelidae). Ecology 60: 557-566.
SAND S. L. STEMBERGER. 1980. Competition, body size and the relative energy
balance of adult tiger beetles (Coleoptera: Cicindelidae). American Midl. Nat.
104: 272-377.
ROBINSON, M. H. 1969. Defences against visually hunting predatores. Evol. Biol. 3:
225-259.
ROTHERAY, G. E. 1986. Colour, shape and defence in aphidophagous syrphid larvae
(Diptera). Zool. J. Linn. Soc. 88: 201-216.
SCHULTZ, T. D. 1986. Role of structural colors in predator avoidance by tiger beetles
of the genus Cicindela (Coleoptera: Cicindelidae). Bull. Ent. Soc. America 32:
142-146.
SPANGLER, H. G. 1988. Hearing in tiger beetles (Cicindelidae). Physiol. Ent. 13:
447-452.
WILLIS, H. L. 1967. Bionomics and zoogeography of tiger beetles of saline habitats
in the central United States (Coleoptera: Cicindelidae). Univ. Kansas Sci. Bull.
47: 145-313.


March, 1990









Insect Behavioral Ecology-'89 Witz


ANTIPREDATOR MECHANISMS IN ARTHROPODS:
A TWENTY YEAR LITERATURE SURVEY

BRIAN W. WITZ
University of South Florida
Biology Department
Tampa, FL 33620

ABSTRACT

Sixteen ecological and entomological journals were surveyed from 1969-1989 for
articles concerning defensive mechanisms in arthropods. Predators and prey are listed
taxonomically by family, and grouped according to the specific defensive mechanism
employed by the prey. A dichotomous categorization scheme is proposed which primar-
ily reflects the hypothesized energetic costs of various antipredator mechanisms.
In total, 354 papers were examined involving 555 potential or real predator/prey
interactions. It is concluded that certain defensive mechanisms are prevalent in particu-
lar taxonomic groups as indicated by the literature. Several predator/prey pairs occur
together in interactions more frequently than would be expected by chance (up to 39
times in one case). Most pairs (71 percent) occur together ten or fewer times. Thirty-four
pairs (nine percent) occurred only once. It is questionable whether this phenomenon is
an accurate representation of the natural distribution of defensive mechanisms in ar-
thropods because of the potential and real bias involved in the investigative process.
Many studies assume but do not demonstrate the efficacy of the alleged defensive
mechanism. Of 555 real or potential interactions, a defensive function was demonstrated
354 times (64 percent). No defensive function was demonstrated 201 times (36 percent).
It is suggested that whenever possible, future investigations incorporate tests of the
hypothesized defensive function against sympatric, and therefore ecologically relevant
predators.
RESUME

Se hizo una encuesta de 16 jornales ecol6gicos y entomol6gicos publicados entire
1969-1989 sobre articulos tratando con mecanismos de defense de artr6podos. Se listen
taxonomicamente y por familiar, depredadores y presa de acuerdo a los mecanismos
especificos de defense empleados por la presa. Se propone una categorizaci6n dic6toma
que primariamente refleja la hip6tesis de los costs energ6ticos de los various mecanismos
antidepredadores.
Se examinaron 354 articulos en total, involucrando la interacci6n de 555 depredadors/
presa real o potential. Se concluye que ciertos mecanismos de defense son prevalentes
en grupos taxon6micos particulares como es indicado por la literature. La interacci6n
de varias parejas de depredadores/presa ocurren mas frequentemente que se pudiera
esperar al azar (hasta 39 veces en un caso). La mayoria de las parejas (71%) ocurren
juntas diez o menos veces. Treinta y cuatro parejas (9%) ocurrieron solo una vez. Es
dudoso que este fen6meno sea una representaci6n precisa de la distribuci6n natural de
mecanismo de defense en artr6podos por el potential verdadero o possible de prejuicios
envueltos en el process investigative.
Muchos studios presumen pero no demuestran, la eficacia de los alegados mecanis-
mos defensivos. De 555 interacciones reales o potenciales, se demostr6 una funci6n
defensive 354 veces (64%). No se demostr6 ninguna funci6n defensive 201 veces (36%).
Se sugiere que cuando sea possible, las investigaciones futuras incorporen pruebas sobre
la hip6tesis de la funci6n defensive contra simpatricos, y de aqui, contra los depreda-
dores ecologicamente relevantes.


As biologists, and as scientists in general, we accumulate voluminous quantities of
information via the inductive process in an attempt to explain natural phenomena.
Inductive observation is the first step in the investigative process which enables us to
deductively generate testable hypotheses. It occasionally is necessary to summarize and









Florida Entomologist 73(1)


examine the accumulated data for patterns and formulate testable hypotheses which
guide research toward understanding any underlying causal mechanisms. In addition,
it often is useful to have ready access to review articles of a particular discipline as they
can alleviate, at least in part, the time consuming (and sometimes expensive) task of
exhaustive literature searches.
It has long been recognized that many organisms defend themselves against their
natural enemies. It also is abundantly clear that not all organisms defend themselves
in the same fashion (see Robinson 1969, Eisner 1970, Blum 1981, Matthews 1982, Her-
man 1984 for reviews on defensive mechanisms in certain arthropod taxa). More re-
cently, it has been noted that even within a particular individual or a particular species,
multiple defensive mechanisms are used (Pearson 1989).
In studying these defensive mechanisms, there is a tendency to focus on those which
are either unique, particularly interesting to the investigator, or readily studied,
perhaps because of logistical ease. This real or potential bias in the investigative process
may either conceal or artificially create patterns. The task in the present paper was to
examine some of the major ecological and entomological journals over the past twenty
years in an attempt to detect such bias. It is hoped that the exposition of these imba-
lances will redirect future research into those areas where minimal information exists,
to aid in evaluating the phylogeny of arthropod antipredator mechanisms. In addition,
the number of papers which actually demonstrated a defensive function for a given
mechanism was quantified.

MATERIALS AND METHODS

Sixteen ecological and entomological journals were examined for articles concerning
antipredator mechanisms in arthropods (Table 1). Although this is a relatively small
sample of the literature, these journals were selected because their format frequently
includes articles on defensive mechanisms. The original intention was to focus primarily
on insects, hence crustacean and arachnid journals were excluded. Articles appearing
in less taxonomically specialized journals, dealing with non-insectan arthropods, were
included as discovered. English articles were surveyed almost exclusively because of
time constraints imposed by translation. The journal volumes from 1969-1989 were
surveyed according to their availability at the University of South Florida library. The
years prior to 1969 were excluded as several reviews from this time period already exist
in the literature (see Blum 1981, Eisner 1970, Hermann 1984, Matthews 1982, Retten-
meyer 1970, Robinson 1969). Short communications, letters to the editor, and scientific
notes were excluded as many of these articles did not include detailed descriptions of
the defensive mechanism or its function. Only feature articles were surveyed. The
article titles were examined in each journal's annual table of contents, whenever avail-
able, for one or more of the following key words: defensive mechanism; antipredator
mechanism; predator-prey interaction; mimicry; crypsis; chemical defense; defensive
secretion; aposematic coloration; escape behavior; biting; stinging; defensive posture;
dilution; predator satiation; tending behavior; predation; mobbing; feigning death; avoid-
ance; and repellent. In addition, titles of suspected antipredator studies were noted.
Abstracts then were scanned to confirm the subject matter of the article, identify the
prey, predator(s), and the specific defensive mechanisms) employed by the prey. If the
abstract failed to provide this information, the remainder of the article was read. A
"general" predator category was devised for articles which neither demonstrated nor
hypothesized the defensive efficacy of the investigated mechanism against a specific
predator taxon.
Only articles dealing with classic predator/prey interactions were included. Interac-
tions of this sort typically are characterized by 1) the predator actually or potentially
killing and consuming the prey, 2) the predator being large relative to the prey, and 3)


March, 1990











Insect Behavioral Ecology-'89 Witz 73

TABLE 1. JOURNALS SURVEYED FOR ANTIPREDATOR MECHANISMS.

