Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00053
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 1996
Copyright Date: 1917
Subject: Florida Entomological Society
Entomology -- Periodicals
Insects -- Florida
Insects -- Florida -- Periodicals
Insects -- Periodicals
General Note: Eigenfactor: Florida Entomologist: http://www.bioone.org/doi/full/10.1653/024.092.0401
 Record Information
Bibliographic ID: UF00098813
Volume ID: VID00053
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Full Text

Behavioral Ecology Symposium '95: Pare and Tumlinson 93


Insect Attractants, Behavior, and Basic Biology Research Laboratory,
Agricultural Research Service, United States Department of Agriculture
Gainesville, Florida 32604


A surge in release ofvolatiles by several plant species can be observed in response
to insect feeding. Oral secretions from these feeding herbivores provide the initial
chemical signal that triggers the release of plant volatiles; one or more elicitors from
the oral secretion allow the plant to identify and differentiate herbivore feeding from
mere mechanical wounding. Elicitor(s), in combination with mechanical wounding,
trigger the release of compounds both locally and systemically. These volatiles, which
may be a blend of constitutive and induced compounds, vary in their relative and ab-
solute concentration over time. They serve as easily detectable and distinctive chem-
ical cues for predators and parasitoids of the herbivores feeding on the plants. Volatile
compounds released from herbivore infested plants include the monoterpenes and
sesquiterpenes of the isoprenoid pathway, green leaf volatiles of the fatty acid/lipoxy-
genase pathway and aromatic metabolites, such as indole and methyl salicylate, of
the shikimic acid/tryptophan pathway.

Key Words: Plant defenses, herbivores, plant volatiles, chemical cues, predators, par-


La liberaci6n de volatiles puede ser observada en varias species de plants
cuando los insects se alimentan de ellas. Las secreciones orales de los herbivoros
mientras se alimentan de las plants provee la seiial quimica inicial que dispara la li-
beraci6n de los volatiles. Una o mas sustancias estimulantes de las secreciones orales
permiten a la plant identificar y diferenciar el efecto de los herbivoros al alimentarse
del efecto de las heridas mecanicas. La(s) sustancia(s) estimulante(s), en combinaci6n
con las heridas mecanicas, disparan la liberaci6n de compuestos local y sist6mica-
mente. Estos volatiles, que pueden ser una mezcla de compuestos constitutivos e in-
ducidos, varian en su concentraci6n absolute y relative en el tiempo. Ellos funcionan
como seiial quimica facilmente detectable y distintiva para los depredadores y para-
sitoides de los herbivoros. Los compuestos volatiles liberados de plants infestadas
con herbivoros incluyen monoterpenos y sesquiterpenos de la via de los isoprenoides,
volatiles de hoja verde de la via del acido graso/lipoxygenase y metabolitos aromati-
cos, tales como el indol y el methyl salicilato de la via del acido shikimic/tryptophan.

When plants are subjected to pathogen or herbivore attack, they activate biochem-
ical defenses, such as phytoalexins (Grayer & Harborne 1994) or proteinase inhibitors
(Nelson et al. 1983, Pearce et al. 1993), that directly target their natural enemies. Ad-
ditionally, the release of plant volatiles can cue insect parasitoids and/or predators
that act as another line of plant defense (Tumlinson et al. 1993, Stowe et al. 1995). It
appears that blends of volatile plant terpenoids, which are released in response to in-

Florida Entomologist 79(2)

sect feeding and not to mechanical damage alone (Turlings et al. 1990, Korth et al.
1995), allow natural enemies of insect herbivores, such as parasitic wasps, to distin-
guish between infested and non-infested plants, thus aiding in the location of hosts or
prey (Turlings et al. 1995). These phytodistress signals, which result in an active in-
teraction between herbivore-damaged plants and a third trophic level, have been de-
scribed for several plant species. Examples include (1) lima beans that produce
volatiles that attract the predatory mite Phytoseiulus persimilis when damaged by
the spider mite Tetranychus urticae (Dicke et al. 1993) or (2) corn plants that produce
volatiles that attract the Hymenopterous larval parasitoids Cotesia marginiventris
and Microplitis croceipes when under attack by Spodoptera exigua caterpillars (Turl-
ings et al. 1991, Turlings & Tumlinson 1991).
There are several reports which describe the chemical cues of particular plant/her-
bivore/carnivore interactions. In this review, examples will be drawn from several
such tritrophic interactions to examine how plants identify herbivore feeding and
what role chemical storage and de novo synthesis may play in the blend of volatiles re-
leased, both locally and systemically from the plant.


With or without insect feeding, plants usually release a variety of hydrocarbons
during periods of high temperatures (Sharkey & Singsaas 1995). Recent evidence sug-
gests that this release is a strategy for coping with high temperatures. It is suspected
that fat soluble hydrocarbons dissolve into the thylakoid membrane which surrounds
chlorophyll molecules and serve as a glue to keep the chloroplast from melting when
temperatures exceed the plant's biological optimum. Because these molecules are so
volatile, they quickly evaporate as the temperature rises, prompting the plant to re-
lease more molecules (Mlot 1995). A large number of compounds have been identified
in the collection of head space volatiles from different plant species. Included in the
list are fatty acid derived aldehydes and alcohols (Croft et al. 1993), terpenes derived
from mevalonic acid, and aromatic metabolites such as indole and methyl salicylate
derived from shikimic acid (Mann 1987). Clearly, more than one biosynthetic pathway
is responsible for this phytochemical release. The total energy channelled into the
synthesis and release of volatiles is uncertain, although it is estimated that isoprene
production alone typically siphons off 2% of the carbon fixed through photosynthesis
(Mlot 1995).
In response to herbivore damage, this blend of volatiles can radically change with
the surge of individual constituents. For example, in cucumber plants infested with
spider mites, there is an increase in release of the terpenes (E)-P-ocimene and (E)-4,8-
dimethyl-1,3,7-nonatriene from undetectable levels to 25 and 54%, respectively, of the
total volatiles. This causes the relative amounts of (Z)-3-hexen-l-yl-acetate and (Z)-3-
hexen-1-ol to drop from 35 to 50% of the total volatiles released from uninfested plants
to 1% from infested plants (Takabayashi et al. 1994a). In the apple cultivar Summer
Red, a similar shift in component ratios is observed with a surge of the terpenes (E,E)-
a-farnesene and (3E,7E)-4,8,12-trimethyl-1,3,7,11-tridecatetraene after spider mite
infestation. This results in a relative drop of volatiles such as n-hexanal and (E)-2-
hexenal compared with uninfested plants (Takabayashi et al. 1994b). It has not yet
been established whether the increased amount of volatile compounds released is
caused by de novo biosythesis due to herbivore feeding, or whether such compounds
are present constitutively in the plant and their release from storage is triggered by
this feeding.
Loughrin et al. (1994) suggest that in cotton some monoterpenes and sesquiterpe-
nes are constitutive, providing built-in protection against invading organisms, while

June, 1996

Behavioral Ecology Symposium '95: Pare and Tumlinson 95

others are induced after feeding begins. The constitutive compounds are identified by
their almost immediate release after feeding starts, their relatively constant release
over time, and their rapid wane after feeding stops. Such metabolites can be synthe-
sized and stored in external epidermal hairs called trichomes or in cells embedded in
the leaf, such as mesophyll pigment glands (Gershenzon et al. 1989, Croteau &
Johnson 1984). In either case, these volatiles can be released by the simple breaking
of the glands as a result of herbivore feeding. In contrast, the induced compounds
show a delay between the time feeding starts and the release of volatiles. For example,
in cotton this delay of compound release is between 12 and 24 hours after herbivore
feeding begins (Loughrin et al. 1994). In addition, these induced compounds show a
diurnal cycling of release which continues after herbivore feeding has ceased. Exam-
ples of proposed constitutive compounds in cotton include a-pinene and caryophyl-
lene; induced metabolites include (E,E)-a-farnesene and (E)-P-farnesene, (E)-P-
ocimene and (E)-4,8-dimethyl nonatriene (Loughrin et al. 1994). Thus far, there has
been no determination of induced versus constitutive compounds with labeling exper-
iments using radioactive or stable isotope precursors to establish the time sequence
of synthesis with respect to the release of volatiles.
In addition to the release of volatiles at the site of herbivore feeding, analysis of
volatile emissions from unharmed leaves of insect damaged plants has established
that there is a systemic response. In both corn and cotton, leaves distal to the site of
herbivore feeding showed an increase in the release of volatiles. The chemical blend
of volatiles from undamaged leaves differs from the volatiles collected from the entire
plant (Turlings & Tumlinson 1992). One set of compounds not systemically released
are the leafy green volatiles of the lipoxygenase pathway, which are usually detected
in freshly cut or damaged tissue. In addition, at least in cotton, some of the monoter-
penes and sesquiterpenes, as well as indole, are only released locally.


The systemic release of specific volatiles by damaged as well as undamaged leaves
suggests a mobile inducer which can travel from the wounded leaf to the undamaged
portions of the plant. The signalling mechanism responsible for the systemic release
of predator and parasitoid attractants by plants is under intense investigation in sev-
eral laboratories (Turlings et al. 1993, Boland et al. 1992). One of the proposed models
of signal transduction is that the plant stores volatile metabolites as glycosides; when
cells are broken by herbivore feeding and exposed to P-glucosidase from the saliva of
the feeding herbivore, stored glucoside molecules are hydrolyzed and the volatile com-
pounds are released. The presence of P-glucosidase activity in saliva from the larvae
Pieris brassicae which feed on cabbage and induce the release of volatiles was re-
ported recently (Mattiacci et al. 1995). In addition, the same laboratory group ob-
served the release of volatiles from cabbage plants when commercially purified
almond P-glucosidase was exogenously applied to mechanically wounded leaves. Ra-
diochemical-labeling studies are needed to establish whether such an enzyme inducer
moves through the plant and accounts for the systemic release of volatiles that is ob-
A second model of signal transduction leading to plant volatile release is the lipid-
based lipoxygenase pathway which has already been shown to activate de novo syn-
thesis of defense related proteins. In a series of experiments, Farmer & Ryan (1992)
demonstrated that intermediates of the lipoxygenase pathway are potent elicitors
causing the synthesis and accumulation of proteinase inhibitors. They have also
shown that compounds such as p-chloromecuribenzene sulfonic acid and phenidone,
which channel intermediates out of the pathway, are effective inhibitors of protein

Florida Entomologist 79(2)

synthesis associated with plant defense responses (Narvaez-Vasquez et al. 1994 and
information within Cucurou et al. 1991). In their model, the interaction of signal mol-
ecules with plasma membrane receptors leads to the activation of lipase and the re-
lease of linolenic acid into the cytoplasm. Linolenic acid is then converted through
phytodienoic acid to jasmonic acid which interacts with receptors to activate protein-
ase inhibitor gene expression.
In tomato, the earliest signal that has been identified with herbivore wounding is
the release of an 18 amino acid polypeptide, systemin. Radiolabeling studies have
shown that this signal moves throughout the plant via the phloem (Pearce et al. 1991)
and inhibitor studies suggest that systemin is loaded apoplastically into the vascular
tissue (Narvaez-Vasquez et al. 1994). In other words, the polypeptide is loaded into
the sieve tube/companion cell complex from the free space of the apoplast and not via
the plasmodesmata.
In this laboratory, an elicitor has been isolated, purified and partially character-
ized from the oral secretions of beet armyworms, Spodoptera exigua Hiibner. The elic-
itor is a low molecular weight compound, too small to be an enzyme. When this
partially purified factor is supplied to young corn plants through cut stems at approx-
imately 100 nM level, volatiles which are specific to insect feeding are released
throughout the plant (Turlings et al. 1993). Experiments will be carried out to deter-
mine whether such an elicitor may activate the lipoxygenase pathway, as has been im-
plicated with the plant elicitor systemin. The procedures for a large scale purification
of this saliva factor have been established so that a complete chemical analysis will be


At the species level, head space volatiles of several crop plants have been analyzed
with and without insect damage. Though the chemical blend, as well as the total vol-
atile release varies, there are some compounds that many species have in common.
Repeatedly, the acyclic C,, homoterpene (E)-4,8-dimethyl-1,3,7-nonatriene and the C,
homoterpene (3E,7E)-4,8,12-trimethyl-1,3,7,11-tridecatetraene have been identified
in the head space of herbivore infested plants including lima bean, apple, cowpea, cu-
cumber, corn and cotton (Dicke 1994, Loughrin et al. 1994, Turlings et al. 1991).
Data on variation in volatiles released at the cultivar level is limited. In cotton, no
significant differences were reported in the volatiles released among commercial cot-
ton varieties, though a naturalized cotton variety did emit substantially greater quan-
tities of volatiles per plant than any of the commercial hybrids (Loughrin et al. 1995).
Five varieties of corn exposed to herbivore injury differed significantly in the quantity
and variety of terpenes released. The genotypes showed variation as well in their re-
sistance to the important corn pest, fall armyworm. However, these differences did
not appear to affect the ability of the parasitoids Cotesia marginiventris and Micropli-
tis croceipes, to locate their host, the fall armyworm (Turlings et al. 1995). Clearly
some variation in the volatile blend released by an individual plant does not com-
pletely eliminate chemical signalling.
Leaves from different parts of an individual plant also can show variation in the
release of herbivore-induced volatiles. In cucumber, although young and old leaves do
not differ much with regard to the total volatiles released from spider mite induced
plants, a greater variety of compounds are released from young than older leaves
(Takabayashi et al. 1994a). Young and mature leaves on beet army worm-infested cot-
ton plants differ in the blend of volatiles that are released (R6se & Tumlinson unpub-
lished data).

June, 1996

Behavioral Ecology Symposium '95: Pare and Tumlinson 97

Differences in the amount of volatiles released between individual plants are often
subtle and can easily be masked by variations in environmental conditions that influ-
ence the plants' physiology. Several species, including corn, cotton (Loughrin & Tum-
linson unpublished data) and lima bean (Takabayashi et al. 1994b) respond to
reduced light, either from lower light intensity or shorter day length, with a decline
in the release of herbivore induced volatiles. Based on studies with lima bean, water
stress seems to be directly related to volatiles released. With less water available for
the plant, elevated levels of volatiles are released from infested individuals relative to
non-water stressed controls. Correlating this with insect preference showed that
predatory mites selected plants which were infested and water stressed over infested
but not water stressed plants (Takabayashi et al. 1994a).
The variation between species in herbivore-induced volatile release and the gen-
erality of their response in the plant kingdom can not yet be satisfactorily addressed
because only a limited number of crop plants have been studied.


Though there are numerous reports on the release of volatiles with insect feeding
as well as the effect of these volatiles on herbivore parasitoids, there is little informa-
tion on the way in which plants regulate the blend of compounds emitted. Are the re-
leased parasitoid attractants synthesized de novo? If so, what are the biosynthetic
pathways that are activated under such regulation? An outline of the biosynthetic
route to plant volatiles is shown in Fig. 1. At least three biosynthetic pathways are ex-
pected to be responsible for the blend of volatiles that are released; the isoprenoid
pathway which produces monoterpenes and sesquiterpenes, the fatty acid/lipoxygen-
ase pathway which generates green leaf volatiles and jasmone, and the shikimic acid/
tryptophan pathway which results in several compounds, including indole (Mann
1987). From data reported on the timing of volatile release with insect feeding in cot-
ton, it appears that the terpene pathway is responsible for the release of most of the
inducible compounds (Loughrin et al. 1994).

glucose agly s cetyl Coenzyme A

shikimic acid/ fatty acid/ isoprenoid
tryptophan lipoxygenase pathway
pathway pathway



Figure 1. Primary and secondary metabolic pathways leading to volatile com-
pounds released in herbivore damaged plants.

Florida Entomologist 79(2)

A key intermediate in the isoprenoid pathway is mevalonic acid which is formed by
the condensation of three acetyl-CoA molecules to (3S)-3-hydroxy-3-methylglutaryl
CoA and a two step reduction to mevalonic acid (Goodwin & Mercer 1990). As shown
in Fig. 2, isopentenyl pyrophosphate (IPP), the five carbon building block of terpenes,
is then formed via decarboxylation of mevalonic acid pyrophosphate (Gershenzon &
Croteau, 1989). Half of the IPP is enzymatically converted to the isomer dimethylallyl
pyrophosphate and the two isomers condense via prenyltransferases to form geranyl
pyrophosphate (GPP). GPP can be channelled into monoterpene biosynthesis cata-
lyzed by cyclase enzymes (Alonso & Croteau 1993) resulting in metabolites including
the acyclic structures linalool and (E)-P-ocimene, the monocyclic structure limonene,
and the bicyclic structure a-pinene, all of which are released with insect feeding in
cotton. GPP can also be condensed with another IPP molecule to form (1) the C,, mol-
ecule, farnesyl pyrophosphate, the precursor of sesquiterpenes such as the acyclic
compounds nerolidol, (2) a- and p-farnesene, the monocyclic compounds humulene
and X-bisabolene and often (3) the bicyclic compound caryophyllene. All of these are
also released from cotton after insect feeding.
Almost all of the identified volatiles systemically released in response to herbivore
feeding are terpenes. At least in cotton, these systemic terpenes are early biosynthetic
products of the isoprenoid pathway. Farnesene and (E)-P-ocimene are likely formed
via an ionization-isomerization-elimination reaction, analogous to a monoterpene
syntheses proposed mechanism (Savage et al. 1994), with farnesyl- and geranyl-pyro-
phosphate as respective precursors. Linalool, another systemically released metabo-
lite, is structurally similar to ocimene except that the former compound undergoes an
additional reaction with the loss of the pyrophosphate moiety to form the tertiary al-

0 0 0 HA
oclyl Coenzyme A acetoacelyl C*A (3S)--hydroxy-3-methylglutoryl-CoA

ilopetenyl 0 _H
pyrophosphate mevolonote

monhoterptenm t .squlterpenas

myrceno Ilmonane nerolidol btisboleno

Figure 2. Biosynthetic steps in the isoprenoid pathway leading to volatile mono-
terpene and sesquiterpene formation from acetyl-CoA. Enzymes in the pathway in-
clude: acetyl-CoA acetyl-CoA transferase (A), hydroxymethylglutaryl-CoA syntheses
(B), hydroxymethylglutaryl-CoA reductase (C), mevalonate kinase, phosphomeva-
lonate kinase and pyrophosphomevalonate decarboxylase (D).

June, 1996

Behavioral Ecology Symposium '95: Pare and Tumlinson 99

cohol instead of proton elimination to form an additional double bond (see Fig. 3). In
the case of the systemically released C,, homomonoterpene, (E)-4,8-dimethyl-1,3,7-
nonatriene and C,, homosesquiterpene (3E,7E)-4,8,12-trimethyl-1,3,7,11-tridecatet-
raene appear to be formed by a series of degradations with an overall loss of four car-
bons from the sesquiterpernoid nerolidol or from the diterpenoid geranyllinalool,
Why acyclic terpenes are released throughout the plant while more biosyntheti-
cally complex derivatives are released only locally is not clear. One explanation is that
unequal enzyme activation increases the flux of terpenoid biosynthesis at the level of
mevalonic acid; however, insufficient cyclase activity limits formation of the cyclic ter-
Perhaps not surprisingly, in cotton, the systemically released terpenes are the
same compounds proposed by Loughrin et al. (1994) to be induced, and which have a
diurnal cycling pattern in their release. A pulse chase experiment using a labeled sub-
strate, such as glucose or carbon dioxide, is needed to establish whether the release of
volatiles can be divided into two chemical responses: (1) the release of stored metab-
olites at the site of wounding and (2) de novo synthesis of particular metabolites
throughout the plant.
The pathway responsible for the synthesis of the leafy green volatiles is the fatty
acid/lipoxygenase pathway. For fatty acid synthesis, C,, palmitic acid is assembled by
a series of condensation steps with malonyl-CoA serving as substrate as shown in Fig.
4. This intermediate length C, acid can then undergo elongation and/or desaturation
reactions to form long chain fatty acids (C,,-C,). The widely distributed fatty acids in
plants are monocarboxylic acids with unbranched, even-numbered carbon chains. The
unsaturated fatty acids have cis double bonds; in polyunsaturated acids, these cis
double bonds are arranged in a methylene-interrupted system precluding conjuga-
In contrast, the lipoxygenase pathway catalyzes the breakdown of lipids resulting
in short chain volatile compounds. The C, volatiles and jasmone are formed from the
polyunsaturated octadecenoid fatty acids (Z,Z)-9,12-octadecadienoic acid (linoleic

pp IonizalHon iaomerallon I lipminfon

geranyl linalyl ocimean
pyrophosphaot pyrophosphote

oppOPP Opp
O Ionizalton Isomerafoon I1 addilian I

geranyl linalool

Figure 3. A biosynthetic mechanism for the formation of ocimene and linalool from
geranyl pyrophosphate (OPP= pyrophosphate).

Florida Entomologist 79(2)

o 0 0

oeilyl CenztynI A moiClyl CA palmtic >eld

CH 00H 0 11
I*-hydrtpore yllnolnick odld l1olonit ocld

l\ily gr.n v0otll e

r,3-hixx'nI 08.041.MDI
r".- llxaif-1- z-Her-s'r- tl -'h x|nnl rosmo,

Figure 4. Biosynthetic steps in the fatty acid (A-C) and lipoxygenase pathway (D-
E). Enzymes in the pathways include: acetyl-CoA carboxylase (A), fatty acid synthase
(B), fatty acid elongation and desaturation enzymes (C), lipoxygenase (D), hydroper-
oxide dehydrase, reductase, and P-oxidation enzymes (E).

acid) or (Z,Z,Z) 9,12,15-octadecatrienoic acid linolenicc acid). The substrates for li-
poxygenase, linoleic acid and linolenic acid are major components in plant mem-
branes, but are not readily acted upon by the enzyme lipoxygenase until in the free
acid form (Croft et al. 1993). In the formation of the C, volatiles, or green leaf volatiles,
linolenic acid is oxidized to either 9- or 13-hydroperoxylinolenic acid or a mixture of
both. Hydroperoxide lyase then catalyzes the formation of the C, Z-3-hexenal and the
C,, 12-oxo-Z-9-dodecenoic acid. Rearrangement, reduction and/or esterification of the
C, product result in induced volatiles such as (Z)-3-hexenal, (E)-2-hexenal, (Z)-3-hex-
enol, and (Z)-3-hexenyl acetate, all of which have been identified in the head space col-
lections of insect damaged cotton (Rose et al. 1996).
Indole, the only volatile detected in the tryptophan pathway from plants attacked
by insects, is the penultimate intermediate in the biosynthesis of tryptophan. The
shikimic acid pathway provides the substrate for tryptophan synthesis as well as
other aromatic plant volatiles that have been identified in association with herbivore
feeding such as methyl salicylate and phenylacetonitrile (Loughrin et al. 1995, Taka-
bayashi et al. 1994a). Figure 5 includes branch point metabolites which have been es-
tablished as intermediates in the shikimic acid and tryptophan pathway leading to
the formation of indole (Crawford 1989).
In vivo labeling experiments can establish whether herbivore feeding results in
the release of stored metabolites, the synthesis of new volatile compounds, or more
likely some combination of the two. A comparison of activity for enzymes in the ter-
pene pathway before and after herbivore feeding may provide insight into why partic-
ular terpenes are or are not induced and/or systemically released from the plant.