Journal Title Years and Volumes

American Naturalist 1969-1989:103-133
Animal Behaviour 1969-1989: 17-38(2)
Annals of the Entomological Society of America 1969-1989:62-82
Annual Review of Entomology 1969-1989:14-34
Behavioral Ecology and Sociobiology 1976-1989: 1-25(1)
Canadian Entomologist 1969-1989:101-121(8)
Ecological Entomology 1976-1989: 1-14(3)
Ecology 1969-1989: 50-70(4)
Ethology 1986-1989:71-81
Florida Entomologist 1969-1989: 52-72(2)
Journal of Animal Ecology 1969-1989: 38-58(2)
Journal of Chemical Ecology 1975-1989: 1-15(9)
Oecologia 1969-1989: 2-8(2)
Oikos 1969-1989:20-55(3)
Psyche 1976-1988:83-95
Zeitschrift fur Tierpsychologie 1974-1985:34-70

the predator consuming all or most of the prey individual. The following types of articles
were excluded from the study: grazing/partial predation; food site defense/territoriality;
predatory behavior if prey behavior was excluded; intraspecific interactions; predator/
prey interactions with simulated prey; applied entomology using predators as biological
control agents; methodological articles; general organismal biology unless the title in-
cluded mention of predator/prey interactions; phylogenetic articles unless antipredator
characteristics were used in the phylogenetic analysis; defensive mechanisms against
parasites; alarm pheromone articles that did not demonstrate a defensive function of
the chemical, and theoretical/modeling articles.
The following information was recorded for each article whenever provided: 1) scien-
tific name of predator and prey species; 2) the defensive mechanisms) employed by the
prey: 3) the taxonomic position of the predator and prey. This includes class, order,
family, genus, and species, whenever available; 4) the journal name, year, volume, page
numbers, authorss, and title.
In general, the taxonomy presented in the current paper is that supplied by the
authors) in each article. It often was necessary, however, to use other sources because
the authors) neglected to include taxonomic information in the article. For arthropods,
Borror, DeLong, and Triplehorn (1976) was the primary reference, supplemented with
Borror & White 1970, Levi & Levi (1987), Barnes (1974), Schultz (1969), and Edmondson
(1959). Burt & Grossenheider (1976) was used for mammals; Robbins et al. (1983) for
birds; Conant (1975) for reptiles and amphibians, and Wainwright (1976), Moyle & Cech
(1988), and Robbins & Ray (1986) for fish. It occasionally was necessary to use a
taxonomic synonym for the author's nomenclature to prevent redundant categories. For
example, several British authors use the ordinal name Heteroptera which was replaced
with Hemiptera. Although other taxonomic schemes may be more current and perhaps
more accurate, these references were chosen for logistic reasons. It was not the intent
of this paper to dispute the phylogenetic validity of any taxon.
After the list of articles was compiled, the following information was quantified: 1)
number of citations occurring in each defensive mechanism category; 2) the taxonomic
distribution of prey within each defensive category at the ordinal level. This reflects
the percent contribution of a given order to the number of articles dealing with a specific
defensive mechanism; 3) the frequency of occurrence of specific predator/prey pairs.
This was an attempt to determine if a given prey taxon consistently defends itself
against a certain predator taxon, as indicated by the literature; and 4) if a defensive










74 Florida Entomologist 73(1) March, 1990

function actually was demonstrated, and if so, the taxonomic position of the predator
against which it was effective.
The following defensive categories were recognized, most of which are well estab-
lished in the literature: 1) chemical (e.g. Blum 1981, Eisner 1970, Hermann 1984); this
includes a broad diversity of both specialized exocrine gland secretions and other "sys-
temic" defenses such as reflexive bleeding, sequestration of toxic or noxious chemicals,
etc. 2) crypsis (e.g. Bishop 1972, Erichsen et al. 1980, Heinrich 1979, Lees & Creed
1975, Robinson 1969)-this includes cryptic coloration, mimicking some portion of the
environment, and behavioral crypsis whereby the organism remains in a concealed
region of the habitat; 3) mimicry (e.g. Brower 1988, Doyen 1974, Hetz & Slobodchikoff
1988, Pough 1988, RettenmeYer 1970)-both Batesian and Mullerian; 4) escape (e.g.
Krasne & Woodsmall 1969, Nentwig 1982, Roitberg et al. 1979)-this involves detection
of the predator by the prey and usually rapid movement away from the predator; 5)
biting (e.g. Hermann 1984); 6) stinging (e.g. Hermann 1984); 7) mobbing (e.g. Wittmann
1985)-this involves physical contact or intimidation displays by a group of prey; 8)
fighting-this includes all other physical contact with the predator such as kicking, wing
beating, shoving, pinching, entangling (e.g. Bildstein et al. 1989, Mills & Partida 1976,
Nutting & Spangler 1969), etc. Although this category is not well established in the
literature, it was devised to include an array of defensive mechanisms which may have
similar energetic requirements (see discussion below); 9) posture/size (e.g. Jakobsen &
Johnsen 1988, Kevan et al. 1983)-this involves predator avoidance of prey because of
behavioral posturing (including phragmosis, e.g. Wheeler and Holldobler 1985) or large
size; 10) dilution/satiation (e.g. Gillett et al. 1979, Heller & Milinski 1979, Milinski 1979,
1984)-the predator is either confused or satiated by large numbers of individual prey,
thereby reducing the probability that any one individual is eaten (safety in numbers);
11) flash/warning coloration (e.g. Coppinger 1970, Malcolm 1986)-the predator is either
confused, startled, or conditioned to avoid prey because of bright coloration or patterns;
12) misdirected attack (e.g. Robinson 1969)-the predator is confused by the morphol-
ogy of the prey and attacks less vulnerable regions of the body; 13) acoustic (e.g. Bauer
1976, Buchler et al. 1981, Sandow & Bailey 1978)-the prey produces a loud noise such
as stridulation which startles the predator; 14) feigning death (thanatosis) (e.g. Capinera
1976, Howard et al. 1982, Otto & Sjostrom 1983); 15) mutualism (e.g. Bristow 1984,
Burns 1973, Cushman & Whitham 1989, Fritz 1984)-this involves defense of prey by
a mutualistic symbiont; and 16) armor (e.g. Klein & Burkholder 1983, Otto & Sjostrom
1983, Pecarsky 1987, Silgerglied & Aiello 1976)-including all anatomical weaponry
such as spines, thick cuticle, horns, etc. It should be noted that certain categories are
not always mutually exclusive. Stinging and biting are usually accompanied by venom
injection, aposematic coloration often is coupled with noxious or toxic chemicals, flash
coloration with crypsis, and armor is sometimes associated with fighting. Hence a single
interaction may appear in multiple defensive categories. Biting, stinging, mobbing, and
fighting were grouped together for statistical analyses as they are all mechanical de-
fenses (at least in part) which involve active physical contact with the predator and
approach by the prey.

RESULTS

A total of 354 articles was recorded with 555 real or potential interactions (Appendix
1). The number of interactions exceeds the number of articles because 1) several articles
dealt with more than one predator/prey interaction and 2) some articles investigated
multiple antipredator mechanisms in a given prey. Because of space limitation, the
species list of predator and prey is not presented here. This list is available upon request
to the author, and is arranged in an alphabetized order of journals with articles listed
in chronological sequence within journals. Appendix 1 lists the predator and prey of


__










Insect Behavioral Ecology-'89 Witz


each interaction according to family (when available) in each of the defensive categories
delineated above.
Many articles appear in more than one defensive category. Combining all categories
results in 452 articles dealing with specific antipredator mechanisms. Chemical defense
was by far the most prevalent (208 or 46 percent of all cases), exceeding other
mechanisms by at least a factor of five (Fig. 1). This is attributed at least in part to the
Journal of Chemical Ecology dealing exclusively with chemical defense (109 citations).
Coleopterans, followed by hymenopterans and lepidopterans, are the most prevalent
taxa in the chemical defense category (Fig. 2), exceeding all other arthropod orders by
at least a factor of two.
Fighting (via biting, stinging, and all other modes) is the second most abundant
mechanism. Hymenopterans dominate this category, exceeding all others by at least a
factor of four (Fig. 2). Lepidopterans dominate the third most abundant category, cryp-
sis. The fourth category, escape, is more evenly distributed among the orders, with
hymenopterans, homopterans, and coleopterans dominating. Mimicry and flash/warning
coloration are nearly equally represented in the literature and constitute the fifth and
sixth most abundant categories respectively (Fig. 3). The taxonomic distribution of the
remaining six categories is presented in Fig. 3 and 4.
The most prevalent predator/prey interaction at the ordinal level is between
hymenopteran predators and hymenopteran prey, representing 39 out of 364 predator/
prey pairs (11 percent). Lepidopteran prey are paired with passeriform predators 24
times (7 percent). The third most prevalent interaction is between coleopteran prey and
hymenopteran predators, comprising 18 interactions (5 percent). All other predator/
prey pairs occur together less than 15 times, with 34 interactions represented only once
(Table 2). This distribution of occurrences of predator/prey order-pairs differs signifi-
cantly from a Poisson distribution (X2, 14 df= 180.25; p<0.005), indicating that pairs are
non-randomly represented in the literature. Certain pairs occur together more fre-
quently than expected by chance, while others are under represented.


ANTIPREDATOR MECHANISMS IN ARTHROPODS
NUMBER AND % OF CITATIONS PER MECHANISM

CHEMICAL 206 48%








FIGHTING 51 511%
ARMOR 12 S%
MUTUALISM 18 a3
DILUTION 17 4%
CRYPTIC 41 9a
R 4- POSTURE/SIZE 18 4%
APOSEMATISM 22 5%
ESCAPE 38 8% MIMICRY 23 5%

Fig. 1. Distribution of references concerning antipredator mechanisms in ar-
thropods. Fighting is a combination of stinging, biting, and all other modes of physical
contact e.g. kicking, wing beating, pinching, etc.















76 Florida Entomologist 73(1)



CHEMICAL DEFENSE DISTRIBUTION
NUMBER AND S OF CITATIONS PER TAXON N



HYMENOPTERA 1241 EqLP7EDaaW2S

YNGETERERAL -: 6/


ORIIOPrEPa a
LEPIDOPTIERA 17T NoI itECAn
nOOPOPERA II 8
IDPTERA 10 80 HEMIPTERA 4 7%


March, 1990



FIGHTING DEFENSE DISTRIBUTION
NUMBER AND S OF CITATIONS PER TAXON


DPTERA 7,_

OOLEOPTERA 8
10%


OTHER INSECTI 2
1ONIOPTERA 2
DEP=OA 2
ORTHOPTEAA 2
4RD4as


CRYPTIC DEFENSE DISTRIBUTION ESCAPE DEFENSE DISTRIBUTION
NUMBER AND S OF STATIONS PER TAXON NUMBER AND $ OF STATIONS PER TAXON


OTHER INSECTA 7 Ho ER1 8NE





4 3 GENERMLrNSECT. 2
OPA A LEP ERA TLOPTEA






Fig. 2. Distribution of chemical, fighting, cryptic, and escape defensive mechanisms





MIMICRY DEFENSE DISTRIBUTION FLASH/W'.RNING COLOR DEFENSE DISTRIBUTION
NUMBER AND S OF CITATIONS PER TAXON NUMBER AND S OF CITATIONS PER TAXON

2E T I PTI P PT RA S

CLE 6 SXX2EER7 OOLPTERA 6 OEPO 1
22GNERAL INSECT^

?SCLEOPTERA 2
22 PPOD 4 as


01PTERA 8
21%


HYI TERA 1


OMENRAL INECTA 2
as


HMIPTERA
as6


as
M~NOPTERA 1
HOMOPTIER 2
e6


POSTURA/SIZE DEFENSE DISTRBUTON DILUTION/SATIAION DEFENSE DISTBUTION
NUMBER AND S OF CITA7ONS PER TAQIV NUMBER AND OF CITATIONS PER TAXON


MY2 RA 4 TRA
EPME24PWTIA 4 h




O.EOPOERA 6 A ERA







Fig. 3. Distribution of mimicry, flash/warning coloration, postural/size, and dilution/
satiation defensive mechanisms among arthropod orders.
LEPIOWTWRA 1 26



Fig. 3. Distribution of mimicry, flash/warning coloration, postural/size, and dilution/
satiation defensive mechanisms among arthropod orders.











Insect Behavioral Ecology-'89 Witz 77


MUTUAUSM DEFENSE DISTRIBUTION ARMOR DEFENSE DISTRIBUTION
NUMBER AND OF STATIONS PER TAXON NUMBER AND s OF CITATIONS PER TAXON
IMOPT9ERA ?





LEPOTEA OCLEEOPTERA







ACOusIITIC DEFENSE DISTRIBUTION FEIONING DEATH DEFENSE DISTRIBUTION
NUMBER AND S OF CITATIONS PER TAXON NUMBER AND S OF CITATION. PER TAXON
LEpTERA O CO-EG4OTERA 2
ET PTEAA 2 I
E0T NEUROPTERA1







PLELEPIDOPTEA OPTE







ORT2M EWEAA I

Fig. 4. Distribution of mutualism, armor, acoustic, and feigning death defensive
mechanisms among arthropod orders.