In several crop plants, feeding by herbivores results in a time and concentration
controlled release of volatile metabolites; such volatiles can attract parasitic and
predatory arthropods which may prevent the plant from sustaining severe damage.
We propose that the combination of a signal strong enough to be detected by the par-

June, 1996

Behavioral Ecology Symposium '95: Pare and Tumlinson 101


0. H o 4: H

shlklmic acid



H 0H
3-mnoylpyrurylshklmlc acld--P

Indaolri-glycerol phosphale



OH 014

N-(5H-p phoribosyl)-
anihranllic acid

Figure 5. Biosynthetic steps in the shikimic acid/tryptophan pathway. Enzymes in
the pathway include: shikimic acid biosynthetic enzymes (A), shikimate kinase and 3-
enoylpyruvylshikimic acid 5-phosphate synthase (B), chorismate syntheses (C) an-
thranilate synthase and anthranilate phosphribosyl transferase (D), N-5'-phosphori-
bosyl-anthranilate ketol-isomerase and indole-3-glycerol-phosphate synthase (E),
tryptophan synthase F.

asitoid, a signal distinctive enough to be uniquely associated with feeding by herbi-
vores, and a signal rhythmic enough to be emitted at the time of day when the
parasitoids forage has resulted in an effective chemical cue for parasitoids and pred-
ators. The biochemical responses by plants involved in this tritrophic interaction are
only now being questioned and addressed. Many of the protocols used to study other
inducible plant responses can readily be adapted to investigate plant signalling in-
volving the release of volatiles triggered by herbivore feeding.


ALONSO, W. R., AND R. CROTEAU. 1993. Prenyltransferases and cyclases. Methods in
Plant Biochemistry 9: 239-260.
BOLAND, W., Z. FENG, J. DONATH, AND A. GABLER 1992. Are acyclic C,, and C,, homot-
erpenes plant volatiles indicating herbivory? Naturwissenswchaften. 79:368-371.
CRAWFORD, I. P. 1989. Evolution of a biosynthetic pathway: the tryptophan paradigm.
Annu. Rev. Microbiol. 43: 567-600.
CROFT, K. P., R. JUTTNER, AND A. J. SLUSARENKO. 1993. Volatile products of the lip-
oxygenase pathway evolved from Phaseolus vulgaris (L.) leaves inoculated
with Pseudomonas syringae svphaseolicola. Plant Physiol. 101: 13-24.
CROTEAU, R., AND M. A. JOHNSON. 1984. Biosynthesis of terpenoids in glandular tri-
chomes, pp. 133-185 in E. Rodriguez, P. L. Healy, I. Mehta [ed.] Biology and
Chemistry of Plant Trichomes. Plenum Publishing Co., New York.
anisms of inactivation of lipoxygenases by phenidone and BW755C. Biochem.
30: 8964-8970.




~~7 1 0.'COOH
chorismc acid

Florida Entomologist 79(2)

DICKE, M. 1994. Local and systemic production of volatile herbivore-induced terpe-
noids: their role in plant-carnivore mutualism. J. Plant Physiol. 143: 465-472.
DICKE, M, P. VAN BAARLEN, R. WESSELS, AND H. DIJKMAN. 1993. Herbivory induces
systemic production of plant volatiles that attract predators of the herbivore:
Extraction of endogenous elicitor. J. Chem. Ecol. 19: 581-599.
DONATH, J., AND G. W. BOLAND. 1994. Biosynthesis of acyclic homoterpenes in higher
plants parallels steroid hormone metabolism. J. Plant Physiol. 143: 473-478.
FARMER, E. E., AND C. A. RYAN. 1992. Octadecanoid precursors ofjasmonic acid acti-
vate the synthesis of wound-inducible proteinase inhibitors. Plant Cell 4: 129-
GABLER, A., W. BOLAND, U. PREISS, AND H. SIMON. 1991. Stereochemical studies on
homoterpene biosynthesis in higher plants; mechanistic, phylogenetic, and eco-
logical aspects. Helvetica Chimica Acta. 74: 1773-1789.
GERSHENZON, J., AND R. CROTEAU. 1989. Regulation of monoterpene Biosynthesis in
higher plants, pp.99-160 in G. H. N. Towers and H. A. Stafford [ed.] Biochemis-
try of the Mevalonic Acid Pathway to Terpenoids. Plenum Publishing Co., New
GERSHENZON, J., M. MAFFEI, AND R. CROTEAU. 1989. Biochemical and histochemical
localization of monoterpene biosynthesis in the glandular trichomes of spear-
mint (Mentha spicata). Plant Physiol. 89: 1351-1357.
GOODWIN, T. W., AND E. I. MERCER. 1990. Introduction to Plant Biochemistry. Second
ed. Pergamon Press. Oxford.
KORTH, L. K., R. A. DIXON, AND B. A. STERMER 1995. HMG-COA reductase is differ-
entially regulated in potato foliage by insect herbivory and mechanical injury.
Plant Physiol. 27: 52.
emitted by different cotton varieties damaged by feeding beet armyworm lar-
vae. J. Chem. Ecol. 21: 1217-1227.
SON. 1994. Diurnal cycle of emission of induced volatile terpenoids by herbi-
vore-injured cotton plants. Proc. Natl. Acad. Sci. U.S.A. 91: 11836-11840.
LOUGHRIN, J. H., D. A. POTTER, AND T. R. HAMILTON-KEMP. 1995. Volatile compounds
induced by herbivory act as aggregation kairomones for the Japanese beetle
(Popilliajaponica Newman). J. Chem. Ecol. 21: 1457-1467.
MANN, J. 1987. Secondary Metabolism. Clarendon Press. Oxford, England.
MATTIACCI, L., M. DICKE, AND M. A. POSTHUMUS. 1995. P-Glucosidase: an elicitor of
herbivore-induced plant odor that attracts host-searching parasitic wasps.
Proc. Natl. Acad. Sci. U.S.A. 92: 2036-2040.
MLOT, C. 1995. A clearer view of why plants make haze. Science 268: 641-642.
reagent modulates systemic signaling for wound-induced and systemin-in-
duced proteinase inhibitor synthesis. Plant Physiol. 105: 725-730.
RYAN. 1983. Regulation of synthesis and accumulation of proteinase inhibitors
in leaves of wounded tomato plants, pp. 103-122 in P. A. Hedin [ed.] Plant re-
sistance to insects ACS symposium series vol 208.
PEARCE, G., S. JOHNSON, AND C. A. RYAN. 1993. Purification and characterization
from tobacco (Nicotiana tabacum) leaves of six small, wound-inducible, pro-
teinase isoinhibitors of the potato inhibitor II family. Plant Physiol. 102: 639-
PEARCE, G., D. STRYDOM, S. JOHNSON, AND C. A. RYAN. 1991. A polypeptide from to-
mato leaves induces wound-inducible proteinase inhibitor proteins. Science
253: 895-898.
ROSE, U. S. R, A. MANUKIAN, R. R. HEATH, AND J. H. TUMLINSON. 1996. Volatile Semi-
ochemicals released from undamaged cotton leaves: a systemic response of liv-
ing plants to caterpillar damage. Plant Physiol. III (in press).

June, 1996

Behavioral Ecology Symposium '95: Pare and Tumlinson 103

SAVAGE, T. J., M. W. HATCH, AND R. CROTEAU. 1994. Monoterpene syntheses of Pinus
contorta and related conifers. J. Biol. Chem. 269: 4012-4020.
SHARKEY, T. D., AND E. L. SINGSAAS. 1995. Why plants emit isoprene. Nature 374:
1993. Analysis of lipids in the salivary glands ofAmblyomma americanum (L.):
Detection of a high level of arachidonic acid. Archives Insect Biochem. Physiol.
23: 37-52.
SON. 1995. The chemistry of eavesdropping, alarm, and deceit. Proc. Natl. Acad.
Sci. U.S.A. 92: 23-28.
1994a. Leaf age affects composition of herbivore-induced synomones and at-
traction of predatory mites. J. Chem Ecol. 20: 373-386.
TAKABAYASHI, J., M. DICKE, AND M. A. POSTHUMUS. 1994b. Volatile herbivore-induced
terpenoids in plant-mite interactions: variation caused by biotic and abiotic
factors. J. Chem. Ecol. 20: 1329-1354.
mental stages of herbivore Pseudaletia separate affects production of herbi-
vore-induced synomone by corn plants. J. Chem. Ecol. 21: 273-287.
TUMLINSON, J. H., W. J. LEWIS, AND L. E. M. VET. 1993. How parasitic wasps find
their hosts. Scientific American 268: 100-106.
patibility of resistance against herbivores and attractiveness to parasitoids in
corn varieties. Biol. Control (submitted).
TUMLINSON. 1995. How caterpillar-damaged plants protect themselves by at-
tracting parasitic wasps. Proc. Natl. Acad. Sci. U.S.A. 92: 4169-4174.
TURLINGS, T. C. J., AND J. H. TUMLINSON. 1991. Do parasitoids use herbivore-induced
plant chemical defenses to locate hosts? Florida Entomol. 74: 42-50.
TURLINGS, T. C. J., AND J. H. TUMLINSON. 1992. Systemic release of chemical signals
by hervbivore-injured corn. Proc. Natl. Acad. Sci. U.S.A. 89: 8399-8402.
TLE. 1991. Isolation and identification of allelochemicals that attract the larval
parasitoid, Cotesia marginiventris (Gesson), to the microhabitat of one of its
hosts. J. Chem. Ecol. 17: 2235-2251.
TURLINGS, T. C. J., J. H. TUMLINSON, AND W. J. LEWIS. 1990. Exploitation of herbi-
vore-induced plant odors by host-seeking parasitic wasps. Science 250: 1251-
itor in caterpillar oral secretions that induces corn seedlings to emit chemical
signals attractive to parasitic wasps. J. Chem. Ecol. 19: 411-425.

Florida Entomologist 79(2)

June, 1996


Office of the Dean for Research
University of Florida, IFAS
Gainesville, FL 32611


Naive females of Macrocentrus grandii Goidanich, a parasitoid of the European
corn borer (ECB), Ostrinia nubilalis Hubner were attracted to some, but not all, host
plant species of the latter; attraction being exceptionally strong for Zea mays L., pro-
gressively weaker for potato, snapbean and pepper, neutral for sunflower, with soy-
bean being slightly repellent. Also, extracts of volatiles from soybean were repellent.
The percent of females that responded to a given ECB host plant increased progres-
sively with age of females to a maximum at 8 days. When various plant species were
damaged by feeding activity of ECB, they became highly attractive to M. grandii fe-
males, and, subsequently, these experienced females were attracted to undamaged
plants, even of plant species that were not attractive to naive females.
Responsiveness of females to all plant species increased as a result of experience
with ECB-damaged plant specimens of host species. Such experience with ECB-dam-
aged potato increased attraction of females to sunflower by 4-fold, while experience
with ECB-damaged sunflower increased attraction to potato by 2-fold. Loss of mem-
ory in females of experience with ECB-damaged plants progressed with time and was
total by the ninth day. Memory was induced primarily by frass, but not because frass
was derived from a particular plant species. Male M.grandii were not attracted to any
host plants.
Female M. grandii emit a series of 9,13 dienes of odd-numbered 27-41 carbon atom
chains, and these, when oxidized by air, produced Z-4-tridecenal, the active volatile of
the pheromone. The latter when emitted at 50 ng per hr proved to be a very effective
lure. Further, we found that this aldehyde is synergized significantly by the lactone,
(3S,5R,6S)-3,5-dimethyl-6-(methylethyl)-3,4,5,6-tetrahydropyran-2-one. It is note-
worthy that this lactone by itself is highly attractive when 1 to 500 ng are applied to
the source. The lactone is produced by the mandibular gland of both males and fe-
males. Possibly the diene/aldehyde has arisen as an important intermediary evolu-
tionary step, in a process wherein opportunistic males that recognize volatile
chemicals unique to the female may have been selected. Natural selection then would
favor females that produce more and more of the attractant precursor a process that
could lead eventually to selection for specialized cells and a pheromone gland.

Key Words: Tritrophic, attractants, pheromone, Macrocentrus, European corn borer,
parasitoid, behavior.


Las hembras sin experiencia de busqueda de Macrocentrus grandii Goidanich, un
parasitoide del barrenador europeo del maiz, Ostrinia nubilalis Hubner, son atraidas
hacia algunas species de plants hospedantes del barrenador. La atraccion es excep-
cionalmente fuerte hacia Zemays L., progresivamente d6bil para la papa, la habi-
chuela y el pimiento, neutral para el girasol y ligeramente repelente para la soya.
Tambien los extractos volatiles de soya fueron repelentes. El porcentaje de hembras
que respondio a una plant hospedante dada aument6 progresivamente con la edad

Behavioral Ecology Symposium '95: Jones

de las hembras hasta alcanzar un maximo a los 8 dias. Varias species de plants que
habian sido danadas previamente por el barrenador se convirtieron en altamente
atractivas a las hembras de M. grandii. Estas ultimas fueron subsecuentemente
atraidas por las plants no danadas y por las no atractivas para las hembras sin ex-
La respuesta a todas las epecies de plants aument6 como resultado de la expe-
riencia de las hembras con plants hospederas danadas por el barrenador. La expe-
riencia con patatas danadas por el barrenador aument6 4 veces la atracci6n de las
hembras hacia el girasol, mientras que la experiencia con el girasol aument6 2 veces
la atracci6n hacia la patata. Lap6rdida de memorial en las hembras con experiencia en
plants danadas aument6 con el tiempo y fue total el noveno dia. La memorial fue in-
ducida primariamente por las excretas, pero no porque las excretas fueran derivadas
de una especie de plant en particular. Los machos de M. grandii no fueron atraidos
por ninguna de las plants hospederas.
La hembra de M. grandii emite una series de 9,13 dienos de numero impar de ca-
denas de carbonos de 27-41 atomos y 6stos, cuando son oxidados por el aire, produce
Z-4-tridecenal, el volatil active de la feromona. Este ultimo es un atrayente muy efec-
tivo cuando es emitido a 50 ng/hr. Mas tarde encontramos que este aldehido es siner-
gizado significativamente por la lactona (3S,5R,6S)-3,5-dimethyl-6-(methylethyl)-
3,4,5,6,-tetrahydropyran-2-uno. Es notorio que esta lactona por si misma es alta-
mente atractiva cuando de 1 a 500 ng son aplicados a la fuente. La lactona es produ-
cida por la glandula mandibular de machos y hembras. Posiblemente el dieno/
aldehido ha creado un important paso evolutivo intermediario, en un process donde
han sido selecccionados los machos oportunistas que reconocen los quimicos volatiles
unicos de la hembra. La selecci6n natural entonces favoreceria a las hembras que pro-
ducen mis y mis del precursor atrayente -un process que podria conducir eventual-
mente a la selecci6n de las celulas especializadas y la glandula de la feromona.

This presentation provides a review of what has been learned about the role of
semiochemicals in the behavior of Macrocentrus grandii Goidanich, a polyembryonic
braconid parasitoid of the European corn borer (ECB), Ostrinia nubilalis Hubner. The
work occurred over a period of about 10 years in the laboratory of the author at the
University of Minnesota. The research reported herein encompassed two areas of be-
havior the host finding behavior of the female and the mate finding behavior of the
male. The semiochemicals involved in these two areas of behavior are only part of a
complex of chemicals to which an insect is exposed in its natural habitat. Many in-
sects rely heavily on semiochemicals to assess their environment and to guide their
movements as well as to cue other aspects of their behavior. For insects such as a par-
asitoid, whose host finding involves three trophic levels, the array of significant chem-
icals is likely to be complex. These studies were initiated to elucidate the nature, role,
and interrelationships of semiochemicals in the tritrophic interactions of M. grandii,
its European corn borer host, and European corn borer host plants.
The initial studies focused on host finding; the results have been published by
Ding et al. 1989a, 1989b, Udayagiri & Jones 1992a, 1992b, and 1993. One of the first
aspects of host finding behavior addressed was the range of host plants to which M.
grandii was attracted and to what extent this group of plants paralleled the host
range of 0. nubilalis. We proposed that tests in a y-tube olfactometer, with only air as
an alternative, would reveal that M. grandii would be attracted to almost any green
plant, regardless of species. This hypothesis proved false, but a very wide range of
plants across several families were attractive, including Gramineae, Cyperaceae, Ro-
saceae, Leguminoceae, Aceraceae, Malvaceae, Asclepiadaceae, Solanaceae and Com-

Florida Entomologist 79(2)

positae. Surprisingly, the most consistent attraction was among the Compositae, even
more so than among the Gramineae, although the strongest attraction by far was to
Zea mays L. A few species repelled the females. This occurred within the Pinaceae,
Lemnaceae, Liliaceae, Tiliaceae, and Oleaceae families. There was no evident correla-
tion of attractiveness to M. grandii with 0. nubilalis recorded host plants.
Next, we investigated the role of learning in the attraction of M. grandii females
to plants. Learning is a well demonstrated phenomena among insect parasitoids, but
we were interested in the degree of discrimination by the female among closely re-
lated plants, the learning response to non-attractive plants, the existence of cross-at-
traction induced by learning, and the rate of forgetting. Naive female M. grandii are
strongly attracted to sweetcorn, somewhat attracted to potato, neutral to sunflower,
and very slightly repelled by soybean. Naive females were attracted to all four plants
when the plant had been damaged by the European corn borer.
Prior exposure of female M. grandii to plant parts damaged by ECB larvae signif-
icantly and markedly increased responses by the female M. grandii to undamaged
plants in the y-tube olfactometer. This was true even with sunflower and soybean. For
example, experience increased attraction to potato three-fold and to sunflower four-
fold. The overall level of responsiveness increased to all plant species as a result of ex-
perience with one European corn borer damaged plant species. Experience with po-
tato increased attraction to sunflower four-fold and experience with sunflower
increased the response to potato two-fold. Females experienced on potato or sunflower
also strongly preferred those plants to sweet corn when offered a choice of the two
plants. Similar results were obtained when leaves or stems were tested. Females also
were able to discriminate between varieties of sweet corn, as a learning experience on
one variety induced preference for that variety.
M. grandii females did forget the learning experience. Over a period of nine days
post-experience, attraction to sunflower decreased to insignificance. It should be
noted that the experience provided exposure to European corn borer damaged plants
only. It did not allow for host finding, nor for parasitism. Thus, this learning does not
involve a "reward" in the classical sense or a consummatory act.
Further studies were conducted to identify the specific components of the host-
damaged plants that cued the learning response in the female. These studies identi-
fled the frass as the key component, but not because the plant material contained in
the frass was derived from a particular plant species. For example, exposure of M.
grandii females to sunflower in combination with frass collected from ECB larvae
that had fed on sweetcorn, induced attraction to sunflower in the female, whereas ex-
posure to sunflower alone does not.
Because the above studies were conducted in a y-tube olfactometer and, therefore,
assayed a walking response, it was of interest to study female behavior in flight. Con-
sequently, a series of experiments were conducted in a small flight chamber provided
with an air flow. These studies were conducted with plant volatile extracts, and they
revealed similar responses to those seen in the y-tube. Naive female M. grandii were
most strongly attracted to sweetcorn, with decreasing attraction to potato, snapbeans,
pepper, and soybean. For example, on average, 9 of 10 females landed on the sweet-
corn source, 5 of 10 on potato, 2 of 10 on snapbean, 1.5 on pepper and 1 on soybean.
In all tests, responses were <1 on the control source. Female response to corn in-
creased over time when females were held unexposed. This response ranged from very
little (<1/10 females) at one day of age to a maximum response (8/10 females) at eight
days. This phenomena of lower threshold with increasing physiological deficit is com-
monly reported in hunger and oviposition studies. Male M.grandii were unresponsive
to all plants in these studies.