A defensive function of the alleged antipredator mechanism was demonstrated in
364 cases (66 percent). In most articles it was not possible to determine if the predator
was sympatric with the prey (and therefore ecologically relevant) (see Witz &
Mushinsky 1989). The efficacy of the alleged defensive mechanism was not demonstrated
in 191 cases (34 percent). Of this 34 percent, the effectiveness of the mechanism against
specific predator taxa was hypothesized, but not demonstrated in 81 cases (15 percent);
the defensive function was neither demonstrated nor hypothesized to be effective
against specific predator groups in 120 cases (21 percent).

DISCUSSION

Although the prevalence of chemical defense in the literature is biased by the pre-
dominance of such articles in the Journal of Chemical Ecology (109 citations), this cate-
gory constitutes the majority of antipredator mechanisms investigated, even with that
journal removed from the analysis (99 citations). Several prolific authors further contrib-
ute to this imbalance, publishing multiple papers on chemical defensive mechanisms.
Two authors combined were involved in 20 percent of the chemical defense articles.
Many papers dealt almost exclusively with analytical chemistry.
It would be interesting to attempt to control for multiple authorship and journal
imbalance statistically. Comparing this corrected literature representation of antipre-
dator mechanisms with an estimated, geographically-corrected number of species per
order may reveal a more accurate picture of the true distribution of defensive
mechanism categories among arthropods. It is questionable however, whether sampling
the literature will ever truly reflect the natural distribution of a particular phenomenon.
This is perhaps particularly true of arthropod defensive mechanisms, given the high
diversity of the phylum, the myriad antipredator categories, and the often cryptic na-










78 Florida Entomologist 73(1) March, 1990

TABLE 2. ORDERS PAIRED IN PREDATOR/PREY INTERACTIONS.

# Occurrences # Pairs # Interactions

1 34 34 (09%)
2 17 34 (09%)
3 10 30 (08%)
4 11 44 (12%)
5 6 30 (08%)
6 7 42 (12%)
7 1 7(02%)
9 3 27 (07%)
10 1 10 (03%)
11 1 11 (03%)
14 1 14 (04%)
18 1 18 (05%)
24 1 24 (07%)
39 1 39 (11%)

ture of certain species. I believe that summarizing the literature, at the very least, can
reveal areas where a paucity of information exists and perhaps stimulate subsequent
investigation.
The fact that 34 percent of the studies examined in the present review did not
demonstrate a defensive function may be attributed to several factors: 1) it is not always
logistically feasible to do so; 2) it is possible that the defensive function was demon-
strated in previous studies which were not discussed in the text; 3) some authors may
choose to postpone demonstration of the defensive function for future publications; 4)
the natural history of sympatric predators may be unknown or unfamiliar to the authors;
5) because previous work on similar species demonstrated the defensive efficacy of a
particular mechanism, the authors assume a similar function in the species which they
are investigating. There are, no doubt, other reasons which have been overlooked. It
is difficult to determine which, if any, of these reasons are prevalent. Of the above
reasons, I believe number 5) to be the most critical. Reason 1) and 4) are perhaps the
most difficult to circumvent. It is suggested, however, that reasons 2) and 3) often can
be avoided by a brief mention of previous or expected investigations which demon-
strated or will demonstrate the defensive function of the mechanism under investiga-
tion. Finally, reason 5) can be avoided by refusing to assume defensive efficacy and
endeavoring to incorporate tests of the alleged antipredator mechanism against ecolog-
ically relevant predators.
It is suggested that the energetic requirements of the aforementioned defensive
categories may differ. It would be interesting to measure the energy required for a
prey to defend itself on a per predation event basis. Chemically defended prey for
example, often need to replenish their chemical arsenal following an attempted preda-
tion event. Conversely, cryptic coloration is a permanent, non-depletable mechanism
which may require energy only to find the appropriate habitat. The following dichotom-
ous categorization for antipredator mechanisms is therefore proposed: passive defenses
are those relatively energetically inexpensive mechanisms that reduce the probability
of an encounter between prey and predator. Included in this category are cryptic,
dilution/satiation, flash/warning coloration, and mimicry antipredator mechanisms; ac-
tive defenses are those energetically expensive mechanisms which increase the probabil-
ity of surviving a predator attack. Included in this category are chemical, fighting,
acoustic, escape, armor, and postural/size antipredator mechanisms. It is suggested
that when possible, future investigations might incorporate estimates of the energetic
involved in defensive mechanisms to aid in understanding the relative importance of
predation pressure in the evolution of antipredator mechanisms.





















Insect Behavioral Ecology-'89 Witz


APPENDIX




EI.ALO. DPEF.N v PREY PRANT
EF. NO. PREY PREDATOR REF.NO. PREY PREDATOR REP.NO. PREY PR88IAlTR


296. TEEBRIONIDAE RODENTIA

351. ACRIDIDAE VERTEBRATA:
GENERAL

338. CHRYSONELIDAE FORMICIDAE

144. FORMICIDAE NANALIA:
GENERAL

344. CARABIDAE MURIDAE

327. TENEBRIONIDAE NANKALIA:
GENERAL

339. CHRYSOMELIDAE FORMICIDAE

14. GYRINIDAE SALNOHIDAE.
CERTRARCHIDAE

179. STAPHYLINIDAE TERNITIDAE

15. PTERNARCIDAE FORICIDAE
PELTOPERLIDAE

217. DYTISCIDAE CYPRINIDAE

172. REDUVIIDAE GENERAL

366. TERMITIDAE FORNICIDAE

326. TENEBRIONIOAE GENERAL

4. COREIDAE GENERAL

17. ACRIDIDAE HOMINIDAE

363. MENBRACIDAE COCCINELLIDAE
REDUIIDAE
SYRPHIDAE

145. FORMICIDAE GENERAL

364. MEMBRACIDAE IGUANIDAE

345. APDAE FORKICIDAE

223. APHIDIDAE INSECTA:
41 SPP.

63. APIDAE IOINIDAE

355. APHIDIDAE GENERAL

230. GYRINIDAE GENERAL

187. FORRICIDAE FORMICIDAE

62. KPIDAE GENERAL

295. HETERONENIIDAE CRICETIDAE
FORKICIDAE
CORVIDAE

65. APIDAE GENERAL:
SIMULATED

349. NOTODONTIDAE BRACONIDAE

180. CUCUJIDAE ANTHOCORIDAE
CUCUJIDAE

146. FORMICIDAE HONINIDAE

352. ACRIDIDAE CRICETIDAE

303. HYMENOPTERA: VERTEBRATA:
GENERAL GENERAL

155. TENEBRIONIDAE GENERAL

194. PENTATONIDAE PORNICIDAE
GENERAL

190. CNRYSOPIDAE FORNICIDAE

18a. HYNENOPTERA: FORMICIDAE
GENERAL VESPIDAE
MALICTIDAE
APIDAE

304. HY1IENOPTERA: VERTEBRATA:
GENERAL GENERAL

348. EMIDAE OST ICTHYES:
10 SPP.,
TRIOWYTHIDAE

266. RHOPALIDAE BUFMOIDA
CORVIDAE

141. VESPIDAE FORMICIDAE

22. APIDAE GENERAL
FORMICIDAE
TERITIDAE

1. VESPIDAR MHOINIDAE
GENERAL


156. FORKICIDAI GENERAL
TIWBRIONIDAE

205. INSECT: GENERAL
GENERAL

231. ARTHROPODA: GENERAL
GENERAL


259. TERMITIDAE GENERAL 158. LYGAEIDAE CORVIDAE

82. COLEOPTERA: GENERAL 91. LINREPHILIDAE FORMICIDAE
GENERAL
25. CHRYSO8ELIDAE FORNICIDAE
24. ARTHROPODA GENERAL
GENERAL 207. BLATTIDAE HOININDAE

301. ARCTIIDAE AVES: 76. CHRYSOMELIDAE AVES:
NOCTUIDAE GENERAL GENERAL

242. CICINDELIDAE GENERAL 342. NOTODONTIDAE GENERAL

3. HENIPTERA: GENERAL 258. ISOPTERA: GENERAL
GENERAL GENERAL

100. TERNITIDAE FORMICIDAE 314. CHRYSOMELIDAE FORMICIDAE

358. FORNICIDAE FORMICIDAE 231. DYTISCIDAE APHIBIA:
GENERAL

GENERAL
33209. THORIDIIDAE OCRICETIDAE TE E
209. FORNICIDAE FORICIDAE 80. CARABIDAE GENERAL,

163. VESPIDAE HOMIIIDAE, SIMULATED
SIMULATED 299. PHASMATIDAE GENERAL

70. TENEBRIONIDAE CICETIDAE 137 ATOPHORIDAE FOMICIDAE
FORMICIDAE DERMAPTERA:
DEERAL
201. APHIDIDAE ARANEIDAE GENERAL
PARIDAE 208. SCOLOPENDRIDAE GENERAL

73. SATURNIIDAE GENERAL: 315. CHRYSOMELIDAE GENERAL
SIMULATED
250. APHIDIDAE GENERAL
320. NEODIPRIONIDAE PENTATOMIDAE
267. COSNETIDAE GENERAL
341. POLYDESMIDA GENERAL GONYLEPTIDAE

87. TENERIONIDAE GENERAL 110. MUTILLIDAE FORMICIDAE

362. XEMBRACIDAE IGUANIDAE 9. TERPITIDAE FORMICIDAE
DASYOPIDAE
117. APIDAE APIDAE MYRMECOPHAGIDAE

281. CTUCHIDAE AVES (6 SPP) 8. TERMITIDAE FORMICIDAE
LACERTIDAE MYNECOPHAGIDAE

29. NYPALIDAE CORVIDAE 261. TERMITIDAE GENERAL

331. DAMAIDAE ARANEIDAE 128. STAPHYLINIDAE GENERAL

135. INSECT FORMICIDAE 93. STYGNOMMATIDAE FORMICIDAE
22 SPP. IGUAIIDAE

241. CICINDELIDAE ASILIDAE 129. TENEBRIONIDAE GENERAL
LANZIDAE
ACCIPITRIDAE 321. FORMICIDAE FOMPICIDAE
TEIIDAE
IGUANIDAE 152. PAPILIONIDAE VESPIDAE:
GENERAL
77. PAPILIONIDAE FORMICIDAE PORMICIDAE:
SALTICIDAE GENERAL
ARANEAE:
340. FORNICIDAE COLUBRIDAE GENERAL
AVES:
175. COPEPODA: SALGONIDAE GENERAL
GENERAL CENTRARC IDAE
CLADOCERA: PERCIDAE 102. CANTHARIDAE SALTICIDAE
GENERAL
81. CHRYSO1ELIDAE GENERAL

139. LEPIDOPTERA: PARIDAE
GENERAL 262. TERMITIDAE GENERAL

290. CULICIDAE NOTONECTIDAE 170. FORMICIDAE GENERAL

249. LA ICAMPIDAE FORNICIDAE 154. CHRYSOMELIDAE FORMICIDAE

284. COREIDAE EBERIZIDAE 64. APIDAE GENERAL:
PYRRHOCORIDAE SIMULATED
PENTATOMIDAE
5PENTATOMIDAE DANAIDAE CORVIDAE

224. VESPIDAE GENERAL:
APIDAE SIMULATED 248. STAPHYLINIDAE GENERAL

92. FORNICIDAE GENERAL 165. APIDAE REDUVIIDAE

219. GYRINIDAE AMPHIBIA: 83 STAPHYLINIDAE GENERAL
GENERAL
AVES; 164. STAPHYLINIDAE FOMICIDAE
GENERAL
OSTEICTHYES: 58. TERMITIDAE FORMICIDAE
61. ARCTIIDAE CHIROPTERA
283. STAPHYLINIDAE GENERAL GENERAL

218. DYTISCIDAE CYPRIIDAE 41. FORMICIDAE GENERAL
GYXINIDAE
282. TERMITIDAE FORMICIDAE
169. GEOPHILIDAE FORNICIDAE
113. DYTISCIDAE OSTEICTHYES:
90. XYSTOoESMIDAE FORNICIDAE GENERAL
AJPHIBIA:
346. FORMICIDAE GENERAL GENERAL

101. COSEIDAE FONICIDAE: 298. APIDAE GENERAL:
GONYLEPTIDAE GENERAL SIMULATED

233. APHIDIDAE GENERAL 51. ANTHOPHORIDAE FORMICIDAE

182. STAPHYLINIDAE FORRICIDAE: 54. GLONERIDAE ARANEAE:
GENERAL GENERAL
HENIPTERA: INSECT:
GENERAL GENERAL
ARAME: VERTEBRATA:
GENERAL GENERAL

260. TERITIDAE FORMICIDA



















Florida Entomologist 73(1)