June, 1996

Behavioral Ecology Symposium '95: Jones

The flight studies also confirmed the strong influence of experience on host choice.
Bioassays of mixtures of plant volatile extracts showed that soybean significantly re-
duced female response to sweetcorn, whereas pepper did not. This strongly indicates
a repellent chemical contained in soybean. In spite of this, experience with European
corn borer damaged soybean induced females to respond positively to soybean vola-
tiles in the flight chamber. Further studies indicated that the leaf and husk of corn
contained the most active volatiles attractive to the female. These chemicals could be
trapped, partially purified and still retain their activity.
Our studies revealed that M. grandii has a pliable behavioral system that no
doubt allows it to adapt to the several plant hosts of its host, the European corn borer.
Although numerous studies have revealed the multiple roles of chemicals in the
mate-finding and mate acceptance behavior of insects, very little similar information
is extant for insect parasitoids. In addition to the contribution that such studies also
could make to our understanding of insect behavior and its evolution, this information
could lead to useful survey tools for parasitoids. Consequently, we pursued the eluci-
dation of the M. grandii mating behavior scheme. These studies have been reported
by Swedenborg & Jones 1992a, 1992b, Shin et al. 1993, Swedenborg et al. 1993, Swe-
denborg et al. 1994. Male M. grandii were shown to respond positively to females and
extracts of females in a flight chamber by exhibiting the common behaviors of upwind
anemotaxis, hovering, landing, wing fanning, and copulatory motions. These behav-
iors were used as the assay tool for identification of the pheromone components in-
The first studies demonstrated that a series of dienes contained in the female cu-
ticle elicited the mating behaviors observed in the males. It was also evident that at
least two more polar components were active. Synthesis of the hydrocarbons con-
firmed that a series of 9,13 dienes of odd numbered 27-41 carbon atoms were active.
We later demonstrated that the activity was due to air oxidation of the dienes to pro-
duce Z-4-tridecenal, the active volatile component of the pheromone. This compound,
as well as the dienes, proved to be active in attracting males in the field. In slow re-
lease Hercon wafers, a release rate of 50 ng per h was a very effective lure as it at-
tracted twice as many males as were attracted by three virgin females. E-tridecenal
was synthesized, tested, and shown to be inactive, and it was without inhibitory ef-
fects when added to Z-4-tridecenal. Further studies indicated that both the dienes
and the aldehyde were significantly synergized by a polar compound found in the ac-
tive extracts of females. Addition of this component to Z-4-tridecenal increased trap
catch in the field two-fold. Extensive analyses and bioassays revealed this chemical to
be (3S,5R,6S)-3,5-dimethyl-6-(meth(methyl)-3,4,5,6-tetrahydropyran-2-one. It and its'
3R,5S,6R enantiomer were synthesized and tested. The combination of Z-4-tridecenal
and the lactone proved to be a powerful lure for males in the field.
Laboratory bioassays revealed that the lactone elicited upwind anemotaxis. Males
demonstrated a positive dose response to the lactone at levels of 1 ng to 500 ng applied
to the source. The RSR enantiomer proved to have very slight activity, but the SRS
enantiomer was 15-fold as active. When added to Z-4-tridecenal (50 ng per h release
rate) in the field, 10 micrograms of the SRS enantiomer compared favorably to mix-
tures using the female derived materials.
The source of the lactone is the mandibular gland of both males and females. Be-
cause it occurs in the male, its primary function as a pheromone is not certain. It per-
haps serves as an aggregation pheromone in the male, but this could not be
demonstrated. Both male and female contain 100-300 ng of the lactone per head.
These studies represent the second report of the role of an oxidation product of hy-
drocarbon dienes in the pheromone scheme of an insect. A similar account was docu-

Florida Entomologist 79(2)

mented by Bartelt & Jones (1983) for another hymenopteran, Pikonema alaskensis.
Because, as far as we know, the aldehyde activity is elicited entirely by the products
of air oxidation of these dienes, there appears to be no specific gland and, conse-
quently, no control of pheromone release. In both species, a potent synergist is in-
volved, the control of which is unknown. One can readily foresee an important
intermediary evolutionary step here, as this represents a case wherein opportunistic
males that recognize volatile chemicals unique to the female may have been selected.
Natural selection then would produce females that produce more and more of the at-
tractant precursor a process that could lead to selection for specialized cells and a
pheromone gland.


BARTELT, R. J., AND R. L. JONES. 1983. (Z)-10-nonadecenal: a pheromonally active air
oxidation product of the (Z,Z)-9,19 dienes in the yellowheaded spruce sawfly. J.
Chem. Ecol. 9: 1333-41.
DING. D., P. D. SWEDENBORG, AND R. L. JONES. 1989a. Chemical stimuli in host-seek-
ing behavior of M. grandii. Ann. Entomol. Soc. America 82: 232-36.
DING, D., P. D. SWEDENBORG, AND R. L. JONES. 1989b. Plant odor preferences and
learning in M. grandii (Hymenoptera: Braconidae), A larval parasitoid of the
European corn borer, Ostrinia nubilalis (Lepidoptera: Pyralidae). J. Kansas
Entomol. Soc. 62: 48-61.
1993. A sex pheromone isolated from Macrocentrus grandii: Absolute stere-
ochemistry determination and implications. J. Org. Chem. 58: 2923-2926.
SWEDENBORG, P. D., AND R. L. JONES. 1992a. Multicomponent sex pheromone in Mac-
rocentrus grandii Goidanich (Hymenoptera: Braconidae). Jour. Chem. Ecol. 18:
SWEDENBORG, P. D., AND R. L. JONES. 1992b. (Z)-4-Tridecenal, a pheromonally active
air oxidation product from a series of(Z,Z)-9,13 dienes in Macrocentrus grandii
Goidanich (Hymenoptera: Braconidae). Jour. Chem. Ecol. 18: 1913-1931.
(3R*,5S*,6R*)-3,5-dimethymethylethylethyl)-3,4,5,6-tetrahydropyran-2-one, a
third sex pheromone component for Macrocentrus grandii (Goidanich) (Hy-
menoptera: Braconidae) and evidence for its utility at eclosion. Jour. Chem.
Ecol. 19: 485-502.
1994. Biological activity of (3R,5S,6R)- and (3S,5R,6S)-3,5-dimethyl-6-(methyl-
ethyl)-3,4,5,6-tetrahydropyran-2-one, a pheromone of Macrocentrus grandii
(Goidanich) (Hymenoptera: Braconidae). Jour. Chem. Ecol. 20: 3373-3380.
UDAYAGIRI, S., AND R. L. JONES. 1992a. Flight behavior of Macrocentrus grandii
Goidanich (Hymenoptera: Braconidae), a specialist parasitoid of European
corn borer (Lepidoptera: Pyralidae): factors influencing response to corn vola-
tiles. Environ. Entomol. 21: 1448-1456.
UDAYAGIRI, S., AND R. L. JONES. 1992b. Role of plant odor in parasitism of European
corn borer by braconid specialist parasitoid Macrocentrus grandii Goidanich:
isolation and characterization of plant synomones eliciting parasitoid flight re-
sponse. Jour. Chem. Ecol. 18: 1841-1855.
UDAYAGIRI, S., AND R. L. JONES. 1993. Variation in flight response of the specialist
parasitoid Macrocenrus grandii Goidanich to odours from food plants of its Eu-
ropean corn borer host. Entomol. exp. appl. 69: 183-193.

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.


'Department of Entomology
Texas A&M University
College Station, TX 77843-2475

'Department of Environmental Science, Policy, and Management
University of California
Berkeley, CA 94720


The reproductive biology of some plants is based on the movement of pollen be-
tween plants by insects. In many plant-insect pollination systems the plant produces
a "reward" usually in the form of nectar. However, a number of plant taxa produce oil
as a floral reward. Bees of the genus Centris are known as "oil-collecting" bees which
are the important mediators in the reproductive success of oil reward producing flow-
ers. These bees are solitary and their own reproductive success depends on the inter-
action between the sexes and the ability of the female to construct and provision a
nest. In this paper we discuss the importance of male territorality and its mainte-
nance to the reproductive success of this taxa of bees. We also discuss the nesting bi-
ology of Centris in Costa Rica focusing on the resource needs of females.

Key Words: Centris reproductive biology, oil-collecting bees, male territorial behavior,
bee nests.


La biologia reproductive de algunas plants esta basada en el movimiento de polen
entire plants llevado a cabo por insects. En muchos sistemas de polinizaci6n planta-
insecto la plant produce un "agradecimiento" usualmente en forma de nectar. Sin
embargo, cierto numero de taxa de plants produce aceite como agradecimiento flo-
ral. Las abejas del g6nero Centris son conocidas como colectoras de aceite y son impor-
tantes mediadores en el exito reproductive de las flores productoras de aceite. Estas
abejas son solitarias y su propio exito reproductive depend de la interacci6n entire los
sexos y la habilidad de la hembra de construir y aprovisionar un nido. En este articulo
discutimos la importancia del mantenimiento de la territorialidad del macho en el
exito reproductive de estos taxa de abejas. Tambien discutimos la biologia de nidifica-
ci6n de Centris en Costa Rica haciendo hincapie en la necesidad de recursos de las

Pollen and nectar have long been recognized as important nesting resources for so-
cial and solitary bees (Baker & Baker 1975, Baker & Hurd 1968). As a result, female
bees are essential to the reproductive biologies of many plants (Kevan & Baker 1983).
However, some plants do not produce nectar, but produce lipids (Vogel 1969, 1974,
1976a, Seigler et al. 1978, Simpson et al. 1990). Solitary bees of the genus Centris,

Florida Entomologist 79(2)

which we have referred to as "Oil Baron Bees" (Vinson & Frankie 1991), are important
pollinators of plants which produce lipid floral rewards rather than nectar (Buck-
mann 1987, Vinson et al. 1996, Vogel 1976b, 1981, 1986, 1988, 1990).
The floral oil reward pollination system was first described by Vogel (1974) and in-
volves a number of plant species (Simpson 1989, Simpson & Neff 1981, Simpson et al.
1979), several of which are members of the family Malpighiaceae. Members of this
family have glands on the abaxial side of the sepals called "elaiophores" (Fig. 1) that
produce the oils (Anderson 1979). The oils are collected by a number of bees (Albu-
querque & Rego 1989, Rego & Albuquerque 1989, Neff& Simpson 1991, Vinson et al.,
1996), including species of the genus Centris (Hymenoptera: Apidae) which are con-
sidered important pollinators of these oil producing plant species (Simpson & Neff
1981, Simpson et al. 1977, 1990).
Although many species of Malpighiaceae are vines or shrubs (Gentry 1993), Byr-
sonima crassifolia (L.) D. C. is a moderate sized tree which produces a fruit known as
Murici (Brazil) or Nance (Costa Rica) eaten by some people in Northern and North-
eastern Brazil and in Central America (Braga 1976, Camargo & Mazucato 1984,
Anderson 1983). Byrsonima crassifolia occurs from Paraguay to Mexico and may be
composed of closely related segregates (Anderson 1978). The flowers of the Malpighi-
aceae are conservative among the species (Anderson 1979).
Of the 68 described species of Centris, 35 are known to occur in Costa Rica (Snel-
ling 1974, 1984) of which 15 species have been collected in the Guanacaste Providence
in dry forest region of Costa Rica (Snelling 1984, GWF & SBV unpublished data).
Eight of these species are ground nesters and seven nest in pre-existing tree cavities
made by other insects (Frankie et al. 1989; Table 1). Regardless of nesting site these


Figure 1. A diagram of a flower of Byrsonima crassifolia showing the location of the
five pairs of elaiophores which produce an oil collected by several species of oil collect-
ing bees. One petal, i.e., the flag petal, is grasped by the mandibles freeing the legs of
the bee and allowing it to scrape the oil from the elaiophores.

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.


C) C)

= r
cc cc
M a


a a 9 c c


cc C






S ewe*e

1b bb


C C0


QQ c




^ ,-

a %. 6

-E IIlt

". "; "
^ a a



E ,,

I- ^|061
g < a g 0
& !.sI 1

a gi ioQO


'IS o

8 Sa .S: .S.S
t^ ^ t ww
o' oa s
e 5 oUUu

^ '- > c 0 n &

Rcc bfl
ea: e c, e

d0 d03 d3

Florida Entomologist 79(2)


c. cc


4 o
















June, 1996





-f I

II.x ~

a %s~Az

II, iiv II i
" 0 n0
4 mE




S >

-9 $-, $
11- ~0


U W ^ c-.


06 06

to 0
to c .

lA~ I?1 I? c^.


s ll

Behavioral Ecology Symposium '95: Vinson et al.

bees require sources of nectar, oil, and pollen for their reproductive success and, there-
fore, are important pollinators of about 50% of all dry forest flowering plants that oc-
cur in the seasonal dry forest life zone (Frankie 1976, Frankie et al. 1976, 1990,
Vinson et al. 1993). We have been interested in the chemical ecology of these bees fo-
cusing on two aspects, male territorality and female nesting.


Several reproductive strategies are employed by males to gain access to females
(Alcock 1979, Emlen & Oring 1977). Among the acelate hymenoptera, these include
female (nest) defense polygyny, resource defense polygyny, and landmark defense po-
lygyny (Marshall & Alcock 1981, Eickwort & Ginsberg 1980). The latter, analogous to
lek polygyny (Alcock & Smith 1987), generally involves a male aggressively defending
a small area, driving off other males and releasing chemicals that both delineate the
males space or "territory" and are, presumably, attractive to virgin females (Minckley
et al. 1991, Frankie et al. 1980).
Territories of Centris vary considerably in their size, plant association, and male
marking activity between species, although they have many features in common (Rau
1975, Frankie et al. 1989). Three species in Guanacaste, Costa Rica, C. adanae Cock-
erel, C. trigonoides Lepeletier, and C. lutea Friese defend territories in grass. C.
adanae territories are usually less than half a meter above the ground in pockets of a
shorter species of grass in which a small stick, shrub, or rock located near the center
can serve as a perch (Frankie et al. 1980). The territorial area defended (Fig. 2) by C.
adanae is about 3-4 m2 although most of the activity occurs within a 1-2 m2 space.
Marking is frequent within the inner marking area and involves volatile compounds
released by the mandibular glands (Table 2). In contrast, C. lutea establishes territo-
ries between 0.5 and 2 meters high at the tips of emergent jaragua grass, Hyparrhe-
nia rufa (Mees.), an introduced species from Africa (Parsons 1976, Pohl 1983). These
territories may encompass a defended 20-30 m2 elliptical area with males perching on
a grass seed head or shrub close to the upwind side of the defended area. Marking is
less frequent and involves compounds released by glands located in the hind leg
(Frankie et al. 1989; Table 3). Males of C. lutea have been observed on perches with
their hind leg elevated so that the tibia and tarsi stick up (Fig. 3). This may provide a
means of increasing the evaporation of compounds released by the tibial gland located
in the hind leg in this species. Territories of C. trigonoides tend to be intermediate in
the area patrolled (15 m ), and males tend to perch on twigs of leafless bushes that
emerge above the jaraqua grass (2-3 m), or they defend territories in B. crassifolia ir-
respective of whether the tree is in flower; these males moderately mark (Frankie et
al. 1989). This species also has leg glands (Table 3) and spends more time hovering,
often with its hind legs extended down (Fig. 4).
Territories of C. flavifrons Fabricius, C. inermis = segregata Friese, and C. nitida
F. Smith (Table 1) tend to occur in downwind depressions in the canopy of B. crassifo-
lia whether in flower or not, while C. vittata Lepeletier prefers the leguminous tree
Cassia grandis Lb. Centris flavofasciata Friese males defend territories in coastal
strand vegetation of Ipomoea pes-caprae (L.) R. Br. and Canavalia maritima (Aubl.),
two species often referred to as Beach Morning Glory that are commonly intermixed
and are similar in appearance (Wilson 1983). In all of these cases, the territories are
initiated and defended during the morning. Males depart usually around noon. An ex-
ception is C. aethiocesta Snelling. A few C. aethiocesta male territories have been re-
corded in several species of small unidentified trees at the edge of the coastal strand.
Male territorial behavior begins in the early morning soon after sun rises as is typical

Florida Entomologist 79(2)

Figure 2. General territorial area of Centris displaying a form of lec defense polyg-
yny reproductive strategy (based on male territories of C. adanae, C. lutea, and C. fla-
vifrons). Distances depend on species and shape may be modified by local physical
parameters. P = perch where the male may rest usually facing down wind. Area
marked refers to the area where substrate marking may occur with the dashed area
indicating where marking is more likely. Defended area is based on area where male
will fly out to investigate an intruding bee-sized flying insect.

of the Centris species in Guanacaste. The C. aethiocesta territories were located on the
ocean side of the bush as the breeze comes off the land. At about noon the males leave
when the breeze usually stops. The breezes begin again about an hour later, but this
time they come from the ocean side. Interestingly, the male bees return with the de-
fended area of the territory on the land side of the bush. The territory is subsequently
maintained for about three hours (SBV, unpublished data).
Although most of the male Centris prefer to defend territories in only one or two
species of plants (Table 1), several species of Centris utilize a wide variety of plants.
For example, C. heithausi Snelling can be found in downwind depressions in the can-
opy of C. grandis, Myrospermum frutescens Jacq. (Frankie et al. 1989), Gliricidia
sepium (Jacq.) Steud., Piscidia carthagenensis Jacq., Cassia emarginata L., Securi-
daca sylvestris Schlecht, and Cochlospermum vitifolium (Willd) Spreng (Coville et al.
1986). However, most territories occur in G. sepium when it is in flower and then shift

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.


Species Chemical % Total

C. adanae Geraniol 80.0
Geranyl acetate 17.2
Nerol 0.8
Ethyl laurate 0.6

C. flavifrons Geraniol 77.0
Geranial 10.1
Neral 7.2
Geranyl acetate 2.3
Nerol T'

C. flavofasciata Neral 56.7
Geraniol 17.2
Geranyl acetate 16.4
Geranial 9.7

C. aethiocesta Neral 57.8
Geraniol 15.8
Nerol 13.4
Linalool 9.8

C. inermis Geraniol 73.7
Geranial 15.3
Neral 6.7
Nerol 1.7

C. inermis Geraniol 45.2
(segregata morph) Geranial 42.9
Neral 9.2
Nerol 2.6

T = Trace.

to the other plants which come into flower as flowering ofG. sepium declines. Marking
is not common in C. heithausi Snelling (Table 1).
Centris aethyctera Snelling also defends territories in a number of different tree
species (Table 1), usually during flowering. No marking by C. aethyctera has been ob-
served (Frankie et al. 1989). The observations suggest that both C. aethyctera and C.
heithausi have evolved to patrol territories where they can intercept females seeking
nectar. Centris fuscata Lepeletier may also employ a similar strategy since it rarely
marks, but C. fuscata tends to defend in either of two important pollen resources dur-
ing their blooming period (Frankie et al. 1989). As shown in Table 1, the males of each

Florida Entomologist 79(2)

AL. 1984).

Species Chemical % Total

C. heithausi Pentacosene 37.0
Nonadecadiene 21.0
Heptadecene 15.0
Tetradecenyl acetate 9.0
Pentacosane 5.0
Heptadecane 3.0
Heptacosane 3.0
Other hydrocarbons (4) 6.0

C. nitida 2-Pentadecanone 53.0
2-Heptanone 32.0
Tetradecyl acetate 9.6
Ethyl myristate 3.7
Ethyl palmitate 1.7
Hexadecyl acetate T'

C. trigonoides subtarsata 2-Pentadecanone 53.0
Hexadecyl acetate 34.9
Ethyl myristate 8.6
Ethyl palmitate 3.5

T = Trace.

species can be recognized based on their plant association, size of territory, and their
territorial behavior. In spite of similar preferences for certain plants, such as B. cras-
sifolia or grass, territories rarely overlap (Frankie et al. 1989).
The territorial pheromones of the species employing a landmark defense strategy
are produced in either mandibular glands or glands located in the hind legs (Frankie
et al. 1989). As can be seen in Table 2 and 3, each species has its own pheromone blend
but those having mandibular glands have terpenes, alcohols, aldehydes, and esters
while those with leg glands tend to be composed of hydrocarbons, ketones, and fatty
acid esters. Males of C. analis Fabricius and C. bicornuta Mocsary are not included in
Table 2 because they are difficult to distinguish and variation in our samples has sug-
gested a mixture of species, thus this data is not provided. Data for C. lutea is based
on only one specimen and is not included. Male territories ofC. dichrootricha (Moure)
have not been found.
The territorial chemicals from C. inermis and C. segregata (Vinson et al. 1984) are
of particular interest because both male and female C. segregata are distinct color
morphs of C. inermis and were considered a separate species until Snelling (1984) rec-
ognized C. segregata as a synonym of C. inermis. We observed that territories of C. in-
ermis in Byrsonima in the early dry season were often replaced by the segregata
morph of C. inermis later in the dry season. We also reported the chemistry differed

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.

Figure 3. Diagrammatic representation of Centris lutea perched on a grass head
with his hind leg extended up into the air stream. The hind leg is the source of a ter-
ritorial marking material.

Florida Entomologist 79(2)

Figure 4. Territory of a male Centris trigonoides in Byrosonima showing the exten-
sion of the hind leg (arrow) where a pheromone is released while hovering.

in the two morphs, but primarily in the percent composition of the two major com-
pounds (Vinson et al. 1984), rather than different compounds (Table 2).
In many Centris species, such as C. inermis, C. flavifrons, and C. flavofasciata in
Costa Rica and C. pallida W. Fox in the southwest United States, metandric forms
(sometimes referred to as "beta" males) occur. These males are larger than the territorial
males and appear to patrol nest sites where they have been reported to wait for females
and to mate as they emerge or to even attempt to dig them up to mate (Alcock et al.
1976a, 1977, Chemsak 1985, Toro et al. 1991). In C. pallida beta males have been ob-
served to be successful while the smaller alpha males defending territories in bushes
nearby rarely appear to intercept females (Alcock et al. 1977, Rozen & Buchmann 1990).
Both the beta male strategy of searching for and digging up emerging females and
the alpha strategy of releasing pheromones and defending territories, appear to be en-
ergy intensive (Vinson et al. 1982). As shown by Frankie et al. (1980), the nectar con-
tent in the crop of male C. adanae significantly decreased during the time males spent
in the territory and the crop was nearly empty when these males abandoned their ter-
ritorial activity around noon. Males of all the Centris species in the area can be col-
lected from a number of nectar sources in the early afternoon, including the plant
species indicated in Table 4. Thus, male Centris are also important pollinators of these
massively blooming nectar reward trees.

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.


Centris Needs Source Function

Nectar' Ai, Dr Pr, My For maintenance. For nest
Se, Tb, CI, Cr provisions (some species).

Pollen' Ca, Cb, Ce, Cg, For nest provisions.
Co, Cu

Oil I' By For nest provisions (some
species). For sand collec-
tion. For wood chip collec-

Water (some situations) Ponds and streams For sand collection. For
wood chip collection.

Sand River banks, For cell wall construction
wind rows (some species).

Wood chips Wood boring insects For cell wall construction
(some species).

Resin I ? For cell wall construction.

Oil II ? For nest defense (some spe-

Resin II ? For nest defense (some

See Table 1 for plant names.


In studies of the chemical ecology of nesting, there are several questions to be ex-
plored. First, nests must be located and the nesting substrate and morphology of nests
described. Secondly, the source of the materials used to construct the nest and cell(s)
must be determined. Thirdly, the resources collected or produced to provision the cells,
to mark the nest and for use in nest defense must be determined. However, these
questions are particularly challenging due to the known diverse nesting habits and
reproductive strategies of the genus (Michener & Lange 1958, Coville et al. 1983).
The diversity in nesting habits ranges from the production of solitary nests where
only one cell is provisioned as occurs in C. adanae (Frankie et al. 1989), through soli-
tary nesting with provisioning of several cells as occurs in C. fusciata (Raw 1984), to
aggregated nesting with several cells per nest as occurs in C. pallida (Alcock et al.
1976b, Rozen & Buchmann 1990). Nests may occur in flat ground or earthen banks
(Vinson & Frankie 1977, Batra & Schuster 1977) or females may utilize existing holes
in wood, but with specific preferences for certain habitats (Frankie et al. 1988). Other
species nest in termite nests (Bennett 1964, Pickel 1928) or in the tumulus ofAtta ant
nests (Vesey-Fitzgerald 1939). Centris flavifrons prefers to initiate a nest in a depres-
sion in the soil and is frequently found to have nest entrances in the side of burrows
excavated by Ctenosaura similis (Vinson & Frankie 1988), a common iguana lizard
that also nests in the soil during the dry season (Fitch & Harkforth-Jones 1983). Thus,
the location of the nests of a particular species, for the first time, is often a product of

Florida Entomologist 79(2)

chance and patience. Further, Centris show some adaptability. For example, C.
aethyctera usually excavates a single linear tunnel in which 3-6 cells, one on top of the
next, are provisioned (Vinson & Frankie 1977), but in rocky soil or under poor re-
source conditions the nest may consist of several cells together in a pocket in the soil
or only one or two cells (Vinson & Frankie 1991).
When we began these studies, there was no data on the nesting of any of the spe-
cies in Costa Rica. Within a few years we had discovered nests of six of the eight spe-
cies which we now know nest in the soil, but had not yet determined the rearing
habits of C. inermis and C. segregata. In 1981 we discovered a mixed nesting aggrega-
tion of C. segregata and C. inermis in a river bank. The nests of each species were
marked, the nests described, and some of the cells from each were collected and trans-
ported to the laboratory. All the collected cells yielded C. segregata even though some
had been clearly provisioned by C. inermis (Coville et al. 1983). These results led Snel-
ling (1984) to reevaluate the characters of these two species and to synomized C. seg-
regata with C. inermis.
There were indications that some Centris would nest in cavities in wood (Jayas-
ingh & Freeman 1980, Kimsey 1978), and even though we had looked in dead trees,
we still lacked nests for seven species. In 1980 we discovered a C. vittata nesting in a
cavity in a tree stump. Although C. vittata had been observed to nest in existing holes
in earthen banks, and species ofHemisiella and Heterocentris appeared to adopt a va-
riety of cavities as nest sites (Coville et al. 1983), this was our first evidence that some
of the Costa Rican species did nest in wooden cavities. This led to an effort to produce
artificial nesting sites in which holes drilled into wood would be tested to determine
if some species could be induced to nest. The technique was successful and tests in
1981 using nest blocks with holes of different sizes produced nests of the remaining
seven species (Table 1).
All the members of the subgenera Centris and Trachina in Costa Rica nest in soil,
but each species has different requirements (Frankie et al. 1989, Table 1). However,
all initially form tunnels of a particular length depending on the soil and species. For
example, C. aethiocesta excavates a 6 cm tunnel in sand while C. flavifrons tunnels
may be 108 cm long in sandy clay loam (Vinson & Frankie 1988). Further, depending
on species, one or more cells may be formed with the first placed at the end of the tun-
nel and with the rest placed sequentially on top. The cells consist of a waxy-resinous-
like material embedding several millimeters of the surrounding soil. These cells are
provisioned with pollen filling about one-third to one-half of the cell. A liquid is then
added covering the pollen to a depth of 3-4 millimeters. Then an egg is oviposited
which floats on the liquid. The cell is capped with what appears to be the same resin-
waxy material mixed with the surrounding soil. When the appropriate number of cells
have been provisioned, the female fills the remaining tunnel with loose soil (Vinson &
Frankie 1977, 1988, 1991, Vinson et al. 1987).
The remaining subgenera nest in cavities in wood which are probably produced by
wood boring insects. Each species prefers a certain size entrance hole (Frankie et al.
1988, Table 1) and habitat (Frankie et al. 1988, 1993). The cavity nesting species can
be separated further on the basis of the material used to build the cells and to fill
spaces between cells. The three species of the subgenus Hemisiella use sand while
both the subgenera Heterocentris and Xanthemisia use wood chips. Several of these
wood cavity nesting species leave an oily or resinous deposit near the nest entrance
(Fig. 6) after the last cells has been completed (Frankie et al. 1988, 1989, Table 1).
From the above discussion it is clear that female bees must collect a number of re-
sources. These include nectar for their own energy; a liquid to mix with the soil to con-
struct the cells, or in the case of some wood cavity nesters, to mix with wood chips or
sand; pollen to provision cells; a liquid to add to the pollen provisions; and resins and

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.

oils to seal the entrance of nests. However, the source and nature of some of these ma-
terials is not clear (Table 4).


Pollen is an essential provision providing the protein required by the developing
bee. Several plants that bloom during the dry season are known to be important ex-
cess pollen producers (Table 4). These include Cochlospermum vitifolium, C. grandis,
C. biflora L., C. emarginata, Curatella americana L., and B. crassifolia (Frankie et al.
1983, 1989), although which species are used most frequently by the different bee spe-
cies as a pollen resource has not been determined.


Centris have been referred to as oil collecting bees and all the 16 species found in
the dry forest life zone of Guanacaste have been observed collecting oil from Byroson-
ima (Vinson et al., 1996). Vogel (1974) first postulated that floral oils could be used in-
stead of nectar for larval development, but gave no evidence. Simpson et al. (1977)

Figure 5. A wooden bee trap nest block showing the entrance to the completed nest
of Centris bicornuta plugged with a "mayonnaise-like" oil (arrow).

Florida Entomologist 79(2)

reported that the nest cell provisions of C. trigonoides, a cavity nester, consisted of pol-
len and floral oil with no appreciable carbohydrate. Neff & Simpson (1981) reported
that another species, C. malcutifrons, in Peru provisioned their cells with pollen and
oil, along with "appreciable" amounts of nectar. This led Neff & Simpson (1981) and
Simpson et al. (1990) to question the absence of nectar in the nest provisions of Cen-
To determine whether the Centris in Costa Rica use only oil, we first examined the
oil produced by B. crassifolia, and we found two different oil types. Most of the trees
we sampled were found to produce a mixture of mono-, di- and tri-glycerides and free
fatty acids (Vinson et al., 1996). The mono-glycerides composed over 50% of the oil
with the di-glycerides being the second most abundant compounds. The tri-glycerides
made up only 10% while the remaining 5% consisted of a mixture of free fatty acids.
The fatty acids released from the glycerides by de-esterification consisted of saturated
C16-20 and unsaturated C18-22. A nearly identical glycerol ester and fatty acid com-
position was found for the liquid in cells of C. aethectera, C. flavifrons, C. flavofasciata,
and C. adanae, the only four ground nesting species examined (Vinson et al., 1996, Ta-
ble 5).
The oil from some B. crassifolia trees contained an abundant unknown that was
considerably more polar than the glycerides reported above and yielded two uniden-
tified GC peaks on de-esterification. No bee cells containing this lipid pattern were
found, but the sample size was limited (Vinson et al. 1996). Whether only certain trees
produce this oil with the unknowns, or the same tree produces both types but at dif-
ferent times or under different conditions, is now under study.
We recently collected the liquid contents of five cells of C. bicornuta, a wood cavity
nester, just after completion and before the embryo hatched. Unlike the oil from B.
crassifolia or the liquid content of the cells of the ground nesters, the liquid from the
C. bicornuta nest cells was not soluble in hexane and no lipids were detected. How-
ever, the liquid consisted of carbohydrates (SBV, GWF, & HJW, unpublished data).
Thus, C. bicornuta uses nectar along with pollen to provision their cells (Table 5).
Since C. bicornuta has been observed collecting oil, its use is a major question. In
the process of opening many C. bicornuta nests, it has been observed that wood chips
fill spaces between cells, and in bigger diameter cavities the space between the cells
and the cell wall also consisted of wood chips. These chips were sticky and the sticky
material was soluble in hexane. An analysis by TLC (Vinson et al. 1996) revealed a
lipid pattern, when exposed to I, vapor, similar to C. crassifolia oil, but with several
additional spots, depending on the sample. These additional trace compounds, which


Species Provision Composition

Ground nesters:
C. adane Pollen Oil
C. flavifrons Pollen Oil
C. flavofasciata Pollen Oil
C. aethyctera Pollen Oil

Cavity nesters:
C. bicornuta Pollen Nectar

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.

remain unidentified, varied and may come from the different woods that are the
source of the wood chips. The collection and transport of sand to their nests by C. nid-
ita, C. trigonoides (Fig. 7), and C. vittata on their legs was reported by Vinson et al.
(1993). The transported sand around and between the cells of C. nidita was also found
to be sticky in most nests, and the sand from these nests rinsed in hexane yielded a
TLC pattern nearly identical to B. crassifolia oil. We suggest that females use the oil
to stick the wood chips or sand to their legs to aid in the transport of these materials
to the wood cavity.
In a few nests, neither the wood chips or sand were sticky and no oil could be de-
tected. We (SBV & GWF, unpublished data) have observed C. nidita and either C. ana-
lis or C. bicornuta (impossible to tell apart unless captured and examined) landing at

Figure 6. Picture of Centris trigonoides in a sand pit collecting sand for nest con

Florida Entomologist 79(2)

the edge of a stream and dipping their hind legs into the water and flying off. This may
be an alternative to the use of oil to transport sand or wood chips and may be more
common when oil is unavailable (Table 4).
The entrance to completed nests of C. bicornuta and C. vittata are covered with an
oily material that appears similar in consistency and color to whipped oil or "mayon-
naise" (Fig. 6). A similar material has been observed on rare occasions at the entrance
of completed nests of C. nitida. In C. vittata after one, two, or three cells are provi-
sioned, a partial cell is constructed and the inner surface is coated with a similar oily
material. In C. bicornuta after a series of 2-6 cells are completed, the female con-
structs a plug of material of similar consistency as a cell. This plug is generally over
6 mm thick. There is usually some space between the last completed cell and the plug.
The plug is usually recessed 4-6 mm and this space is filled with an oily material (Fig.
6). Observations of 16 C. bicornuta females developing this oily plug suggests the fe-
male brings in a liquid on her scopae which she removes at the entrance. She then
turns around and appears to add a regurgitate. She then again turns around and
sticks the tip of her abdomen into the material and rapidly twists her abdomen in a
whipping motion appearing to add air to the mixture and to whip the material into a
"mayonnaise" like material. Although a number of bee components are likely added,
the TLC pattern has no resemblance to B. crassifolia oil. The source of the initial liq-
uid (Oil II) is unknown (Table 6). This material hardens over a period of several days
to resemble a crumbly cheese which persists for a year.
The cell wall of most Centris nests is composed of either soil, sand, or wood chips
embedded in a hard waxy-resinous-like material of unknown origin. The suggestion
has been made (Vogel 1974, Neff & Simpson 1981, Simpson 1989, Buchmann 1987)
that the cell is composed of oils collected from such plants as B. crassifolia to which se-
cretions are added to cause the hardening and the material is mixed with the soil or
wood chips and hardens. However, we have been unable to support this suggestion, al-
though the source and nature of the material remains unknown. Our evidence is
based on observations of the nesting behavior of C. flavofaciata (SBV & GWF, unpub-
lished data) which only begins to construct a cell when returning with a brownish liq-
uid on her scopae.
Females of C. flavofaciata construct a tunnel and then construct a single cell at the
end of the tunnel. This cell is provisioned, capped, and then the tunnel is then filled
with sand (Vinson et al. 1987); the process requires about one day. Females initiating
a nest in the evening complete the tunnel, prepare the cell chamber, and then wait un-
til dawn to begin to construct a cell. By digging up nests before and after the first
morning trips were made by the bee (SBV & GWF, unpublished data), we determined
that no cells were initiated before the first trip; however, nests dug up during the bees'
second trip had the base of the cell formed. The partially formed cell was soft with the
sand being oily and the binding material soluble in ethanol, but not hexane. After the
bee made 3-4 trips, the cells were almost completely formed although they remained
soft. Bees collected during these trips to the nest had their scopula filled with an oily
material that hardened with time (4-6 h) and was soluble before hardening in ethanol,
but not hexane.
We tried to determine if females added something by examining the crop, Dufour's
gland, labial glands; we also examined females for other possible glands. Dissections
of bees returning upon the first trip had a full crop that appeared to contain nectar (it
did not harden) that was partly soluble in ethanol (becoming cloudy). The Dufour's
gland of C. flavofasciata is similar to that of C. flavifrons which is reported by Cane &
Brooks (1983) to be a long, thin tubular gland. The Dufour's gland showed no differ-
ences in size between the first and fifth trip. Further, Cane & Brooks (1983) reported
that the Dufour's glands of several Centris contain a series of hydrocarbons. Hydro-

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.

carbons are not very reactive and are soluble in hexane. The small mandibular and la-
bial glands also appeared identical in bees collected on the first and fifth trip back to
the nest. No other obvious glands were found. The fecal material in the hind gut was
not completely soluble in water, ethanol or hexane and when mixed with the oil of B.
crassifolia did not result in hardening. These observations, along with the natural
hardening of the material on the females scopula, suggest the females collect some-
thing (Resin I) that hardens without female additives (Table 4). Further, the sugges-
tion that B. crassifolia oil is the base material used in cell wall production (Simpson
1989, Neff & Simpson 1981, Buchmann 1987) is not supported unless the B. crassifo-
lia oil from some trees is substantially different. Although substantial differences in
B. crassifolia oil from some B. crassifolia trees was reported (Vinson et al., 1996), the
oil containing the large amount of an unknown also is soluble in hexane and does not
harden over time, suggesting this oil is also not responsible. Thus, the source of the
cell soil or wood chip binding material remains unknown.


Whether the cell wall material is an oil or resin awaits further analysis. Resins are
used by a species ofAnthodioct and a species of Chalicodoma that also nest in our
wooden nest blocks.
The source of the various resins and oils is difficult to determine. The TLC separa-
tions of the ethanol soluble cell wall material of C. flavofaciata indicates a complex
mixture; and while the cell wall material appears physically similar among the
ground nesters, the cell walls of tree nesters have some slight differences which may
or may not be due to the presence of a small amount of B. crassifolia oil. More impor-
tant than the chemical composition, is the sources of these materials. The population
of these oil-collecting bees appears to be in decline (Vinson et al. 1993). The reason for
the decline is not clear, but a reduction in any one of the resources required by these
bees could be responsible. However, there are several pollen and nectar sources (Table
4), and these are probably not resource limits. The situation with the Oil I used in nest
provisioning is less clear. There are several species of Malpighaceae in the area (Vin-
son et al. 1993) that may provide usable oil for some of the tree cavity nesting species
as indicated in Table 5, and the oil may not be the limiting factor. However, the resin
used for constructing the cell could be limiting if all the Centris species in the dry for-
est depended on a particular plant. The problem is to determine the source of the cell
wall material. There are many resins and gums released by a variety of trees attacked
by various insects along with many other resin and gum compounds associated with
plants. Although a number of these materials have been collected from nests and the
chemical composition of some of these are partially known, knowing the chemistry
may not be very helpful since the chemistry of only a very small percent of the possible
resins, gums, and other plant exudates are known. Thus, knowing the chemistry of
the cell material or defensive materials may be of little help in identifying the impor-
tant resources, particularly if combinations are used. It is only through a complete
knowledge of the resource needs of these bees that efforts can be made to insure their
survival in the remaining Pacific dry forest which is also under threat from a variety
of forces (Vinson & Frankie 1993).


We wish to thank Dr. Waldy Klassen for the opportunity to review our solitary bee
research in Costa Rica. We also acknowledge the research support of the National
Geographic Society (4941-92 and 5382-94), the Friends of Lomas Barbudhal, and the

Florida Entomologist 79(2)

Florida Entomological Society for allowing us to present this material. This manu-
script reports research conducted by the Texas Agricultural Experiment Station, The
Texas A&M University System.


ALBUQUERQUE, P. M. C., AND M. M. C. REGO. 1989. Fenologia das abelhas visitantes
de murici (Byrsonima crassifolia, Malpighiaceae). Bol. Mus. Para. Emflio.
Goeldi. sfr. Zool. 5: 163-178.
ALCOCK, J., C. E. JONES, AND S. L. BUCHMANN. 1976a. Location before emergence of
the female bee, Centris pallida by its male (Hymenoptera: Anthophoridae). J.
Zool. 179: 189-199.
ALCOCK, J., C. E. JONES, AND S. L. BUCHMANN. 1976b. The nesting behavior of three
species of Centris bees (Hymenoptera: Anthophoridae). J. Kansas Entomol. Soc.
49: 469-474.
ALCOCK, J., C. E. JONES, AND S. L. BUCHMANN. 1977. Male mating strategies in the
bee Centris pallida Fox (Hymenoptera: Anthophoridae). American Nat. 111:
ALCOCK, J. 1979. The evolution of intraspecific diversity in male reproductive strate-
gies in some bees and wasps, pp. 381-402 in M. S. Blum and N. A. Blum [eds.].
"Sexual Selection and Reproductive Competition in Insects". Academic Press,
NY. 463 pp.
ALCOCK, J., AND A. P. SMITH. 1987. Hilltopping, leks and female choice in the carpen-
ter bee Xylocopa (Neoxylocopa) varipuncta. J. Zool. Lond. 211: 1-10.
ANDERSON, W. R. 1978. Byrsonimodea: A new subfamily of the Malpighiaceae. Lean-
dra 7: 5-18.
ANDERSON, W. R. 1979. Floral conservatism in Neotropical Malpighiaceae. Biotropica
ANDERSON, W. R. 1983. Brysonima crassifolia (Nance, Nancile, Shoemaker's tree), pp.
202-204 in D. H. Janzen, [ed.]. "Costa Rican Natural History". Univ. Chicago
Press, Chicago, IL.
BAKER, H. C., AND I. BAKER 1975. Studies of nectar constitution and pollinator-plant
coevolution, pp. 100-140 in L. E. Gilbert and P. H. Raven [eds.]. "Animal and
Plant Coevolution". Univ. Texas Press, Austin, TX.
BAKER, H. C., AND P. D. HURD. 1968. Intrafloral ecology. Ann. Rev. Entomol. 13:385-414.
BATRA, S. W. T., AND J. C. SCHUSTER 1977. Nests of Centris, Melissodes, and Colletes
in Guatemala (Hymenoptera: Apoidea). Biotropica 9: 135-138.
BENNETT, F. D. 1964. Notes on the nesting sets of Centris derasa (Hymenoptera:
Apoidea) Pan-Pacific Entomol. 40: 125-128.
BRAGA, R. 1976. Murici, in "Plantes do nordeste, especialmenie do Ceura". 3rd Ed.
Fortaleza, p. 61.
BUCHMANN, S. L. 1987. The ecology of oil flowers and their bees. Annu. Rev. Ecol. Sys-
temat. 18: 343-369.
CAMARGO, J. M. DE, AND M. MAZUCATO. 1984. Inventario da fauna e flora apicola de
Ribeirao Preto, SP, Brazil. Dusenia 14: 55-87.
CANE, J. H., AND R. W. BROOKS. 1983. Dufour's gland lipid chemistry of three species
of Centris bees (Hymenoptera: Apoidea, Anthophoridae). R. Comp. Bio. B76 (4):
CHEMSAK, J. A. 1985. Observations on adult behavior of Centris flavofasciata Friese
(Hymenoptera: Anthophoridae). Pan-Pacific Entomol. 61: 265.
COVILLE, R. E., G. W. FRANKIE, AND S. B. VINSON. 1983. Nests of Centris segregata
(Hymenoptera: Anthophoridae) with a review of the nesting habits of the ge-
nus. J. Kansas Entomol. Soc. 56: 109-122.
1986. Nesting and male behavior of Centris heithausi in Costa Rica (Hy-
menoptera: Anthophoridae) with chemical analysis of the hind leg glands of
males. J. Kansas Entomol. Soc. 59: 325-336.

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.