March, 1990


PENNO. PUNS PUESATO*R


322. TERNITIDAE


RF- NO. PREY PREDATOR


PnomrTCTIn 30. LYCAENIDAE GENERAL


37. DANAIDAE CORVIDAE

184. LYGAZDAE GENERAL

131. TERNITIDAE GENERAL

31. TERKITIDAE GENERAL

255. VESPIDAE VESPIDAE

239. CHRYSOMELIDAE TENTHREDINIDAE

256. VESPIDAE FORRICIDAE

36. DANAIDAZ GENERAL

60. DANAIDAE GENERAL
NYINPILIDAS

112. CHRYSONELIDAE NANTIDAE

66. APIDAE GENERAL:
SIULAUTED

132. SCUTELLERIDAE FORNICIDAE

166. APIDAE GENERAL

130. TENHERIONIDAE GENERAL

12. FORMICIDAE FORMICIDAE
VERTEBRATA:
GENERAL

221. LYGARIDA GENERAL

84. STAPKYLINIDAZ DROSOPHILIDAE
FORNICIDAE

26. TENTREDINIDAE FORNICIDAE
PARIDMA

150. TENEBRIONIDAE GENERAL

45. PAPILIONIDAE FORMICIDAE

133. SCUTELLERIDAE GENERAL

59. TERNITIDAE GENERAL

55. OEDENERIDAE HOINIDAE

167. ACRIDIDAZ FORNICIDAE

288. DOrAIDAS CORVIDAE
RODENTIA:
GENERAL
287. LYGAIDAE AIUTIDAE

276. CHRYSOELIDAE POMWICIDAZ:
GENERAL

50. APIDAE FORNICIDAE
REIDTIIDAE
ASILIDAE
AGELRIIDAB
ARUANIDAI
104. SILPHIDAE PORNICIDA

213. 8ILPHIDAZ SALTICIDAZ

197. DAAIDAR CORVIDA

228. YGAENIDAI VMRT ATA:
GENERAL
319. FUmrCIDAR PFOIICIDA

168. ACRIDIDAI FOII? CIDA

275. APIDAR HCGNID~
VIRTEBRATA:

157. TEIEBRI 0IDA GENERAL

354. LEPIDOPT2ERA GIEERAL
GEEmMAL

205. DAUAIDAE AVE:
GENERAL
RODEITIA:
GENERAL
40. CArTlARIDAE GENERAL

308. GREnETRIDA CORVIDAE

214. FORIICIDAI GENERAL

171. TETHRlDINIDAE ARANEAEJ:
GENERAL
FORPICIDAE:
GENERAL
119. YPOWKEUTIDAE GENERAL

121. RIYN IAIDAI VERTEBRATA:
GENERAL
4. PFOMICIDAE rFOICIDAE

203. DMRIDAE COVIDAE

261. CARAIDAU SCINCIDAE


240. FORMICIDAE FORMICIDAE

79. CARABIDAE HOHINIDAE

202. APHIDIDAE ARANEIDAE

67. APIDAE GENERAL:
SIMULATED

232. APHIDIDAE COCCINELLIDAE

264. ACRAEINAE GENERAL

142. ARTHROPODA: CICHLIDAE
GENERAL

206. FORMICIDAE FORMICIDAE:
GENERAL

226. PERGIDAE FORMICIDAE
REDUVIIDAE
NUSCICAPIDAE
CRICETIDAE

138. NOTODONTIDAE PASSERIFORKES;
DAMAIDAE 5 SPP.
LASIOCAMPIDAE
NYMPHALIDAE
NOCTUIDAE
LYPARIDAE

178. PIERIDAE CORVIDAE
148. TENEBRIONIDAE VESPERTILIONIDAE
NUSTELIDAE
BASSARISCIDAE
18. LEPIDOPTERA: FOREICIDAE
GENERAL

236. DANAIDAE AVES:
GENERAL
237. DANAIDAE GENERAL
ACRAEIDAE

195. FORIICIDAE FORNICIDAE

162. PAPILIONIDAE PARIDAE

292. LYGAEIDAE PARIDAE

212. MIRIDAE IGUANIDAE
LYGAEIDAE

109. LYGAEIDAE PHABIANIDAE

153. ARANEIDAR CORVIDAE

307. ACRIDIDAZ SPHECIDAE

103. SILPHIDAR GENERAL

69. CARABIDAE NYREELEOHTIDAE

11. CARABIDAE 8COLOPACIDAE

34. FORKICIDAE FOIICIDAI
BUPOMIDAE
AILADIDA!
PRICKLIDA
NUtIDAR

191. TNVEBRIOSIDAE CRICETIDAE


CRTPTIC DRPEnEIV KNCH3UWIEBM
REF.NO. PREY PREDATOR

280. NOCTUIDAE AVES:
GENERAL

2. ACRIDIDAE AVES:
GENERAL

125. ACRIDIDAE HOMINIDAE
SQUAKATA
CRACIDAE

127. ACRIDIDAE AVES:
5 SPP.

94. PAGURIDAE HOMINIDAE

111. PERILIDAE SALMONIDAE

13. CULICIDAE GENERAL

120. ACRIDIDAE GENERAL
PIERIDAE

310. GEOMETRIDAE GENERAL
HOMINIDAE

254. PSOCIDAE AVES:
GENERAL

124. ORTHOPTERA AVES:
GENERAL
SAURIA:
GENERAL

134. PAPILIONIDAE AVES:
GENERAL

200. ENNOMINAE RUSCICAPIDAE
TROGLODYTIDAE
STURNIDAE

135. INSECT: FORNICIDAE
22 SPP.

350. FORIICIDAE IGUANIDAE

48. COPEPODA: SALMONIDAE

139. LEPIDOPTERA: PARIDAE
GENERAL,

118. COPEPODA ATHERINIDAE

290. CULICIDA! NOTONECTIDAE

78. PYRALIDAE FORRICIDAE
GENERAL
ARANEAE:
GENERAL

199. HIPPOLYTIDAE SPARIDAE

183. PQ.ALNONIDAZ CYPRINODONTIDAE

20. GEONETRIDAE AVES:
GENERAL

192. GEONETRIDAE TROGLODYTIDAE
MUSCICAPIDAE
PARIDAB
PASSERIDAE

311. GEOZETRIDAE AVES:
GENERAL
HOMINIDAE
107. TENBRMIONIDAE PARIDAE

81. CHRYSONELIDAE GENERAL

26. TERTHREDINIDAE FORNICIDAE
PARIDAB
138. NOTODONTIDAE PASSERIFORKES:
DANAIDAE 5 SPP.
LASIOCANPIDAE
NYMPHALIDAE
OCTUIDAE
LYPARIDAE

185. CYNOHOIDDAE LARRIDAE

71. NOTONECTIDAE CENTRARCHIDAE

343. LIBELLULIDAE CENTRARCHIDAE

251. LIBELLULIDAE CBNTRARCHIDAE

300. EPHMERERLLIDAE PERLIDAE

353. CERCOPIDAE AVES:
GENERAL
356. LIKNXPHILIDAE GENERAL

273. CHRYSONRLIDAE PENTATONIDAE
MABIDAR
COCCINELLIDAR
147. VISPIDAE GENERAL

347. FORNICIDAE GENERAL

105. CHRYSOPIDAR GENERAL

204. GONhTR IDAJ MPLLM IAIDAZ
UMBRISIDAS



















Insect Behavioral Ecology-'89 Witz 81




IMIICRT D1MaI cIK llMfuu XCAPX Damo IuICMXuIuM IITIM D'S MCI K MSIUMS (ITIMIMIG)
REF..O. PREY PRlEDATOR REF.NO. PREY PREDATOR 2rEF. O. PRLY PREDATOR