EICKWORT, G. C., AND H. S. GINSBERG. 1980. Foraging and mating behavior in
Apoidea. Annu. Rev. Entomol. 25: 421-446.
EMLEN, S. T., AND L. W. ORING. 1977. Ecology, sexual selection, and the evolution of
mating systems. Science 197: 215-223.
FITCH, H. S., AND J. HACKFORTH-JONES. 1983. Ctensaura similis (Carrolo, Iguana,
Negra, Ctensaus), pp. 394-396 in D. H. Jansen, [ed.]. "Costa Rican Natural His-
tory". Chicago Press, Chicago, IL.
FRANKIE, G. W. 1976. Pollination of widely dispersed trees by animals in Central
America, with emphasis of bee pollination systems, pp. 151-159 in J. Burley
and B. T. Styles [eds.]. "Variation, Breeding, and Conservation of Tropical For-
est Trees". Academic Press, London.
FRANKIE, G. W., P. A. OPLER, AND K. S. BAWA. 1976. Foraging behavior of solitary
bees. Implications for outcrossing of a neotropical forest tree species. J. Ecol.
64: 1049-1057.
FRANKIE, G. W., S. B. VINSON, AND R. E. COVILLE. 1980. Male behavior of Centris
adani in the Costa Rican dry forest, with an emphasis in territorially (Hy-
menoptera: Anthophoridae). J. Kansas Entomol. Soc. 53: 837-857.
FRANKIE, G. W., S. B. VINSON, AND P. A. OPLER 1983. Centris aethyctera and C. fus-
cata in the lowland dry forest of Costa Rica (Hymenoptera: Anthophoridae), pp.
707-708 in D. H. Jansen [ed.]. "Costa Rica Natural History". Univ. Chicago
Press, Chicago, IL.
and habitual preferences of Centris bees in Costa Rican dry forest. Biotropica
20: 301-310.
FRANKIE, G. W., S. B. VINSON, AND H. WILLIAMS. 1989. Ecological and evolutionary
sorting of 12 sympatric species of Centris bees in Costa Rican dry forest, pp.
535-549 in J. H. Bock and Y. B. Linhart [eds.]. "The Evolutionary Ecology of
Plants". Westview Press, Boulder, CO.
K. FRANKIE. 1990. Plant phenology, pollination ecology, pollinator behavior, and
conservation of pollinators in neotropical dry forest, pp. 37-47 in K. S. Bawa
and M. Hadley [eds.]. "Reproductive Ecology of Tropical Forest Plants". Par-
thenon Pub., Park Ridge, NJ.
FRANKIE, G. W., L. E. NEWSTROM, AND S. B. VINSON. 1993. Nesting-habitat preferences
of selected Centris bee species in Costa Rican dry forest. Biotropica 25: 322-333.
GENTRY, A. H. 1993. A field guide to the families and genera of woody plants of North-
west South America (Colombia, Ecuador, Peru) with supplementary notes on
Hebaceous taxa. Conservation Instit. Washington, DC.
JAYASINGH, D. B., AND B. E. FREEMAN. 1980. Trap-nesting solitary aculeates (Insecta:
Hymenoptera) in St. Catherine, Jamaica. Caribbean J. Sci. 15: 69-78.
KEVAN, P. G., AND H. G. BAKER 1983. Insects as flower visitors and pollinators. Annu.
Rev. Entomol. 28: 407-453.
KIMSEY, L. S. 1978. Nesting and male behavior in Dynatus nigriceps spinolae (Lepe-
letier) (Hymenoptera: Sphecidae). Pan-Pacific Entomol. 54: 65-68.
MARSHALL, L. D., AND J. ALCOCK. 1981. The evolution of the mating system of the Car-
penter bee Xylocopa varipuncta (Hymenoptera: Anthophoridae). J. Zool., Lon-
don 193: 315-324.
MICHENER, C. D., AND R. B. LANGE. 1958. Observations on the ethology of neotropical
anthophorine bees (Hymenoptera: Apoidea). Univ. Kansas Sci. Bull. 39: 69-96.
MINCKLEY, R. L., S. L. BUCHMANN, AND W. T. WCISLO. 1991. Bioassay evidence for a
sex attractant pheromone in the large carpenter bee Xylocopa varipuncta (Hy-
menoptera: Anthophoridae). J. Zool., London 224: 285-291.
NEFF, J. L., AND B. B. SIMPSON. 1981. Oil-collecting structures in the Anthophoridae
(Hymenoptera): Morphology, function and use in systematics. J. Kansas Ento-
mol. Soc. 54: 95-123.
PARSONS, J. J. 1976. Forest into pasture: Development or destruction? Review of Bi-
ology in the Tropics 24 (Suppl. 1): 121-138.

Florida Entomologist 79(2)

PICKEL, B. 1928. Contribuicao para a biologia de Centris sponsa e Acanthopus exce-
llens (Hymen.). Bol. Biol. (Sao Paulo) 14: 135-143.
POHL, R. W. 1983. Hyparrhenia rufa (Jaragua), pp. 77-79 in D. H. Janzen [ed.]. "Costa
Rican Natural History". Univ. Chicago Press, Chicago, IL.
RAW, A. 1975. Territorality and scent marking by Centris males (Hymenoptera: An-
thophoridae) in Jamaica. Behaviour 54: 311-321.
RAW, A. 1984. The nesting biology of nine species of Jamaican bees (Hymenoptera).
Revta. Bras. Entomol. 28: 497-506.
REGO, M. M. C., AND P. M. C. DE ALBUQUERQUE. 1989. Compartamento des abelhas
visitantes de Murici, Byrsonima crassifolia (L.) Kunth, Malpigheaceae. Bol.
Mus. Para. Emflio. Goeldi. sfr. Zool. 5(2): 179-193.
ROZEN, J. G., AND S. L. BUCHMANN. 1990. Nesting biology and immature stages of
bees Centris caesalpiniae, C. pallida, and the cleptoparasite Ericrocis lata (Hy-
menoptera: Apoidea, Anthophoridae). Am. Mus. Novitates 2985: 30 pp.
SEIGLER, D. S., B. B. SIMPSON, C. MARTIN, AND J. L. NEFF. 1978. Free 3-acetoxy fatty
acids in floral glands ofKrameria species. Phytochem. 17: 995-996.
SIMPSON, B. B. 1989. Pollination biology and taxonomy of Dinemandra and Dinema-
gonum (Malpighiaceae). Systematic Botany 14: 408-426.
SIMPSON, B. B., AND J. L. NEFF. 1981. Floral rewards, alternatives to pollen, and nec-
tar. Ann. Missouri Bot. Gard. 68: 301-322.
SIMPSON, B. B., J. L. NEFF, AND D. L. SEIGLER 1977. Krameria, free fatty acids and
oil collecting bees. Nature 267: 150-151.
SIMPSON, B. B., D. S. SEIGLER, AND J. L. NEFF. 1979. Lipids from the floral glands of
Krameria. Biochem. Syst. Ecol. 7: 193-194.
SIMPSON, B. B., J. L. NEFF, AND G. DIERINGER 1990. The production of floral oils by
Monttea (Scrophulariaceae) and the function of tarsal pads in Centris bees.
Plant Systematics and Evolution 173: 209-222.
SNELLING, R. R. 1974. Notes on the distribution and taxonomy of some North Ameri-
can Centris (Hymenoptera: Anthophoridae). Contrib. Science, Los Angeles, Cal-
ifornia Museum Nat. Hist. 259: 1-41.
SNELLING, R. R. 1984. Studies on the taxonomy and distribution of American Centri-
dine bees (Hymenoptera: Anthophoridae). Contrib. Science, Los Angeles, Cali-
fornia Museum Nat. Hist. 347: 1-69.
TORO, H., E. CHIAPPA, L. RUZ, AND V. CABEZAS. 1991. Compartamiento reproductive
de Centris mixta tamarugalis (Hymenoptera: Anthophoridae). I Parte. Acta.
Ent. Chilena 16: 97-112.
VESEY-FITZGERALD, D. 1939. Observations on bees (Hymenoptera: Apoidea) in Trin-
idad, B.W.I. Proc. Royal Entomol. Soc. London (A) 14: 107-110.
VINSON, S. B., AND G. W. FRANKIE. 1977. Nests of Centris aethyctera (Hymenoptera:
Apoidea: Anthophoridae) in the dry forest of Cost Rica. J. Kansas Entomol. Soc.
E. COVILLE. 1982. Mandibular glands of male Centris adani (Hymenoptera:
Anthophoridae): their morphology, chemical constituents, and function in scent
marking and territorial behavior. J. Chem. Ecol. 8: 319-327.
VINSON, S. B., H. J. WILLIAMS, G. W. FRANKIE, AND R. E. COVILLE. 1984. Comparative
morphology and chemical contents of male mandibular glands of several Cen-
tris species in Costa Rica. Comp. Biochem. Physiol. 77a: 685-688.
VINSON, S. B., G. W. FRANKIE, AND R. E. COVILLE. 1987. Nesting habits of Centris fla-
vofasciata (Hymenoptera: Apoidea: Anthrophoridae) in Costa Rica. J. Kansas
Entomol. Soc. 60: 249-263.
VINSON, S. B., AND G. W. FRANKIE. 1988. A comparative study of the ground nests of
Centris flavifrons and Centris aethiocesta (Hymenoptera: Anthophoridae). En-
tomol. Expt. Appl. 49: 181-187.
contents of male mandibular glands of three Centris species (Hymenoptera:
Anthophoridae) from Costs Rica. Comp. Biochem. Physiol. 93B: 73-75.

June, 1996

Behavioral Ecology Symposium '95: Vinson et al.

VINSON, S. B., AND G. W. FRANKIE. 1991. Nest variability in Centris aethyctera (Hy-
menoptera: Anthophoridae) in response to nesting site conditions. J. Kansas
Entomol. Soc. 64: 156-162.
VINSON, S. B., AND G. W. FRANKIE. 1991. The oil barron bees of Lomas Barbudal. Bee-
line 5 (1-3): 16-17. Pub. by Friends of Lomas Barbudal, Berkeley, CA 94707.
VINSON, S. B., G. W. FRANKIE, AND J. BARTHELL. 1993. Threats to the diversity of sol-
itary bees in a neotropical dry forest in central America, pp. 53-81 in J. LaSalle
and I. D. Gauld [eds.]. "Hymenoptera and Biodiversity". C.A.B. International,
Wallingford, UK.
VINSON, S. B., H. J. WILLIAMS, G. W. FRANKIE, AND G. SHRUM. 1996. Floral lipid
chemistry of Byrsonima crassifolia (Malpigheacea) and a use of floral lipids by
Centris bees (Hymenoptera: Apidae). Biotropica (in press).
VOGEL, S. 1969. Flowers offering oil instead of nectar. XI Bot. Congr. Abstr. Seattle,
Washington. pp. 229.
VOGEL, S. 1974. Olproduzierende Blumen, die durch olsammelnde Bienen. Tropishe
und subtropishce Pflanzenwelt, Nr. 7. F. Steiner, Wiesbaden 267 pp.
VOGEL, S. 1976a. Lysimachia: Olblumen der Holarktis. Naturwissenschaften 63: 44-
VOGEL, S. 1976b. Oil-collecting bees of the Old World and their flowers. Abst. 15th Int.
Entomol. Congr., Washington, DC. Aug. 1976.
VOGEL, S. 1981. Abdominal oil-mopping-a new type of foraging in bees. Naturwis-
senschaften 67: 5627.
VOGEL, S. 1986. Olblumen and olsammelnde Bienen: Zweite Folge Lysemachia und
Macropis, 168 p. in "Akademie der Wissenschaften und der Literatur Mainz,
1986." Stuttgart: Steiner.
VOGEL, S. 1988. Die Olblumen-Symbiosen: Parrallelismus und andere Aspekte ihrer
Entwicklung in Raum and Zeit. Z. Zool. Syst. u Evolut. Forsch. 26: 341-362.
VOGEL, S. 1990. The role of scent glands in pollination: On the structure and function
of osmophores. (Translation: Smithsonian Instit. Libraries; Amerind Pub., New
Delhi). Rotterdam, Balkema, 202 pp.
Morphology, chemical contents and possible function of the tibial gland of
males of the Costa Rican solitary bees Centris nitida and Centris trigonoides
subtarsata. J. Kansas Entomol. Soc. 57: 50-54.
WILSON, D. E. 1983. Ipomea pes-capae (Pudre Oreja, Beach Morning Glory), pp. 261-
262 in: D. H. Janzen [ed.]. "Costa Rican Natural History". Univ. Chicago Press,
Chicago, IL.

Florida Entomologist 79(2)


'Escuela de Biologia, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica

present address: Instituto Nacional de Recursos Naturales Renovables
(INRENARE), Areas Protegidas, Apartado 2016,
Paraiso, Corregimiento Ancon, Panama

"and Smithsonian Tropical Research Institute


Video recordings of Ceratitis capitata courtship behavior revealed several hitherto
unreported details. The droplet of pheromone at the tip of the male's abdomen during
wing vibration was partially or completely retracted during wing buzzing and head
rocking. Wing vibration gave way to wing buzzing and head rocking during the last 1-
2 sec before the male attempted to mount the female; the male ceased rocking his
head during the last 0.3 sec before a mounting attempt. Immediately after landing on
the female, the male performed up to three additional types of apparent courtship be-
fore achieving intromission. The circumstances in which males attempted to mount
females differed from those in which males abandoned courtship: the male was closer
to the female and the two flies were oriented more directly toward each other. Neither
reductions in distance nor more precise orientations appeared to be the immediate
cues releasing mounting attempts, however.

Key Words: Medflies, Ceratitis, courtship, behavior.


Video grabaciones del comportamiento de cortejo de Ceratitis capitata revelaron
various detalles antes desconocidos. La gotita de feromona que aparece en la punta del
abdomen del macho durante la fase de vibraci6n de las alas fue resorbida cuando dib
inicio la fase de zumbido de las alas y balance de la cabeza. La vibraci6n de las alas
fue reemplazada por el zumbido de las alas y el balance de la cabeza durante los ul-
timos 1-2 segundos antes de que el macho intentara montar a la hembra; la oscilacion
de la cabeza termin6 0.3 segundos antes del intent de montar a la hembra. Una vez
sobre la hembra, el macho ejecut6 hasta tres tipos de cortejo aparente antes de vol-
tearse e intentar copularla. Las circunstancias en las cuales los machos intentaron
montar a las hembras difirieron de aquellas en las cuales los machos abandonaron el
cortejo: el macho estaba mas cerca de la hembra y las dos moscas estuvieron mas di-
rectamente orientadas una hacia la otra. Ni la reducci6n en la distancia entire las mos-
cas ni una orientaci6n mas precisa parecieron ser las sefiales inmediatas que liberan
la conduct de la monta.

The basic nature of courtship and copulation behavior in the medfly, Ceratitis cap-
itata (Wiedemann) is well known. Males mate in leks, in which males behave territo-
rially and each occupies a separate leaf, and at oviposition sites where territorial
behavior is reduced (Prokopy & Hendrichs 1979, Hendrichs & Hendrichs 1990, Whit-

June, 1996

Behavioral Ecology Symposium '95: Bricefio et al.

tier et al. 1992, Shelly et al. 1993). In leks, and less often when alone, males release
a long distance attractant pheromone, exposing a droplet of liquid by everting a bal-
loon-like structure formed by a membranous portion of the rectal epithelium (Arita &
Kaneshiro 1986). Male courtship begins when a female approaches a male, and in-
cludes the following behavior patterns (Feron 1962): (1) continuous wing vibration, of-
ten performed while the male is facing the female and the tip of his abdomen is bent
ventrally with the pheromone droplet present on the everted rectal membrane ("vi-
brate wings", or stage II of Feron 1962) (a plume of pheromone is probably thereby
wafted toward the female); (2) wing buzzing, during which the wings are rhythmically
moved forward and back ("buzz !ii,- and (3) rapid rotations of the head ("head
rock"). Wing buzzing and head rocking often occur simultaneously (stage III of Feron
1962; see also Rolli 1976 on the resulting songs).
After courting the female, the male leaps onto her and at least sometimes buzzes
his wings. These vibrations may serve as further courtship (Zapien et al. 1983), or to
maintain the male's balance while he positions himself to intromit. A recently discov-
ered courtship behavior sometimes occurs after the male has mounted the female but
before he has achieved intromission. The male nips the female at the tip of her abdo-
men with his genitalic surstyli. This is often followed by her extending her aculeus,
which allows him to clamp it with his surstyli so he can intromit (Eberhard & Pereira
Presumably most of these male behavior patterns serve to elicit crucial female re-
sponses: to approach, or to stop and allow the male to approach close enough that he
can mount; to align her body properly for a mounting attempt; and to subsequently al-
low him to intromit. More often than not, however, courtship fails to result in female
acceptance. For instance, the female walked away from a courting male in more than
90% of courtships in the lab (Whittier et al. 1994); and the female dislodged the male
after he had mounted in more than 90% of mounting attempts (Kaneshiro 1991) (see
also Feron 1962). Thus certain male behavior patterns, and appropriate transitions
from one type of courtship behavior to another, may have important effects on a male's
chances of copulating.
The role of visual stimuli from the female in eliciting wing vibration (stage I of
Feron 1962) and of olfactory stimuli from other males in eliciting pheromone release
have been demonstrated experimentally, though only with qualitative data (Feron
1962). Many other important aspects of the coordination of male behavior remain un-
studied, however. For instance, there are apparently no studies of the factors that may
affect male decisions such as whether to turn toward the female, to advance toward
the female, to shift from wing vibration to wing buzzing, to rock the head, or to at-
tempt to mount. Decisions by females (e.g. whether to leave, to allow the male to ap-
proach, to allow him to mount, to extend the aculeus and allow intromission) are also
Understanding these sorts of details may have important practical consequences.
The medfly is a notorious agricultural pest, and large sums of money are spent annu-
ally in raising and releasing sterile males to mate with wild females. A better under-
standing of the details of medfly courtship may have important implications in
attempts to maintain the competitive quality of mass-reared males in the face of pos-
sible differences in selective regimes in the field and in mass-rearing facilities (e.g.
Spates & Hightower 1967, Boller et al. 1981, Kaneshiro 1991), as well as substantial
variations in some behavioral aspects such as songs (Rolli 1976).
This paper attempts to lay the groundwork for such studies by describing the de-
tails of male courtship behavior. These descriptions permit preliminary conclusions
regarding the cues used by males to trigger attempts to mount females.

Florida Entomologist 79(2)


Three recordings of 0.5 h each were made of a strain of flies that had been mass-
reared for 14 years at the Organizaci6n Internacional Regional de Sanidad Agropec-
uaria (OIRSA) in San Jos6, Costa Rica using a National "Omnipro" video camera (30
images per s). All events were taped which occurred in an area of 588 cm2 on the cloth
wall of the cage through which females oviposited, and the tapes were then studied to
follow individual males during courtship sequences. The density of flies resting on
this cloth in 20 counts separated by 5 min, averaged 0.550.083 flies per cm2. Supple-
mentary recordings of both wild flies and of a strain of flies that had been mass-reared
for 4.5 years were used to clarify some behavioral details. Unless stated otherwise, all
data are from the 14 year strain.
Data taken from the recordings included the distance of the male from the female
(measured on screen of the monitor and later converted to cm), the angle of male ori-
entation toward the female (male orientation), and the angle of alignment between
male and female (female orientation) (Fig. 1). Male behavior (turn, walk forward or
backward, vibrate wings, buzz wings, rock head, leap onto female) was also noted. All
drawings are based on images traced from videotapes; body parts (legs, wings, etc.)
that were out of focus or otherwise unclear were omitted. All averages are followed by
one standard deviation. The flies were not marked, thus it was not certain that differ-
ent courting pairs involved different individuals. The cage contained over 10,000 flies,
however, so it is very likely that each male seen courting was a different individual.

The most common sequences of behavior (Fig. 2) were similar to those reported by
other authors (e.g. Feron 1962). The following details, however, were different.
(A) Resorption of the pheromone droplet. As has been described by Feron
(1962), males consistently bent their abdomens to hold the everted pher-
omone droplet ventrally while they vibrated their wings (Fig. 3). When
wing buzzing replaced wing vibration, however, the droplet was usually
resorbed (completely resorbed in 75% of 12 sequences, 25% partially re-
sorbed in the others).
(B) A burst of rapid forward-backward rocking of the male's body after he
landed on the female (Fig. 4). Small details of the flies' movements dur-
ing body rocking, such as deflections of the female's wings, made it clear
that the male rather than the female produced these movements. In a
sample of 15 mountings from the 4.5 year strain, the time elapsed be-
tween landing and the beginning of body rocking averaged 0.430.31 s,
the number of rocking movements averaged 42.1 per mount (range 1-
7), and the duration of body rocking averaged 1.640.23 s. Rocking
movements of this sort are not a normal part of landing behavior and
were never observed in flies landing on the wall of the rearing cage.
(C) Cross-legged raising of the female oviscape and
(D) Rubbing on the ventral surface of the female abdomen with the male's
hind tarsi and tibiae. In at least those pairs in which the male had diffi-
culty intromitting, he repeatedly raised the tip of the female's abdomen
and rubbed its ventral surface with his hind legs (Fig. 5). In one wild pair
in which more than 5 min of this behavior was taped before the male fi-
nally achieved intromission, some movements of the male's legs were
produced by pushing movements of the female's hind legs; nevertheless
the male clearly rubbed her on other occasions when her legs were im-

June, 1996

Behavioral Ecology Symposium '95: Bricefio et al.

I /





/ I

Figure 1. Distances and angles measured from video recordings.

Slow motion analyses of video recordings also revealed new details of some previ-
ously described behavior patterns. Wing vibration involved both rapid vibration (too
fast to be resolved in 30 frames per s images), and slower, small deflections from side
to side about 5-6 times per s (Fig. 6). Wing buzzing included two different components:
a rapid continuous vibration of the wings (too fast to be resolved in 30 frames per s im-

Florida Entomologist 79(2)

Figure 2. Frequencies of transitions between different male behavior patterns in
17 courtship sequences in which the male attempted to mount the female. Widths of
arrows are proportional to frequencies. A = reorient; B = vibrate wings; C = buzz
wings; D = rock head; E = mount female; F = walk toward female; G = immobile; H =
move away; I = rub legs together.

ages); and a rapid, repeated forward-backward movement about 3-4 times per s (av-
erage 1/0.260.11 s, N=34) (Fig. 7). The number of forward-backward movements per
buzz averaged 10.159.95 (N=34). Head rocking also involved more complex move-
ments than previously appreciated. As described by Feron (1962), the largest dis-
placements of the male's head were rotatory (A in Fig. 8); but lateral as well as dorso-
ventral movements also occurred (B and C in Fig. 5). Rotation in one direction (e.g. to
the left) was nearly always followed by rotation in the other direction. Rotations were
too rapid to allow precise measurements of angles, but it appeared that the first rota-
tions in a series tended to be larger than later ones. The frequency of rotations aver-
aged 24.622.6 per s in 21 courtships.
Possible stimuli triggering male mounting attempts were checked by analyzing
details of the courtship behavior of 17 pairs during the 6 s period immediately preced-
ing the male's leap onto a female that was more or less facing him (3 other leaps from
behind the female were excluded). The data in the three images during each lapse of
0.1 s were averaged during the second immediately proceeding the leap; similar aver-
ages were calculated for each of the 0.3 s during the proceeding 5 s. As shown in Fig.
9A, wing vibration was the earliest courtship behavior in all interactions, occurred at

June, 1996

Behavioral Ecology Symposium '95: Bricefio et al.



Figure 3. A droplet of presumed pheromone (black) was exposed dorsally by calling
males, and the droplet was directed ventrally during wing vibration. The droplet was
usually resorbed when the male began wing buzzing. The drawing at the bottom rep-
resents an image 0.4 sec. after the drawing in the middle.