296. CERAMBYCIDAE RODENTIA 186. HOMARIDAE XAA3IA: 49. FORMICIDAE FORMICIDAE:
GENERAL GENERAL
265. INSECT: GENERAL
GENERAL 294. BLATTIDAE VERTEBRATA: 97. APIDAE PYRALIDAE
GENERAL
39. NYMPHALIDAE GENERAL 65. APIDAE GENREAL:
189. FORMICIDAE FORMICIDAE SIMULATED
87. TENEBRIONIDAE GENERAL
89. FORMICIDAE FORMICIDAE 146. FORMICIDAE HOMINIDAE
313. BRASSOLIDAE AVES:
GENERAL 53. CALLIPHORIDAE LANIIDAE 303. HYKENOPTERA: VERTEBRATA:
IGUAHIDAE GENERAL GENERAL
367. PORTUNIDAE OSTEICTHYES:
281. ZYGAENIDAE AVES GENERAL 188. FORMICIDAE VERTEBRATA:
6 SPP. AVES: VESPIDAE GENERAL
LACERTIDAE GENERAL HALICTIDAE
CRUSTACEA: APIDAE
136. SYRPHIDAE PASSERIFORMES: GENERAL
GENERAL 304. HYENOPTERA: VERTEBRATA:
196. FOR ICIDAE NMYRELEOGTIDAE GENERAL GENERAL
29. NYRPHALIDAE CORVIDAE
52. CHRYSOKELIDAE CORVIDAE 1. VESPIDAE HONINIDAE
274. INSECTA: GENERAL PARIDAE
GENERAL 163. VESPIDAE MONINIDAE,
301. NOCTUIDAE CHIROPTERA: GENERAL:
337. DIPTERA: AVES: ARCTIIDAE GENERAL SLATE
21. SPP. GENERAL
242. CICINDELIDAE GENERAL 225. APIDAE GENERAL
336. SYRPHIDAE PASSERIDAE,
CONOPIDAE IMIMDAE, 88. FORMICIDAE FORMICIDAE 137. AMTHOPHORIDAE FORMICIDAE
TACHINIDAE EMBERIZIDAE, DELAPTERA:
ASILIDAE CORVIDAE 73. SATURNIIDAE GENERAL GENERAL:
MANTISPIDAE SIMULATED
64. APIDAE GENERAL:
160. REDUVIIDAE GENERAL 225. APIDAE GENERAL SIMULATED

81. CHRYSONELIDAE GENERAL 272. APHIDIDAE COCCINELLIDAE 214. FORnICIDAE GENERAL

308. GEOMETRIDAE CORVIDAE 368. APHIDIDAE COCCINELLIDAE 67. APIDAE GENERAL:
CICADELLIDAE SIMULATED
148. TENEBRIONIDAE VESPERTILIONIDAE
MUSTELIDAE 159. CICADIDAE FORMICIDAE 191. VAEJOVIDAE CRICETIDAE
BASSARISCIDAE
135. INSECT: FORNICIDAE 108. APIDAE EMBERIZIDAE
236. ACRAEIDAE AVES: 22 SPP.
GENERA L C D D333. FORMICIDAE IGUANIDAE
241. CICIRDELIDAP ASILIDA.
237. NYMPHALIDAE GENERAL LANIIDAE
SATYRIDAE ACCIPITRIDAE
PAPILIONIDAE TEIIDAE
LYCAENIDAE IGUANIDAE

212. MIRIDAE IGUANIDAE 122. FORMICIDAE HOMINIDAE
LYGAEIDAE
306. PALAEMONIDAE CENTRARCHIDAE
291. CERAMBYCIDAE GENERAL,
HONINIDAE 305. PALAEONIDAE CENTRARCHIDAE

99. TEPHRITIDAE GENERAL 350. PORMICIDAE IGUANIDAE

85. NORDELLIDAE GENERAL 244. HEPTAGENIIDAE PERLIDAE,
MELOIDAE RAETIDAE PERLODIDAE
PTERONARCIDAE
108. SYRPHIDAE ENBERIZIDAE
289. NOTONECTIDA MNIPTCERA
333. FORMICIDAE IGUANIDAE
74. COPEPODA: BLENIIDAE
GENERAL

272. APHIIDAE COCCINELLIDAE



233. APHIDIDAE GENERAL

81. CHRYSOMELIDAE GENERAL

154. CHRYSOMELIDAE FORMICIDAE

229. IN SECTA ARANEIDAE
10 ORDERS

312. CICADIDAE EMBERIZIDAE

277. APHIDIDAE ANTHOCORIDAE

56. VESPIDAE FORKICIDAE

307. ACRDIDAE SPHECIDAZ

191. ACRIDIDAE CRICETIDAE



















Florida Entomologist 73(1)


March, 1990


nlm RIn DoraI xmmIU (ITIMNGI FrrIMO DMs MZ1cnImmS (EMONINGl) Fr iTING DUrES lacaNUIIl (IO1B)
REF.P. PREY PREABTOR REF.NO. PRY PREDATOR REF.0N. PREY PREDATOR


49. FORMICIDAE FORMICIDAE 57. THEIDIIDAE SPHECIDAE

144. rORMICIDAE RAMKALIA:
GENERAL

259. TERNITIDAE GENERAL

100. TERMITIDAE PORMICIDAE

360. APIDAB APIDAR

260. TERNITIDAE FORMICIDAE

370. RHINOTERNITIDAE FORKICIDAE

321. FORMICIDAE FORKICIDAE

298. APID GENERAL:
SIMULATED

322. TERMITIDAE FORKICIDAE

166. APIDAE AHGALIA:
GENERAL

275. APIDAE HOMIKIDAE
VERTEBRATA:
GENERAL

79. CARABIDAE HONINIDAE

307. ACRIDIDAE SPHECIDAE


TURAL/ISIl Dm US! RUCR 2ISMS FLAB/IJARnM COLORATION
REP.NO. PREY PREDATOR REF.. = 0 PREY PREDATOR

326. TENEBRIONIDAE GENERAL 242. CICINDELIDAE GENERAL

259. TERMITIDAE GENERAL 201. APHIDIDAE ARANEIDAE
PARIDAE
176. ACRIDIDAE HONINIDAE
341. POLYDESMIDA GENERAL
241. CICINDELIDAE ASILIDAE
LAN IIDAE 87. TENEBRIONIDAE GENERAL
ACCIPITRIDAE
TEIIDAE 362. NENBRACIDAE IGUANIDAE
IGUAIIDAE
313. BRASSOLIDAE AVES:
306. PALAENONIDAE CENTRARCHIDAE GENERAL
IGUANIDAE
318. LASIOCAMIPIDAE FORICIDAE
241. CICINDELIDAE ASILIDAE
245. EPHEMERELLIDAE PERLIDAE LAIIDAE
PERLODIDAE ACCIPITRIDAE
PTERONARCIDAE TEIUDAES
IGUAHIDAE
86. DAPHIKIDAE TENORIDAE
284. COREIDAE EUBERIZIDAE
370. RHINOTERNITIDAE FORNICIDAE PYRRHOCORIDAE
PENTATOMIDAE
154. CHRYSOLELIDAE PORIICIDAE
81. CHRYSONELIDAE GENERAL
50. APIDAE REDUVIIDAE
ASILIDAE 154. CHRYSONELIDAE FORNICIDAE
AGELENIDAE
ARANEIDAE 61. ARCTIIDAE CHIROPTER:
GENERAL
171. TENTHREDINIDAE ALRNEAZ:
GENERAL 26. TENTHREDINIDAE FORNICIDAE
FORNICIDAE: PARIDAS
GENERAL
207. LYGAEIDAZE MATIDAE
235. TAERIOPTERYGIDAE SAIOMNIDAE
PERLIDAE 228. ZYGAENIDAE VERTBRATA:
GENERAL
245. EPHMEERELLIDAE PERLODIDAE
BAETI DA PTERONARCIDAE 308. GEONETRIDAE CORVIDAE
PERLIDAE
202. APHIDIDAE ARANEIDAE
246. EPHMENERELLI DAE PERLODIDAE
369. MORPHOIDAE TYRANUIDAE
247. EPHZERELLIDAE PERLODIDAE GA9LULIDAE

307. ACRIZIDAE SPRECIDAE 178. PIERIDAE CORVIDAR
147. VESPIDAE GENERAL 162. PAPILIONIDAE PARIDAS

347. FORNICIDAE GENERAL 292. LYGAEIDAE PARIDAN

212. MIRIDAE IGUANIDAE
LYGAIIDAE

307. ACRIDIDAE SPHECIDAZ


106. PAGURIDAE GENERAL

115. OXYOPIDAE TORKICIDAE:
GENERAL

19. OXYPODIDAE THRESKIORNITHIDAE

234. DERMESTIDAE INSECA:
GENERAL
VERTIERATA:
GENERAL
ARACIDA:
GENERAL

220. DEESTIDAE TENEBRIONIDAE

363. NEMBRACIDAE COCCIHELLIDAE
REDUVIIDAE
SYRPHIDAE

198. DERBESTIDAE CHELONETHIDA
HYMEROPTERA

209. FORMICIDAE FORMICIDAE

316. TINGIDAE COCCINELLIDAE:
SIMULATED

80. FORMICIDAE FORMICIDAE

278. TETRANYCHIDAE PHYTOSEIDAE

73. SATURNIIDAE GENERAL:
SIMULATED

320. NEODIPRIONIDAE PENTATONIDAE

181. SCARABEIDAE SCARABEIDAE

137. ANTHOPHORIDAE FORMICIDAE
DERMAPTERA:
GENERAL

277. APHIDIDAE ANTHOCORIDAE

307. ACRIIDAE SPHECIDAE

147. VISPIDAE GENERAL





DILrUTrIOfS/TIATIOM rFICT
REF.NO. PREY PREATOR

309. KAJIDAE AVES:
GENERAL

125. ACRIDIDDAE RHOIMIDAE
SQUANATA
CRACIDAE

126. ACRIDIDAE HOINIDAE
LACERTIDAE

140. DAPHINIDAE GASTEROSTEIDAE

215. DAPHINIDAE GASTEROSTEIDAE

323. GERRIDAE AVES:
SIMULATED

324. GERRIDA AVES :
SIMULATED

325. GERRIDAE CLUPEIDAE
MUGILIDAE

216. DAPHINIDAE GASTEROSTEIDAE

161. MOSMIHIDAE GASTEROSTEIDAE

201. APHIDIDAE ARANEIDA
PARIDAE

159. CICADIDAZ FORRICIDAX

241. CICINDELIDAE ASILIDAE
LANIIDAR
ACCIPITRIDAN
TEIIDAE
IGUAMIDAR

173. CICADIDAR STURMIDAE
MUSCICAPIDAE
CORVIDAE
DEBRIZIDAE

328. APHIDIDAB COCCINDLLIDAE

7. MOTONECTIDAR SIMULATED

312. CICADIDAE EBRERISIDAE
















Insect Behavioral Ecology-'89 Witz 83



RORIULZHI DRYWIN EROdNI*RU ACOUSTIC
IREI .. PREY PREDATOR RP.HO. PREY PREDATOR
365. IERBRACIDAE GENERAL 279. TETTIGONIDAE SCINCIDAE
253. LYCAEIIDAE TKERIDIIDAE 42. PABSALIDAE CORVIDAE
SPARASSIDAE
THOMISIDAE 52. CHRYSOKELIDAE CORVIDAE
CLUBIONIDAE PARIDAR
ARAMEIDAE
PORIICIDAE 242. CICINDELIDAE GENERAL
VESPIDAE
211. NUTILLIDAE LYCOSIDAE
32. COCCIDAE COCCINELLIDAE HYDROPHILIDAE CRICETIDAE
CARABIDAE
46. COCCIDAE COCCIRNLLIDAE
PYRALIDAE 81. CHRYS8NELIDAE GENERAL
317. APHIDIDAE COCCINELLIDAE 21. ACRIDIDAE GEKKONIDAE
RIRIDAK
SPHECIDAE 11. CARABIDAE SCOLOPACIDAE
LINYPHIIDAE
116. CHRYSOMELIDAE FORMICIDAE
74. MEMBRACIDAE SALTICIDAE
33. APHIDIDAE COCCINELLIDAE PEGIGING D fTZ
MENBRACIDAE CRYSOPIDAE REF. O. PREY PREAUTOR
151. COCCIDAE COCINELLIDAE 52. CHRYSONELIDAE CORVIDAE
PARIDAK
43. REMBRACIDAE GENERAL 154. CYSONELIDAE FORICIDAE
114. LYCAEIDAE GENERA TAENIPTERYGIAE S I
235. TAENOPTERYG I DAE A IDAE
297. LYCAEIDAE GENERAL PERLIDAE
149. TERNITIDAE FORMICIDAE


NMXDIRENCTUD AT"IC
REF.NO. PREY PREDIOR
atom Dnine RUSISIR
pRF. O. PREY PREDATOR 291. CERANBYCIDAE GENERAL
ROKIMIDAB
70. TEMEBRIONIDAE FORMICIDAE
371. DAPHINIDAE ATHERINIDAE
330. PACURIDAE CANCRIDAE
174. BOSNINIDAE TIEONIDA
CYCLOPIDAZ
86. DAPHINIDAE TENDORIDAE
96. FOIMICIDAE SALTICIDAE
81. CHRYRKDELIDAE GENERAL
79. CARABIDAE HONIIDAE
235. TAENIOPTERYGIDAE BSAIONIDA
PERLIDAE
291. CEAMBYCIDAE GENERAL
1HONIMDAX
273. CHRYSONELIDAE PENTATOMIDAE
NABIDAN
COCCINELLIDAR
105. CHRYSOPIDAE GENERAL



ACKNOWLEDGMENTS


I thank the Florida Entomological Society for inviting me to participate in this

symposium. I would especially like to thank Howard Frank and Earl McCoy for their

interest in my research. I thank Earl McCoy additionally for reading the first draft of

this manuscript. Thanks to Pablo Delis for translating the abstract.



REFERENCES CITED

(Numbers refer to the appendix)


1. AKRE, R. D., AND H. G. DAVIS. 1978. Biology and pest status of venomous
wasps. Annu. Rev. Ent. 23: 215-38.
2. ALCOCK, J. 1972. Observations on the behaviour of the grasshopper,
Taeniopoda eques Burmeister, Orthoptera: Acrididae. Anim. Behav. 20: 237-42.
3. ALDRICH, J. R. 1988. Chemical ecology of the Heteroptera. Annu. Rev. Ent.
33: 211-38.
4. AND T. R. YONKE. 1975. Natural products of abdominal and
metathoracic scent glands of coreoid bugs. Ann. Ent. Soc. America 68: 955-60.
5. ATSATT, P. R. 1981. Lycaenid butterflies and ants: selection for enemy-free
space. American Nat. 118: 638-54.




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

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