Florida Entomologist 79(2)

A .-
,, -.: /

leap; it became especially common about s before the leap, and then disappeared

abruptly 0.3 s before the leap.
^w\ ... "


Figure 4. After landing on a female (left; the upper image of the male followed the
other by 0.06 s), the male rocked his body briskly anteriorly and posteriorly (right; the
dotted lines in the figu, represent the male's position 0.03 se after the solid lines)
(drawings of wild male and female; male wings omitted for clarity).

intermediate frequences during the last seconds before a leap, and was replaced com
pletely by buzzing during the last second. Head rocking was absent >5.8 s before the
leap; it became especially common about 1 s before the leap, and then disappeared
abruptly 0.3 s before the leap.
The male remained oriented toward the female and showed no clear trends during
the 6 s preceding a leap, other than a possible decrease in the variance during the fi
nal s (Fig. 9C). The angle of the female's alignment toward the male was also rela
tively constant and variance declined head-to-head during last s before the jump (Fig.
9B). The average distance between the two flies was also nearly constant during the
entire courtship period, decreasing slightly and becoming less variable just before the
leap (Fig. 9D).
Using the same 17 interactions, male and female behavior were analyzed as a
function of the distance between the two flies during the 6 s proceeding a leap. The
likelihood that a leap would occur was greater when the flies were closer, although in
a few cases males leapt from 0.75-1.0 cm away (Fig. 10A). The relationships between
distance and the different male courtship behavior patterns were similar to those in
Fig. 9. Wing vibration was more common farther from the female, and wing buzzing
and head rocking more common closer to her (Fig. 10B). Similarly, male orientation
was low and nearly constant at different distances, while female alignment was closer
to 0 at shorter distances. Males were more likely not to be walking than females at
all distances, and both males and females were more likely to be immobile when they
were closer together (Fig. 10C).
These analyses suggest that neither male or female orientation nor the distance
between the flies is the immediate cue triggering mounting attempts by males. These
factors are, nevertheless, probably important in predisposing males to attempt to
mount, as shown by comparisons between the relative positions between males that

June, 1996

Behavioral Ecology Symposium '95: Bricefio et al.


( .1-

Figure 5. Behavior of a mounted male attempting to achieve intromission. A. Male
moves his abdomen dorsally after pressing his genitalia against those of the female,
but having failed to clasp her aculeus with his surstyli. The tip of the male's abdomen
was touching that of the female 0.3 s before the position indicated by solid lines; 0.3 s
later it was in the position indicated by dotted lines. B. Mounted male raises the tip
of the female's abdomen with his crossed hind legs as he bends his abdomen ventrally
and anteriorly to bring his genitalia into contact with her. C. The male rubs the ven-
tral surface of the female's abdomen with one hind tarsus, which first moves anteri-
orly (dotted lines follow solid lines by 0.1 s), and then posteriorly (0.2 s later).

they attempted to mount, and males that stopped courting without being interrupted.
Males which abandoned courtship attempts were farther from the female (0.790.38
as compared to 0.450.23 cm), were less oriented toward the female (42.330.0 vs.
4.36.2'), and the female was less aligned toward the male (88.952.7 vs. 14.415.2)
(N=35 for abandoned attempts) (all p<0.01 with Mann-Whitney U Test).
When the overall averages for the behavior of a random sample of males seen
courting were compared with averages for courtship sequences in which the male at-
tempted to mount the female, the durations of wing vibration but not wing buzzing or
reorientation movements by the male head rocking were shorter for courtships pre-
ceding mounting attempts (Table 1). Thus, if a male was going to mount a female, he
tended to do so relatively rapidly. The possible importance of rapid male mounting
was illustrated by the fact that most courtships in mass-rearing cages were inter-
rupted when another fly walked or landed nearby. Of 30 randomly chosen cases in
which a male began courting a female, 63% were interrupted by another fly. Of 29
cases in which a male was interrupted, in only 17% did he resume courtship of the
same female.

Florida Entomologist 79(2)

Figure 6. Wing vibration involved both rapid vibrations (area covered by displace-
ments indicated by solid lines) and slower, somewhat irregular side to side displace-
ments (dotted lines). Dotted lines represent wing positions 0.12 s after solid lines.

1 second

Figure 7. Wing buzzing involved rapid wing vibrations during which there were
quick forward and backward movements. The graph shows approximate changes in
wing positions over time (angle with longitudinal body axis; measurements were
somewhat imprecise due to the blurred outlines of the wings, especially when the
wings were moved anteriorly). Forward and backward movements were relatively
rapid, and the wings were kept in the anterior position for shorter times than in the
posterior position.

June, 1996

Behavioral Ecology Symposium '95: Bricefio et al.



Figure 8. Ranges of movement of the male's head during head rocking. The stron-
gest movements were rotatory (A), but dorso-ventral (B) and lateral (C) deflections
also occurred. B and C were drawn directly from video images; the displacements in
A were extrapolated from changes in the positions of the male's capitate setae seen in
lateral views.


Our observations clarify some behavioral details on which there have been previ-
ous contradictory reports. Zapien et al. (1983) indicated that when the male lept to-
ward a female, he "jumped over the head of the female, turned and mounted her." The
videos showed, however, that not only did the male consistently land on the female
more or less facing the posterior end of her body and then turn 180 as described by
Feron (1962), but he also rocked his body in a distinctive manner before making the
Arita & Kaneshiro (1986) reported that the rectal epithelium is extended during
the entire courtship sequence prior to mounting. At least in the strains we observed,
the epithelium was everted during wing vibration, but was retracted during wing
buzzing. Arita & Kaneshiro (1985) also mention that the male approaches the female
"until he is almost touching the female's antennae with his second pair of fronto-or-
bital bristles." The flies of this study also approached females closely before leaping,
but in at least some cases it was certain that they did not touch them with these bris-
tles (e.g. Fig. 4) (the female's arista was generally not visible in the images, but some
mounting attempts were launched from distances larger than the length of the
arista;) (V. Mendez, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica, for

140 Florida Entomologist 79(2) June, 1996


o o
-7 Z

Co C L

-- /Cp ) 'I U (-) u o
C" I c C

0 (D = I o
\E L E
E s b

o o o 0 o in o o
Co (b 4 C' N 6- 6
(sau6nap) Gauuv (V3) aBLa aoue siU !

I -a.- -- bfl

Ii ;" E E ^

< "'"- o i

*-- --- l --- --- l --- t -- --------.---- ID **
n E IC) a c


Co bf l
ir C

QL or Nrn b l

Behavioral Ecology Symposium '95: Bricerio et al.





0 V.


(seejBap) a|6uv

I a
4 a

E 05

- E


0 O




i +


- C)
o a>
C^ cc
(U C)
at a


Florida Entomologist 79(2)


Behavior Prior to
Random Leaping onto a
Sample N Female N

Reorient when female moves (sec) 0.820.83 212 0.840.76 50
Wing vibration 7.445.12 79 2.483.48 50
Wing buzz 3.333.31 34 3.633.17 50
Head rock (sec) 4.654.72 7 0.830.75 17
Distance prior to jump (cm) 0.420.80 29

further data on this point) (in contrast, both bristles and antennae touched in some
aggressive head-to-head interactions between males-RDB & WGE, unpublished
data). This suggests that the bristles may form part of the visual stimuli provided by
head rocking (Arita 1983 cited in Arita & Kaneshiro 1986), and a positive effect on
possible male mating success has been confirmed by experimental removal of the bris-
tles (V. Mendez).
The detailed analysis of male behavior just before mounting attempts also re-
vealed new aspects of courtship not previously described. As previously reported,
males buzzed their wings and rocked their heads prior to leaping onto the female, but
head rocking was intermittent, and ceased during the 0.3 s immediately preceding the
leap. Perhaps the male was better able to orient his leap by holding his head still just
before leaping. The decision to mount the female may thus occur several tenths of a
second before the leap itself.
The courtship song recorded by Webb et al. (1983) and signal II of Rolli (1976) prob-
ably correspond to wing buzzing. The frequency of pulses of sound (1 per 0.18 s Webb
et al. 1983) was similar to the frequency of forward wing displacements observed in
the videos (1 per 0.26 s). Feron (1962) mentions a slightly slower frequency of 1 per 0.5
s, and notes that the female's wings were apparently slightly raised by the air cur-
rents produced by the male's wing movements. Since the rectal epithelium is re-
tracted during wing buzzing, this behavior presumably represents an acoustic (or
mechanical?) rather than a chemical stimulus.
The male's distance to the female was smaller, he was oriented more directly to-
ward her, and she was oriented more directly toward him when a male attempted to
mount the female than when he abandoned courtship, suggesting that these variables
may be involved in triggering mounting attempts. None of these variables showed
clear changes in the few seconds preceding mounting attempts, however, so they are
probably not immediate triggers of mounting.
Using male behavior to tentatively deduce which possible stimuli are important,
it appears that there are three phases of male courtship in medflies: a preliminary
phase (initial attraction; wing vibration) in which odors play an important role; a sub-
sequent close-range phase (head rocking, wing buzzing) in which visual and vibratory
stimuli are important; and a final contact phase (body rocking, rubbing the female's
abdomen and lifting her oviscape, nipping with genitalic surstyli) in which mechani-
cal stimuli dominate. Further observations are needed to test these ideas and the pos-
sibility that (1) the behavior of mass-reared and wild flies is different and (2) the flies

June, 1996

Behavioral Ecology Symposium '95: Bricefio et al.

facultatively modify their behavior when conditions, such as crowding, are changed
(RDB & WGE, unpublished data).


We are grateful to Hernan Camacho for allowing us access to the OIRSA facilities,
M. J. West-Eberhard for criticizing a preliminary draft, and W. Klassen for inviting us
to participate in the Symposium. The Vicerrectorid de Investigaci6n of the Univer-
sidad de Costa Rica, the Smithsonian Tropical Research Institute, and the Interna-
tional Atomic Energy Agency provided financial support.


ARITA, L., AND K. KANESHIRO. 1985. The dynamics of the lek system and mating suc-
cess in males of the Mediterranean fruit fly, Ceratitis capitata (Wiedemann).
Proc. Hawaiian Entomol. Soc. 25: 39-48.
ARITA, L., AND K. KANESHIRO. 1986. Structure and function of the rectal epithelium
and anal glands during mating behavior in the Mediterranean fruit fly male.
Proc. Hawaiian Entomol. Soc. 26: 27-30.
ing, monitoring and improving the quality of mass-reared Mediterranean fruit
flies, Ceratitis capitata Wied. Z. angew. Entomol. 92: 67-83.
EBERHARD, W. G., AND F. PEREIRA. 1993. Functions of the male genitalic surstyli in
the Mediterranean fruit fly, Ceratitis capitata (Diptera: Tephritidae). J. Kansas
Entomol. Soc. 66: 427-433.
FERON, M. 1962. L'instinct de reproduction chez la mouche Mediterraneenne des
fruits Ceratitis capitata Wied. (Dipt. Trypetidae). Comportemont sexuel. -
Comportement de ponte. Rev. Pat. Veg. Entomol. Veg. 41: 1-129.
HENDRICHS, J., AND M. A. HENDRICHS. 1990. Mediterranean fruit fly (Diptera: Te-
phritidae) in nature: location and diel pattern of feeding and other activities on
fruiting and nonfruiting hosts and nonhosts. Ann. Entomol. Soc. America 83:
KANESHIRO, K. Y. 1991. Quality control of mass-reared strains of fruit flies. pp. 137-
145 in K. Kawasaki, 0. Iwahashi and K. Kaneshiro [eds.] Proceedings of the in-
ternational symposium on the biology and control of fruit flies. Ginowan, Oki-
nawa, Japan.
PROKOPY, R. J., AND J. HENDRICHS. 1979. Mating behavior ofCeratitis capitata in field
cages on host trees. Ann. Entomol. Soc. America 72: 642-648.
ROLLI, V. K. 1976. Die akustischen Sexualsignale von Ceratitis capitata Wied. und
Dacus oleae Gmel. Z. ang. Entomol. 81: 219-223.
SHELLY, T. E., T. S. WHITTIER, AND K. Y. KANESHIRO. 1993. Behavioral responses of
Mediterranean fruit flies (Diptera: Tephritidae) to trimedlure baits: can leks be
created artificially? Ann. Entomol. Soc. America 86: 341-351.
SPATES, G., AND B. HIGHTOWER 1967. Sexual aggressiveness of male screw-worm flies
affected by laboratory rearing. J. Econ. Entomol. 60: 752-755.
WHITTIER, T. S., K. Y KANESHIRO, AND L. D. PRESCOTT. 1992. Mating behavior of Med-
iterranean fruit flies (Diptera: Tephritidae) in a natural environment. Ann. En-
tomol. Soc. America 85: 214-218.
WHITTIER, T. S., F. Y. NAM, T. E. SHELLY, AND K. Y KANESHIRO. 1994. Male courtship
success and female discrimination in the Mediterranean fruit fly (Diptera: Te-
phritidae). J. Ins. Behav. 7: 159-170.
ZAPIEN, G., J. HENDRICHS, P. LIEDO, AND A. CISNEROS. 1983. Comparative mating be-
havior of wild and mass-reared sterile medfly Ceratitis capitata on a field-caged
host tree. II. Female mate choice. pp. 397-409 in R. Cavalloro [ed.] Fruit Flies
of Economic Importance. Balkema, Rotterdam, The Netherlands.

Florida Entomologist 79(2)


'United States Department of Agriculture,
Agricultural Research Service, Insect Attractants,
Behavior and Basic Biology Research Laboratory, Gainesville, FL 32604

'Department of Entomology, Tropical Research and Education Center,
University of Florida, Homestead, FL 33031


A synthetic food-based attractant, and a painted cylindrical dry trap that protects
the synthetic lures from the environment, were developed to replace liquid protein-
baited traps. This trapping system was tested for capture of the Mediterranean fruit
fly, Ceratitis capitata (Wiedemann). The dry trap is constructed of acetate film with a
painted band that provides a visual cue; it contains internally-placed toxicant panels
to kill flies that enter the trap. Field trials conducted in Guatemala suggested that a
solid-colored material could be substituted for the painted trap body. We also evalu-
ated a sticky insert made from commercially produced adhesive paper as a replace-
ment for the toxicant panels. Unlike paintable sticky adhesives, the sticky material on
the adhesive paper insert does not adhere to the skin of personnel who service the
traps. An open-bottom trap made of green opaque plastic with a sticky insert captured
more C. capitata than the closed-bottom painted trap with a toxicant panel. When
used in conjunction with sterile insect release technology, the open-bottom dry trap
baited with food-based synthetic attractant often caught wild C. capitata in numbers
equal to those caught by trimedlure-baited Jackson traps, but the dry trap caught
many fewer sterile C. capitata.

Key Words: Ceratitis capitata, trapping, food-based attractant, dry trap.


Para sustituir trampas cebadas con protein liquid fueron desarrollados un fa-
goatrayente y una trampa seca cilindrica pintada que protege los cebos sinteticos del
ambiente. Este sistema de trampa fue probado para la capture de la mosca medite-
rranea de las frutas, Ceratitis capitata (Wiedemann). La trampa seca esta construida
con una pelicula de acetato con una banda pintada que provee una seiial visual y con-
tiene panels internos t6xicos para matar a las moscas que entren. Las pruebas de
campo conducidas en Guatemala sugieren que un material solidamente coloreado po-
dria ser substituido por el cuerpo pintado de la trampa. Tambien evaluamos un dis-
positivo adhesive hecho de papel engomado commercial para sustituir los panels
toxicos. A diferencia de los adhesivos que pueden pintarse, el material adhesive en el
dispositivo de papel no se adhiere a la piel del personal que debe trabajar con las
trampas. Una trampa de fondo abierto hecha de plastico opaco verde con un disposi-
tivo adhesive capture mas C. capitata que una trampa pintada de fondo cerrado con
un panel t6xico. Cuando es usada con la tecnologia de sueltas de insects est6riles, la
trampa seca de fondo abierto cebada con fagoatrayente sintetico a menudo capture C.
capitata salvaje en numero igual al capturado por las trampas de Jackson cebadas con
trimedlure, pero la trampa seca capture much menos C. capitata esteriles.

June, 1996

Behavioral Ecology Symposium '95: Heath et al.

A number of traps have been developed for detecting or monitoring populations of
adult pest tephritids (reviewed in Cunningham 1989, Economopoulos 1989). Some of
the earliest traps were baited with carbohydrates and fermenting fruit or sugar solu-
tions. Subsequently traps have used hydrolyzed proteins as bait. All of these sub-
stances emit food-based chemical cues. Synthetic chemicals and food-related
chemicals are the primary basis for the baits in traps used currently to monitor pop-
ulations of the Mediterranean fruit fly, Ceratitis capitata (Wiedemann). These traps
include trimedlure-baited Jackson traps (Beroza et al. 1961, Harris et al. 1971) for
male C. capitata (Nakagawa et al. 1970) and McPhail traps with aqueous protein bait
solutions (Newell 1936) which capture both female and male C. capitata. Traps baited
with aqueous protein solutions have been the primary tool for capturing female C.
capitata. These traps are cumbersome and have numerous other operational disad-
vantages as follows. Bait solution may be spilled easily while handling the trap and
spilled bait provides an attractant source outside the trap. Removal of insects trapped
is time consuming and tedious. The contents of the trap must be filtered through a
screen to separate the insects from the bait solution. Trapped fruit flies are often
found severely decomposed with parts missing and thus, when these traps are used in
conjunction with marked flies dispersed in sterile insect release programs, it is often
difficult to determine if a trapped fly is sterile or wild. Additionally deployment of
McPhail traps is difficult because of the size and weight of the trap. Also protein baits
attract a number of non-target insects and considerable time is required to sort the
trapped insects. Due to these difficulties associated with the use of liquid protein
baits, we investigated the development of a synthetic food-based lure (ammonium ac-
etate and putrescine) that could be used in a dry insect trap that would capture both
female and male C. capitata (Heath et al. 1995, Epsky et al. 1995).
We report herein a brief description of the original dry trap with food-based syn-
thetic attractant and of subsequent improvements.


Cylindrical Dry Trap with Painted 7.5 cm Band

The original trap is fully described in Heath et al. (1995). Briefly, it consists of a
clear plastic cylinder (9 cm diam by 15 cm long) made from acetate film with a painted
band (7.5 cm wide) in the middle of the trap which provides a visual cue for the fruit
fly (Fig. 1A). The ends of the traps are plastic Petri dishes glued shut that fit snugly
into the cylindrical trap body. A wire hanger that pierces both bottom and top allows
the trap to be hung in the crop canopy. The trap body has three equally-spaced holes
(2.2 cm diam) around the circumference of the cylinder midway between the ends to
allow release of volatile chemicals and access for the responding flies. Flies enter the
trap in response to the attractant and are killed after feeding on toxicant-bait panels
that are placed on the inside surface of the top and bottom of the trap. Captured flies
are removed from the trap by lifting the cylinder from the bottom and dumping the
flies into a small container for transport to the laboratory.

Comparison of Cylindrical Dry Trap and Standard McPhail Trap

Field trials were conducted in an orange and coffee finca (farm) located near Gua-
temala City, Guatemala. Traps were placed in orange trees for all tests, with place-
ment within a tree following standard protocol (Anonymous 1989). Traps tested were
dry traps with green painted bands, dry traps with orange painted bands, and

Florida Entomologist 79(2)


Figure 1. Illustration of the three types of cylindrical dry traps (9 cm diam by 15 cm
long) that were used in field studies conducted in Guatemala to test a food-based syn-
thetic attractant for C. capitata. A) Trap body constructed from clear acetate film with
a painted band (7.5 cm). The closed-bottom trap had toxicant panels placed on the top
and bottom to retain responding flies. B) Trap body constructed from clear acetate film
with a painted band (12 cm). The closed-bottom trap had toxicant panels placed on the
top and bottom to retain responding flies. C) Trap body constructed of green opaque
plastic. The open-bottom trap had a sticky paper insert to retain responding flies. Syn-
thetic attractant, which was formulated in membrane-based lures, can be seen in the
painted, closed-bottom traps.

McPhail traps baited with five torula yeast-borax pellets. Separate comparisons were
conducted with the dry traps baited with low, medium and high dosages of the syn-
thetic lures (Heath et al. 1995). To obtain the different doses, the release rate from the
ammonium acetate lure (Concep, Bend, OR) was adjusted by reducing the exposed
area of the release membrane. The area of exposed surface of the membrane tested
was 1.0 cm2 (low dose), 1.4 cm2 (medium dose), and a full patch having a membrane
surface of 4.0 cm in diam (high dose). Putrescine was formulated using polypropylene
vials (1 cm internal diam by 2.2 cm) and was tested at 50, 100, or 200 p1 (low, medium
and high doses, respectively). The vials were closed and hung in the trap in an upright
position so that the putrescine was allowed to pool in the bottom of the vial. Flies were
removed from traps every 2-3 d, and numbers of male and female C. capitata recorded.
The experiment was replicated over time and there were 18 consecutive replicates for
each dose. The mating status of C. capitata females from the different traps was de-
termined by dissecting a subsample of the trapped females and checking for presence
or absence of sperm in the spermathecae.

Dry Trap Visual Cue Modifications

Field trials were conducted in an orange and coffee finca located near Guatemala
City, Guatemala, to determine optimal width of the visual cue band on the dry trap for

June, 1996

Behavioral Ecology Symposium '95: Heath et al.

capture of C. capitata. All traps were baited with ammonium acetate lure at the me-
dium dose. The putrescine was formulated in a manner similar to the lures used for
Japanese beetle monitoring (Consep, Bend, OR) with a 3 mm diam area of exposed
membrane surface. In all trials, traps were placed in orange trees that were bearing
fruit. Treatments tested were traps with visual cue widths of 3, 7.5, 12 or 15 cm. Sep-
arate tests were conducted for traps with green bands and with orange bands. All four
treatments were placed around the periphery of a tree with approximately one to two
meters between adjacent traps. Initially, treatments were placed randomly within a
given tree, followed by sequential displacement of each trap to the next position
around the tree when the traps were sampled every 2-3 d. On each sample date, flies
were removed from traps, and numbers of male and female C. capitata were recorded.
Sets of traps were placed in three trees spaced approximately 20-30 m apart, and the
total number of flies trapped per treatment was the sum of each sex collected that day.
The experiment was replicated over time, and there were 14 consecutive replicates of
tests of traps with each color. Sum totals were converted to percentage trapped per
trap type for analysis to facilitate comparisons among the range of fruit fly population
densities sampled.

Development of Open-Bottom Dry Trap with Sticky Insert

All tests were conducted with dry traps baited with the medium dosage of ammo-
nium acetate and putrescine. Trap bodies were either acetate cylinders 15 cm tall
painted with 12 cm green visual cues (designated "painted") or green opaque cylinders
15 cm tall (designated "opaque"). Moreover, the cylinders were either closed at both
ends with toxicant panels on the inside of the top and bottom (designated "closed-bot-
tom"), or they lacked a bottom and had a sticky insert (8 by 13 cm rectangle with ad-
hesive on both front and back) hung from the top (designated "open-bottom"). Traps
were tested in a citrus orchard near Guatemala City, Guatemala. There were four
lines of traps with all four trap types placed randomly within each line, one per tree,
according to standard protocol. Traps were checked every 3-7 days depending on pop-
ulation levels, and the numbers of C. capitata females and males were recorded. Data
were converted to percentage per trap type per sample date for statistical analysis.
The experiment was replicated over time and there were 6 consecutive replicates of
the test.

Open-Bottom Dry Trap Jackson Trap Comparison for Wild and Sterile Fly Capture

Green opaque cylindrical traps with sticky inserts were field-tested in Guatemala
to compare performance of the dry trap with that of the trimedlure-baited Jackson
trap. One of the tests was conducted in a coffee finca with a population of wild flies.
This site was at a high elevation (about 1300 m) and was located near Coatepeque,
Guatemala. Five lines of traps were placed at the site, with 50 m between each line.
Jackson traps with standard white inserts and open-bottom dry traps were placed in
each line along with other trap configurations. Traps were placed alternately 30 m
apart along a line. Ten traps of each type were placed following standard protocol. Ap-
proximately 20,000 sterile flies were released weekly distributed evenly over the trap-
ping area. Traps were checked weekly, and numbers ofC. capitata females and males
were recorded. All flies captured were sent to the Moscamed laboratory in Coatepeque
to be examined to determine if they were sterile or wild, following standard protocol.
Tests were conducted for 8 consecutive weeks.

Florida Entomologist 79(2)

Statistical Analysis

Data were analyzed with two sample t-tests using Proc TTEST or factorial analy-
sis of variance (ANOVA) using Proc GLM (SAS Institute 1985), depending on the
number of test factors in the study. Data were assessed by the Box-Cox procedure (Box
et al. 1978) and were transformed appropriately when necessary to stabilize the vari-
ance prior to ANOVA. Chi-square analysis using Proc FREQ (SAS Institute 1985) of
contingency tables of mating status by trap type within each dosage were made to
compare mating status of females trapped in the dry trap-McPhail trap comparisons.


Dry Trap and McPhail Trap Comparison

The dry traps baited with either the medium or the high dosage of synthetic at-
tractant caught numbers of Mediterranean fruit flies equal to those captured in
McPhail traps (Table 1). Although McPhail traps caught more female Mediterranean
fruit flies than dry traps baited with the low dosage of synthetic blend, more unmated
females were captured in the dry traps (Table 2). Increase in dosage of the synthetic
lure increased the percentage of mated females captured in the dry traps. A green dry
trap baited with a medium dose of synthetic attractant was selected for further eval-
uation because it was equal to a protein-baited McPhail trap in capture of females and
approximately 50% of the females captured were unmated.


Orange Green McPhail F P

Low Dose

C. capitata females 13.2 (2.08) 23.6 (3.95) 35.4 (5.46) 8.0 0.0009
C. capitata males 4.8 (0.71) 9.8 (1.44) 14.2 (2.37) 13.0 0.0001

Medium Dose

C. capitata females 10.6 (1.19) 10.9 (2.11) 18.6 (3.89) 1.5 ns
C. capitata males 3.8 (0.62) 3.4 (0.59) 6.6 (1.65) 0.6 ns

High Dose

C. capitata females 12.6 (2.07) 16.9 (4.44) 17.2 (3.15) 0.6 ns
C. capitata males 4.3 (1.18) 6.2 (2.19) 5.4 (1.32) 0.1 ns

Oneway analysis of variance on log [x + 1.0] transformed data, non-transformed means present; ns = not sig-
nificant. P > 0.05.

June, 1996

Behavioral Ecology Symposium '95: Heath et al.


Dosage of Unmated Females Trapped (%)
Blend Orange Green McPhail Chisquare P

Low 54.8b 69.0b 22.0a 21.00 0.0001
Medium 45.4b 51.7b 25.0a 7.25 0.027
High 12.6ab 4.0a 21.0b 6.59 0.037

'Chisquare analysis based on 2 by 3 contingency table within each dosage. Means within a row followed by
the same letter are not significantly different (2 by 2 contingency tables of two-at-a-time comparisons within a
dose, P= 0.05).

Dry Trap Visual Cue Modifications

Average (std. dev.) captures ofC. capitata females and males were 104.9 (69.2) and
49.0 (59.5) respectively. Width of the visual cue affected percentage capture of C. cap-
itata females but not of C. capitata males (Table 3). More C. capitata females were
captured in traps with green visual cues that were 12-15 cm wide than in traps with
3-7.5 cm wide cues. We used a green dry trap with a 12-cm wide visual cue and baited
with the medium dose of synthetic attractant for subsequent tests (Fig. 1B).

Development of Open-Bottom Dry Trap with Sticky Insert

The results of the comparisons among all four trap body/insect retention system
configurations are shown in Fig. 2. Capture of females was significantly higher in the
green opaque open-bottom trap than in the other traps tested (F = 16.09; df = 3, 20; P
= 0.0001). The green opaque open-bottom trap captured an average of 3.4 times as
many females as the painted traps. Capture of males was also affected by trap type (F
= 5.16; df = 3,20; P = 0.0084). The female:male ratio in the open-bottom dry trap was
about 6:1 and 82% of the C. capitata captured were females. We used a green opaque


3 cm 7.5 cm 12 cm 15 cm F P

Females 18.4 (2.16) 18.5 (1.91) 31.5 (2.46) 31.7 (2.55) 10.36 0.0001
Males 20.8 (4.04) 24.6 (4.29) 32.4 (3.18) 22.2 (3.07) 1.90 ns

Oneway analysis of variance on square root [x + 0.5] transformed data, non-transformed means presented;
ns = not significant, P > 0.05.

Florida Entomologist 79(2)

June, 1996

open-bottom dry trap baited with the medium dose of synthetic attractant for subse-
quent tests (Fig. 1C).

Open-Bottom Dry Trap Jackson Trap Comparison for Wild and Sterile Fly Capture

Numbers of wild and sterile C. capitata captured over time are shown in Fig. 3.
Capture with the open-bottom dry trap was greater than with the Jackson trap for

50 A



a, -__ fflJflfiii'NB ^^^f^'T^f^ ^^nf~f~

Closed-Bottom Open-Bottom Closed-Bottom Open-Bottom

Painted Trap Body

Opaque Trap Body

Figure 2. Percentage and standard error of C. capitata females (A), males (B) and
total females plus males, C) captured in four configurations of cylindrical dry traps
baited with food-based synthetic attractants in field trials conducted in Palin, Guate-



Behavioral Ecology Symposium '95: Heath et al.


Figure 3. Capture of (A) wild male C. capitata in Jackson traps (open circle, dashed
line), wild male plus female C. capitata in open-bottom dry traps (solid diamond, solid
line), (B) wild male C. capitata in Jackson traps (open circle, dashed line), wild male
(solid circle, solid line), and female (star, solid line) C. capitata in open-bottom dry
traps, and (C) sterile males in Jackson traps (open circle, dashed line) and sterile
males plus females (solid diamond, solid line) in open-bottom dry traps in field trials
conducted in Guatemala.

seven of the eight weeks and was approximately equal in the remaining week. The fe-
male to male ratio in the open-bottom dry trap was 4:1. There were 3,875 sterile males
captured in the Jackson traps compared to 235 sterile males captured in the open-bot-
tom dry trap. Thus, approximately 16.5 times more sterile males were captured in the
Jackson trap than in the open-bottom dry trap. The open-bottom dry trap captured
415 sterile females.

Florida Entomologist 79(2)


The development of the open-bottom dry trap offers several advantages in addition
to the increase in capture of C. capitata. The panel is easy to insert and to remove from
the trap, and no pesticide is needed. Field personnel in Guatemala reported that they
like this trap because it is similar to the insert system currently used in Jackson traps
containing trimedlure, it is very "user friendly", and it is even more durable than the
closed-bottom, painted dry trap. It also appears to be more sensitive to low density
populations of C. capitata and could provide an improved detection tool for areas in
which C. capitata populations are not established.
Continued research in the development of facile trapping systems will afford sev-
eral new opportunities in efforts to control and eradicate the Mediterranean fruit fly.
It is envisioned that additional attractants will be identified and further optimization
of visual cues and trap design will occur. Use of the trapping systems described will
enable better utilization of resources related to sterile insect release technology, in-
cluding but not limited to decreased efforts and greater accuracy in sterile/wild insect
identification and minimized capture of sterile males after release. Related research
indicates that the synthetic lure could be used in traps for capture of several of the
Anastrepha spp. that also are economically important pests of fruit.


The authors thank Ara Manukian (USDA/ARS, Gainesville, FL), John Howell,
Sanjay Kulkarni, Yasmin Cardoza, Jennifer Braum (Univ. of Florida, Gainesville, FL)
for technical assistance; Antonio Guzman, (Guatemala City, Guatemala) and Oscar
Castro (Moscamed, Guatemala) for conducting field trials; and Waldemar Klassen, Ri-
chard Baranowski (Univ. of Florida, Homestead, FL), Tim Holler (USDA/APHIS,
Gainesville, FL) and Peter Landolt (USDA/ARS, Gainesville, FL) for critically review-
ing the manuscript. Carlos Lira and the staff at Moscamed, Guatemala, and Gordon
Tween and the staff at the USDA, APHIS-International Services, U.S. Embassy, Gua-
temala provided invaluable support of this research. This research was funded in part
by the California Department of Food and Agriculture (Grants #91-0621, #93-0478).
This article reports the results of research only. Mention of a proprietary product does
not constitute an endorsement or recommendation for its use by the USDA or the Uni-
versity of Florida.


ANONYMOUS. 1989. Florida fruit fly detection manual. USDA/APHIS, Plant Protec-
tion and Quarantine, and Florida Department of Agriculture and Consumer
Services, Division of Plant Industry, Gainesville, FL.
sect attractants. New attractants for the Mediterranean fruit fly. J. Agric. Food
Chem. 9: 361-365.
Box, G. E. P., W. G. HUNTER, AND J. S. HUNTER 1978. Statistics for Experimenters.
An Introduction to Design, Data Analysis, and Model Building. J. Wiley & Sons,
New York, NY.
CUNNINGHAM, R. T. 1989. Population detection, pp. 169-173 in A. S. Robinson and G.
Hooper [eds.], World Crop Pests, Volume 3B, Fruit flies, their biology, natural
enemies and control. Elsevier Science Publishers B. V., Amsterdam. 447 pp.
ECONOMOPOULOS, A. P. 1989. Use of traps based on color and/or shape, pp. 315-327 in
A. S. Robinson and G. Hooper [eds.], World Crop Pests, Volume 3B, Fruit flies,
their biology, natural enemies and control. Elsevier Science Publishers B. V.,
Amsterdam. 447 pp.

June, 1996

Behavioral Ecology Symposium '95: Heath et al. 153

EPSKY, N. D., R. R. HEATH, A. GUZMAN, AND W. L. MEYER 1995. Visual cue and chem-
ical cue interactions in a dry trap with food-based synthetic attractant for Cer-
atitis capitata and Anastrepha ludens (Diptera: Tephritidae). Environ.
Entomol. 24: 1387-1395.
HARRIS, E. J., S. NAKAGAWA, AND T. URAGO. 1971. Sticky traps for detection and sur-
vey of three tephritids. J. Econ. Entomol. 64: 62-65.
MEYER 1995. Development of a dry plastic insect trap with food-based syn-
thetic attractant for the Mediterranean and Mexican fruit flies (Diptera: Te-
phritidae). J. Econ. Entomol. 88: 1307-1315.
NAKAGAWA, S., G. J. FARIAS, AND L. F. STEINER 1970. Response of female Mediterra-
nean fruit flies to male lures in the relative absence of males. J. Econ. Entomol.
63: 227-229.
NEWELL, W. 1936. Progress report on the Key West (Florida) fruit fly eradication
project. J. Econ. Entomol. 29: 116-120.
SAS INSTITUTE. 1985. SAS/STAT guide for personal computers, version 6 edition. SAS
Institute, Cary, NC.

Behavioral Ecology Symposium '95: Giblin-Davis et al.


'Fort Lauderdale Research and Education Center
University of Florida, IFAS
3205 College Avenue, Fort Lauderdale, FL 33314

'Department of Chemistry, Simon Fraser University
Burnaby, British Columbia, Canada, V5A 1S6

3Department of Biology, Simon Fraser University
Burnaby, British Columbia, Canada, V5A 1S6

4ASD de Palma Tica, San Jose, Costa Rica

'Tropical Research and Education Center
University of Florida, IFAS
18905 S. W. 280 Street, Homestead, FL 33031

"ChemTica International, Apdo 159-2150, San Jose, Costa Rica


Palm weevils in the subfamily Rhynchophorinae (Curculionidae) (Rhynchophorus
spp., Dynamis borassi, Metamasius hemipterus, Rhabdoscelus obscurus, and Parama-
sius distortus) use male-produced aggregation pheromones for intraspecific chemical
communication. Pheromones comprise 8, 9, or 10 carbon, methyl-branched, secondary
alcohols. (4S,5S)-4-Methyl-5-nonanol (ferrugineol) is the major aggregation phero-

Florida Entomologist 79(2)

mone for R. ferrugineus, R. vulneratus, R. bilineatus, M. hemipterus, and D. borassi
and a minor component for R. palmarum. (5S,4S)-5-Methyl-4-octanol (cruentol),
(3S,4S)-3-methyl-4-octanol (phoenicol), and (4S,2E)-6-methyl-2-hepten-4-ol (rhyn-
chophorol) are the main aggregation pheromones for R. cruentatus, R. phoenicis, and
R. palmarum, respectively. Plant kairomones strongly enhance pheromone attrac-
tiveness but none of the identified volatiles, such as ethyl acetate, ethyl propionate, or
ethyl butyrate are as synergistic as fermenting plant (palm or sugarcane) tissue.
Studying orientation behavior of foraging weevils to semiochemical devices helped to
design and test traps for weevil capture. Generally, 3 mg per day of synthetic phero-
mone (with non-natural stereoisomers being benign) plus insecticide-treated plant
tissue constitute highly attractive trap baits. Potential exists for pheromone-based
mass-trapping of weevils to reduce their populations and the spread of the weevil-vec-
tored red ring disease, for monitoring their population dynamics to facilitate pest
management decisions, and for detection and possible interception of non-native wee-
vils at ports of entry.

Key Words: Aggregation pheromones, behavior, Bursaphelenchus cocophilus, chemi-
cal ecology, Dynamis borassi, kairomones, Metamasius hemipterus, nematode, Para-
masius distortus, red ring nematode, Rhabdoscelus obscurus, Rhynchophorus spp.

Los picudos de la palma pertenecientes a la familiar Rhyncophorinae (Curculioni-
dae) (Rhynchophorus spp., Dynamis borassi, Metamasius hemipterus, Rhabdoscelus
obscurus, yParamasisus distortus) utilizan feromonas de agregaci6n de los machos en
la comunicaci6n quimica intraespecifica. Las feromonas comprenden alcoholes secun-
darios con ramas de methyl de 8, 9 o 10 carbonos. (4S,5S)-4-Methyl-5-nonanol (ferru-
gineol) es la feromona de agregaci6n principal Para R. ferrugineus, R. vulneratus, R.
bilineatus, M. hemipterus y D. borassi y un component menor para R. palmarum.
(5S,4S)-5-Methyl-4-octanol (cruentol), (3S,4S)-3-methyl-4-octanol (phoenicol) y (4S,
2E)-6-Methyl-2-hepten-4-ol (rhyncophorol)son las feromonas de agregaci6n principa-
les para R. cruentatus, R. phoenicis, and R. palmarum, respectivamente. Las kairo-
monas vegetables incrementan fuertemente la atractividad de la feromona pero
ninguno de los volatiles identificados, tales como acetato de etilo, propionato de etilo,
o burtirato de etilo, son tan sinergicos como el tejido de una plant (palma o cania de
azucar) en estado de fermentaci6n. El studio del comportamiento de la orientaci6n de
los picudos forrajeando hacia dispositivos semioquimicos ayud6 a diseiar y probar
trampas para la capture de los mismos. Generalmente, 3 mg por dia de feromona sin-
tetica (con estereois6meros no naturales) mas tejido vegetal tratado con insecticide
constitute un cebo muy atractivo. Existe el potential para la capture masiva de picu-
dos con feromonas para reducir sus poblaciones y contener la diseminaci6n de la en-
fermedad que ellos transmiten, enfermedad del anillo rojo, asi como para muestrear
su dinamica de poblaci6n y facilitar las deciciones de manejo de la plaga, y para la de-
tecci6n y possible intercepci6n de picudos no natives en los puertos de entrada.

The observation that adults of Rhynchophorus cruentatus Fab. were attracted to
nitrocellulose lacquer-based automobile paint in Sanford, FL (Bare 1929) led to re-
search in the last few years on semiochemical-based communication of palm weevils
in the subfamily Rhynchophorinae. Semiochemicals are defined as chemicals that act
as signals between organisms (Dusenbery 1992). Aggregation or sex pheromones
serve as intraspecific transitory signals, whereas kairomones, such as host-plant vol-
atiles, act interspecifically with the receiver benefiting (Dusenbery 1992). As we now
know, several possible components or contaminants of lacquer paints of the 1920s act

June, 1996

Behavioral Ecology Symposium '95: Giblin-Davis et al. 155

as semiochemicals for foraging R. cruentatus. For example, 5-methyl-4-octanol and
ethyl acetate, both possible constituents of lacquer paint, function as aggregation
pheromone and host kairomone, respectively, for R. cruentatus (Giblin-Davis et al.
1994b). Thus, the freshly painted automobiles standing outside Kent's garage during
the summer of 1928 (Bare 1929), superficially resembling behemoth beetles in the
sweltering heat and humidity, sent a clear "come hither" message to all palmetto wee-
vils that could detect them and fly.
For the purpose of this paper, the discussion is limited to palm weevils in the sub-
family Rhynchophorinae, including Rhynchophorus spp., Dynamis borassi Fab., Meta-
masius hemipterus L., Rhabdoscelus obscurus (Boisduval), and Paramasius distortus
(Gemminger & Harold) (=Metamasius inaequalis [Gyllenhal 1838]). Palm weevils are
defined here as weevils that use palms as a major host for larval development, even
though they may colonize other monocots such as sugarcane and banana (Giblin-
Davis et al. 1989, Napompeth et al. 1972, Pefia et al. 1995, Vaurie 1966, Wattanapong-
siri 1966). Other species of weevils that are associated with palms, such as the small
pollinators of the African oil palm, Elaeidobius spp. (subfamily: Erirhininae) (O'Brien
& Woodruff 1986), Parisoschoenus spp. (Baridinae) (O'Brien pers. comm.), Homalino-
tus spp. (Cholinae) (Vaurie 1973), and Rhinostomus spp. (Rhynchophorinae) (Vaurie
1970) will not be discussed.


The basic life cycles of Rhynchophorus spp., D. borassi, M. hemipterus, R. obscurus,
and P. distortus are similar. Adult female weevils are attracted to, and oviposit 30-400
eggs in, damaged (pruning or chain saw wounds), stressed, or healthy palms (Napom-
peth et al. 1972, Wattanapongsiri 1966, Weissling & Giblin-Davis 1994). Larvae bore
into the palms and after several instars develop into adults in about 2 months (Giblin-
Davis et al. 1989, Napompeth et al. 1972, Watananpongsiri 1966). Larvae of Rhyn-
chophorus spp., D. borassi, M. hemipterus, R. obscurus, and P. distortus appear to in-
habit slightly different niches in palms, with Rhynchophorus spp. preferring the
crown and/or stem (Wattanapongsiri 1966, Giblin-Davis & Howard 1989, Giblin-
Davis et al. 1995b); D. borassi living tissue of coconut inflorescence and stem (Gerber
et al. 1990, Wattanpongsiri 1966); and M. hemipterus, R. obscurus, and P. distortus
petioles and stem periphery, sometimes entering the crown (Giblin-Davis et al. 1995b,
R.M.G-D. & C.M.C., unpublished observation, Halfpapp & Story 1991). The last instar
larvae of these weevils move to the petioles or to the rind of the stem to prepare a co-
coon from coarse fibers and become a prepupa, pupa, and adult within several weeks
(Wattanapongsiri 1966, Vaurie 1966). As borer damage from most of these species ac-
cumulates, fermentation volatiles increase rendering the host more apparent to
adults of con- and heterospecifics.
Fermented sap exuding from dead or wounded palms (Sabal palmetto [Walter] and
Phoenix canariensis Hortorum ex Chabaud) is highly attractive to R. cruentatus
(Chittenden 1902). Attractiveness of chopped, fermenting, S. palmetto crown and
stem tissue peaks 24-72 hours after cutting, whereas cut surfaces of felled palms re-
main attractive for 35 days (Weissling et al. 1992). Moist fermenting (stem) tissues
from various palm species, fruits, sugarcane, pineapple, and molasses (plus water) are
similarly attractive to palm weevils (Diegado & Moreno 1986, Giblin-Davis et al.
1994a,b, 1995a). Plant tissue [Serrenoa repens (Bartram) stem] or molasses with min-
imal moisture content are significantly less attractive to R. cruentatus and M. hemi-
pterus, respectively (Giblin-Davis et al. 1994b, 1995a). Early fermentation volatiles of
moist and stressed, damaged, or dying host plant tissues with high sugar content ob-

Florida Entomologist 79(2)

viously provide olfactory cues to attract palm weevils. Dynamis borassi, in contrast,
usually prefers fresh tissue (R.M.G-D. unpublished observation).


Laboratory and field work with R. obscurus provided the first evidence that males
of palm weevils produce male and female-attracting aggregation pheromones (Chang
et al. 1971, Chang & Curtis 1972). Subsequently, male-produced aggregation phero-
mones have been demonstrated for many species in the subfamily Rhynchophorinae
[i.e., R. palmarum (L.) (Moura et al. 1989, Rochat et al. 1991a), R. cruentatus
(Weissling et al. 1993), R. phoenicis (Fab.) (Gries et al. 1993), R. ferrugineus (Olivier)
and R. vulneratus (Panzer) (Hallett et al. 1993b), M. hemipterus (Giblin-Davis et al.
1994a, Rochat et al. 1993b), Cosmopolites sordidus Germar (Budenberg et al. 1993),
Sitophilus spp. (Walgenbach et al. 1983)].
Successful elucidation of pheromones and host kairomones for palm weevils has
required collaboration between field entomologists, chemists, and physiologists. For
semiochemical identification, 10-50 male or female weevils with different mating,
rearing and/or feeding status are placed in modified 9-liter Nalgene polycarbonate
desiccators with a source of moisture (2 Petri dishes filled with 40 ml of deionized wa-
ter and tissue paper each) or with plant tissue (e.g., 250 g of chopped palm stem or
sugarcane). Charcoal prefiltered air is drawn (1.5 cm3 per min) for 5-7 days through
the desiccator and downstream through a glass tube filled with an adsorbent matrix
(Porapak Q). Volatile chemicals are eluted from the Porapak Q with redistilled pen-
tane, concentrated by distillation (Oehlschlager et al. 1988,1992) and subjected to gas
chromatographic (GC) analyses [Hewlett Packard (HP) 5885B gas chromatograph]
employing both flame ionization (FID) and electroantennographic detection (EAD)
(Arn et al. 1975). A variety of GC-columns and GC-temperature programs are used for
separation of semiochemicals (Perez et al. 1994). In GC-EAD recordings, volatiles elic-
iting responses by male or female weevil antennae are analyzed by GC-mass spec-
trometry (GC-MS) in both electron impact (EI; 70 eV) and chemical ionization (CI)
modes [HP 5985B GC-mass spectrometer (MS) equipped with the same column as
used in GC-EAD recordings]. If required, Kratos MS80RFA high resolution MS and/
or nuclear magnetic resonance (NMR) spectroscopy is used to elucidate chemical
Enantioselective pheromone production and response are common in the Co-
leoptera and help fine-tune intra- and interspecific semiochemical communication
(Seybold 1993). Because non-weevil-produced stereoisomers in synthetic pheromones
may cause interruption in response (Seybold 1993), enantiomer(s) and/or stereoiso-
mers of the weevil-produced pheromone are determined on a chiral, isomer separat-
ing column (Perez et al. 1994). Weevil-produced and EAD-active isomers in synthetic
stereoisomeric mixtures are stereoselectively synthesized and tested for behavioral
activity in field experiments (Gries et al. 1993, Oehlschlager et al. 1992, 1995b, Perez
et al. 1994, 1995a,b). Methods of synthesis depend upon the chemical structure of the
semiochemical (Gries et al. 1993, Mori et al. 1993a,b, Oehlschlager et al. 1992, 1995b,
Perez et al. 1994, 1995a, Rochat et al. 1991b, Weissling et al. 1994b).
Major and minor male-produced aggregation pheromones for palm weevils (Table
1) are 8, 9, or 10 carbon, methyl-branched, secondary alcohols. Only the pheromone
for R. palmarum, (2E)-6-methyl-2-hepten-4-ol, is unsaturated. Typically, the (S)-
enantiomer or (S,S) stereoisomer is produced by the weevils and elicits a behavioral
response [(R. palmarum, Oehlschlager et al., 1992) (R. bilineatus (Montrouzier) (Oe-
hlschlager et al. 1995b), R. cruentatus and R. phoenicis (Perez et al. 1994, Rochat et
al. 1995), R. ferrugineus and R. vulneratus (Perez et al. 1995b), and M. hemipterus

June, 1996

Behavioral Ecology Symposium '95: Giblin-Davis et al. 157

(Perez et al. 1995a)]. Synergistic behavioral activity of (4S), (4R), or () 2-methyl-4-
heptanol in M. hemipterus constitutes the only exception to this general rule (Perez et
al. 1995a). Non-natural occurring (non-weevil-produced) stereoisomers in synthetic
pheromones are behaviorally benign (non-interruptive) thus allowing application of
racemic blends for operational use.
Lethal traps baited only with aggregation pheromones or host kairomones are not
very attractive to palm weevils, but in combination synergize attractiveness 8-20 fold
[R. palmarum (Oehlschlager et al. 1992), R. phoenicus (Gries et al. 1993, Perez et al.
1994, Rochat et al. 1995), R. vulneratus and R. ferrugineus (Hallett et al. 1993a,b) R.
bilineatus (Oehlschlager et al. 1995b), R. cruentatus (Giblin-Davis et al. 1994b,
Weissling et al. 1993, 1994b, Perez et al. 1994), and M. hemipterus (Giblin-Davis et al.
1995a, Perez et al. 1995a)]. These weevils appear to be opportunistic oligophages, re-
sponding to early fermentation volatiles from wounded or stressed hosts and recruit-
ing conspecifics over long distances with male-produced aggregation pheromones
(Jaff6 et al. 1993, Weissling et al. 1994b).
Ferrugineol [(4S,5S]-4-methyl-5-nonanol)] is the major aggregation pheromone for
Asian R. ferrugineus, R. vulneratus, and R. bilineatus and for the neotropical M. hemi-
pterus and D. borassi (Oehlschlager et al. 1995b, Perez et al. 1995a,b) and a possible
minor component for R. palmarum (Hallett et al. 1993a). Other members of the Rhyn-
chophorinae, i.e. Sitophilus spp., use the same stereoisomer of their major male-pro-
duced aggregation pheromone (Walgenbach et al. 1987). Identical male-produced
aggregation pheromones have also been reported for geographically isolated African
and Asian rhinoceros beetles (Oryctes spp.; Scarabaeidae) (Gries et al. 1994b, Hallett
et al. 1995).
Production of, and response to, ferrugineol which is present in many of the palm
weevils may be pleisomorphic, whereas other pheromones such as (5S,4S)-5-methyl-
4-octanol (cruentol; R. cruentatus), (3S,4S)-3-methyl-4-octanol (phoenicol; R. phoeni-
cis), and (4S,2E)-6-methyl-4-hepten-2-ol (rhynchophorol; R. palmarum) may repre-
sent a more derived (apomorphic) character. Alternatively, M. hemipterus may
represent a more primitive form because males produce a larger number of alcohols
and corresponding ketones (Table 1) than most Rhynchophorus species which typi-
cally produce one or two male produced volatiles. The premise that R. palmarum
might have speciated from Asian weevils (Griffith 1987, Oehlschlager et al. 1995b)
with the movement of coconut, Cocos nucifera L., from Melanesia to the Neotropics is
not tenable because of the recency (< 1000 years ago) of coconut introduction by man,
and the fact that Rhynchophorus species do not appear to have greatly expanded their
ranges until very recently (Wattanapongsiri 1966). Fossil evidence suggests that
there was a diverse and widespread palm flora during the early to mid-Cretaceous at
about the time when Laurasia and Gondwanaland separated (Uhl & Dransfield
1987). Early protopalm weevils might have been associated with early palms (Thri-
nacinae-like). Radiation of different palm groups and their associated protopalm wee-
vils might have begun prior to the separation of Laurasia and Gondwanaland.
Continued continental drift may have prevented gene flow between African, North
and South American, and Asian palm weevils resulting in the contemporary pattern
of reproductive behavior and aggregation pheromones. Knowledge about pheromone
biosyntheses and dissemination for palm weevils will be important for testing evolu-
tionary hypotheses.
The role of minor EAD-active, male-produced volatiles in palm weevils is not well
understood. High doses of the four ketones in males ofM. hemipterus (Giblin-Davis et
al. 1994a, Perez et al. 1995) reduce attractiveness of the main pheromone. In contrast,
low doses of () 2-methyl-4-heptanol in M. hemipterus (Perez et al. 1995a), slightly en-
hance pheromonal attractiveness. 3-Pentanol and 2-methyl-4-octanol in M. hemi-

Florida Entomologist 79(2)



o C -r
E h~9t t ~WE-
c N c j a 3
E d3 _,M
5 5 Cs c C C4 Cj


e C



E ~
E ~

rn, P

June, 1996


q,) c^

"+3 oj



^ m
c~ -t
^0 0

0 0



- -S ca

fCl -,

I0 I
g in



c~ c~

ct 10
n 1

C) o

-t -
10 O

%% kk
.E E ~E ~E
c~C7 c~c~


Behavioral Ecology Symposium '95: Giblin-Davis et al. 159

pterus (Perez et al. 1994) and 2,3-epoxy-6-methyl-4-heptanol (Rochat et al. 1993a) and
4-methyl-5-nonanol in R. palmarum (Hallett et al. 1993a) are behaviorally benign.
Aeration of many confined weevils for long periods may have induced atypical produc-
tion of compounds or distorted ratios thereof. For example, 24-hour aerations of 5-12
M. hemipterus (subspecies M. h. hemipterus?) from Guadeloupe revealed 5 male-spe-
cific compounds (Rochat et al. 1993b), whereas 96 hour aerations of 38 males ofM. h.
sericeus from Florida disclosed 8 male-specific volatiles (Table 1) (Perez et al. 1995a).
The question remains whether these chemical differences represent divergence in the
chemical ecology of the two subspecies or artifacts from different methodology.
Capture of neotropical palm weevils in lethal traps baited with aggregation pher-
omones of heterospecifics suggests synomonal pheromone activity. For example, D. bo-
rassi, M. hemipterus, and P. distortus are attracted to the aggregation pheromone of
R. palmarum (Chinchilla et al. unpublished data, Giblin-Davis et al. unpublished
data) and P. distortus also responds to the aggregation pheromone ofM. hemipterus
(Chinchilla et al. unpublished data). Cross-attraction of sympatric weevils may have
evolved because of its adaptive significance in overcoming a palm's defense and time-
efficient use of a temporarily suitable resource. Minor volatile components may serve
as intra- or interspecific semiochemicals mediating resource partitioning. Niche di-
vergence and larval cannibalism in R. palmarum, M. hemipterus, D. borassi, and P.
distortus may reduce interspecific competition in an occupied host. Further work with
pheromone combinations in palm weevils, such as has been done for Carpophilus spe-
cies (Bartelt et al. 1995), could simplify the use of pheromones for monitoring or mass-
trapping palm weevils in the Neotropics.
In dose response trapping experiments, 3-7 mg per d of synthetic pheromone cost-
effectively attracts palm weevils (Gries et al. 1994a, Hallett et al 1993b, Oehlschlager
et al. 1992, 1993b, Weissling et al. 1994b). In the presence of plant tissue, 0.3, 20, or
200 mg per d of ()-rhynchophorol (Oehlschlager et al. 1993b), 0.3, 1, or 3 mg per d of
(+)-ferrugineol (Hallett et al. 1993b), 0.4 or 4 mg per d of (+)-cruentol (Weissling et al.
1994b) equally attract their respective weevils. Release rates of 3-4 mg per d rather
than 0.3-0.4 mg per d of pheromone effectively attract R. phoenicis, R. ferrugineus,
and R. bilineatus (Gries et al. 1993, Hallett et al. 1993b, Oehlschlager et al. 1995b).
Release rates exceeding 3 mg per d of ()-ferrugineol reduced attraction ofR. bilinea-
tus (Oehlschlager et al. 1995b). Ten-fold increase in the binary pheromone blend of
(+)-4-methyl-5-nonanol and ()-2-methyl-4-heptanol (8:1 ratio; "metalure"; 30 mg per
d) plus sugarcane only doubled trap captures of M. hemipterus in Costa Rica (Oeh-
schlager et al. unpublished data), thus not justifying the increased lure costs. For ex-
perimental and commercial use, racemic pheromone (about 90 mg total) with an
indicator dye is hermetically sealed in a polymer membrane release device (ChemTica
International, San Jose, Costa Rica) for slow and constant pheromone dissemination
[about 3 or 7 mg per d at 25 C for 2-3 months (Oehlschlager et al. 1995a,b)].
Some of the palm weevils seem to respond to cuticle-bound courtship pheromones
of conspecific females (Rochat & Zagatti 1993). Male R. palmarum exhibit precopula-
tory behavior (a jerky swinging motion of the body) after antennating dead females
but not dead male decoys or decoys that had been washed with hexane (Rochat &
Zagatti 1993). As part of their precopulatory behavior, male R. cruentatus antennate
the pronotum of live conspecifics and mount males and females with equal aplomb
(Weissling unpublished observation). Cuticular lipids in hexane or methylene chlo-
ride extracts of R. cruentatus adults lack sex-specific differences when analyzed by
GC-MS or thin layer chromatography (Giblin-Davis et al. unpublished data). Inter-
specific mating pairs of R. ferrugineus and R. vulneratus suggest that these weevils
have similar or identical precopulatory behavior, rely on postzygotic reproductive iso-
lating mechanisms, or are conspecifics (Hallett et al. 1993b).

Florida Entomologist 79(2)


Early research provided evidence that general fermentation volatiles, such as eth-
anol, appeared attractive to Rhynchophorus weevils (Gunatilake & Gunawardena
1986, Hagley 1965). The recent discovery of aggregation pheromones which greatly
enhance attractiveness of plant tissues facilitated identification of individual palm
kairomones (Giblin-Davis et al. 1994b, 1995a, Gries et al. 1994a, Jaffe et al. 1993,
Perez et al. 1995a).
Kairomones identified by GC-EAD and GC-MS include the "palm esters" ethyl ac-
etate, ethyl propionate, ethyl butyrate, and ethyl isobutyrate for R. phoenicis, R.
cruentatus, R. palmarum, R. bilineatus, R. ferrugineus, and R. vulneratus (Gries et al.
1994a). Ethyl propionate (30 mg per d) was the only "palm ester" tested that syner-
gized attraction of R. phoenicis to ()-phoenicol (Gries et al. 1994a). Ethyl acetate,
ethyl propionate, ethyl butyrate, ethyl isobutyrate, ethyl (S)-(-)-lactate, and ethanol
at various release rates synergized attraction ofR. cruentatus to ()-cruentol (Giblin-
Davis et al. 1994b, Giblin-Davis, unpublished), and ethyl acetate of unknown release
rate enhanced attraction of R. palmarum to ()-rhynchophorol (Jaffe et al. 1993).
Each of the fermenting sugarcane volatiles, ethyl acetate (30 mg per d), ethyl propi-
onate (20 mg per d), and ethyl butyrate (20 mg per d) equally enhanced attraction of
M. hemipterus to metalure (Perez et al. 1995a). None of the "palm esters" tested to
date with pheromone has been as effective as palm or sugarcane tissue in enhancing
pheromonal attractiveness (Giblin-Davis et al. 1994b, 1995a, Gries et al. 1994a, Jaffe
et al. 1993, Perez et al. 1995a), suggesting additional as yet unknown palm
kairomones or improper ratios of tested components. Peak field attraction of chopped
palm or sugarcane tissue within 2 to 5 days indicates proportional changes in volatile
chemicals from fermentation over time, affecting optimal attraction of weevils to
traps (Gries et al. 1993, Hallett et al. 1993a, Weissling et al. 1992). Proportional
changes in volatile chemicals from fermenting or fermented palm sap can be attrib-
uted to the abiotic conditions and microflora present (Nagnan et al. 1992, Samara-
jeewa et al. 1981).
Ethyl acetate released at 400-900 mg per d significantly enhanced attraction ofM.
hemipterus to metalure, sugarcane, or both (Giblin-Davis et al. 1995a). Because ethyl
acetate at 400-1600 mg per d is not repellent to R. cruentatus or M. hemipterus (Gib-
lin-Davis et al. 1994b, 1995a), it may be used to increase the "active space" of a trap,
improve short range orientation towards, or arrest weevils near traps. Ethyl acetate
release rates of 400-1600 mg per d may seem high, but may be competitive with levels
produced by large damaged palms (Giblin-Davis et al. 1995a). Ethyl acetate perceived
by male R. palmarum also stimulates pheromone production (Jaffe et al. 1993). If true
for M. hemipterus or other palm weevils, high rather than low relative release of ethyl
acetate from traps should induce pheromone production by nearby weevils.
Sugarcane is one of the cheapest and best sources ofkairomones to enhance attrac-
tion of palm weevils to pheromone-baited traps. Other tissues such as pineapple, S.
palmetto, or molasses plus water were equally synergistic in trapping trials with R.
cruentatus and M. hemipterus (Giblin-Davis et al. 1994a,b, 1995a). African oil palm
stem cubes, molasses on a sponge, or molasses-impregnated mesocarp were not as ef-
fective as sugarcane for R. palmarum (Oehlschlager et al. 1993b). Increasing quanti-
ties of sugarcane or host tissue generally increase attractiveness of the pheromone-
baited trap (Giblin-Davis et al. 1994a, 1995a, Oehlschlager et al. 1993a,b), but opti-
mal quantities should be based on a cost-benefit analysis.
Endemic R. cruentatus, together with the recently introduced M. hemipterus are
more effective in attacking and colonizing the introduced P. canariensis palm in Flor-
ida than either weevil by itself (Giblin-Davis et al. unpublished observation). Feeding

June, 1996

Behavioral Ecology Symposium '95: Giblin-Davis et al. 161

by M. hemipterus in the petioles causes intense kairomone production, attracting the
crown-dwelling R. cruentatus which kills the tree. Palm weevils may mutually benefit
by "cross-responding" to kairomones of their hosts. Rhynchophorus spp., M. hemi-
pterus, R. obscurus, and P. distortus appear to respond to identical or similar host
kairomones (Chang & Curtis 1972, Giblin-Davis et al. 1994a,b, Gries et al. 1994a,
Hallett et al. 1993b). Dynamis borassi, in contrast, may occupy a more specific "feed-
ing" niche (coconut palm inflorescences), and may respond to host kairomones differ-
ent than those of other palm weevils (Gerber et al. 1990, Giblin-Davis unpublished
observation, Wattanapongsiri 1966).
Palms appear to have defenses that must be overcome by weevils or weevil-associ-
ated organisms such as the red ring nematode, Bursaphelenchus cocophilus. Stressed,
unlike healthy S. palmetto palms, become infested with larval progeny when bagged
with R. cruentatus adults (Giblin-Davis & Howard 1989). Even healthy P. canariensis
palm, in contrast, is a susceptible and suitable host for R. cruentatus (Giblin-Davis et
al. 1995b), suggesting that there are differential physical and/or chemical defenses
conveying host suitability and susceptibility (Giblin-Davis et al. unpublished data,
Schuiling & Dinther 1981).
Association of R. palmarum (Gerber & Giblin-Davis 1990, Giblin-Davis 1993), D.
borassi (Gerber et al. 1990), and M. hemipterus (Toquica 1993) in the Neotropics with
the red ring nematode may affect the population dynamics of the weevils. Deposition
of nematodes into wounds of healthy palms can cause a lethal wilt (red ring disease;
RRD) within 2-4 months (Giblin-Davis 1993). Palms dying from RRD, in turn, produce
kairomones that attract palm weevils. Thus, red ring nematode is a lethal agent that
increases the potential number of hosts for palm weevils. Their cross-attraction to host
kairomones and aggregation pheromones increases the probability of associating with
and vectoring the dispersal stage of the nematode. Rhynchophorus spp. and M. hemi-
pterus are also associated with, or even rely on, other microorganisms in dead or dying
hosts (Giblin-Davis et al. 1989, Griffith 1987). Diets lacking debittered yeast, e.g., are
not suitable for rearing R. cruentatus (Weissling & Giblin-Davis 1995). In addition,
various potential fungal and bacterial plant and weevil pathogens may contaminate
and affect behavior and population dynamics of palm weevils (Griffith 1987).


The abundance of adult palm weevils is affected by seasonal changes. R. pal-
marum populations appear to peak at the end of the rainy season and throughout
most of the dry season in coconut plantations in Trinidad (Hagley 1963), and in the
dry season in oil palm plantations in Brazil (Schuiling & Dinther 1981), Costa Rica
(Morales & Chinchilla 1990), and Honduras (Chinchilla et al. 1990). In Florida, R.
cruentatus (Weissling et al. 1994a) and M. hemipterus (Pefia et al. 1995) are more
abundant in spring, before the onset of the rainy season. Rhynchophorus palmarum
(Rochat 1987) and M. hemipterus (Giblin-Davis et al. unpublished observation) have
crepuscular flight patterns. Flight of R. cruentatus in the laboratory was not corre-
lated with time of day and feeding status but increased with increasing temperature
and decreasing relative humidity (Weissling et al. 1994a).
Rhynchophorus spp. seek harborage in leaf axils of healthy palms (Weissling &
Giblin-Davis 1993) and moist fermenting garbage (Chittenden 1902). Cryptic behav-
ior ofR. cruentatus may help to conserve water because this weevil has high cuticular
permeability which causes desiccation in dry environments (Weissling & Giblin-Davis
1993). In a bioassay with choices between low and high relative humidity environ-
ments, R. cruentatus preferred high relative humidity, suggesting that this weevil
possesses hygroreceptors to locate moist harborage sites (Weissling & Giblin-Davis

Florida Entomologist 79(2)

1993). High relative humidity in semiochemical-baited traps can be provided by use
of soapy water or by pesticide-treated sugarcane or palm, or wet sponges or towels.
Different trap designs have been tested to optimize capture of palm weevils (Gib-
lin-Davis et al. 1994a,b, 1995a, Oehlschlager et al. 1992, 1993a,b, Weissling et al.
1992, 1993). Large bucket traps with good surface area placed on the ground or at-
tached to palm trunks are available (Oehlschlager et al. 1993b). Captured weevils are
killed with pesticide-treated (i.e., carbaryl, carbofuran, lannate) sugarcane (Oe-
hlschlager et al. 1993a) or with soapy water in the bottom of traps (Weissling et al.
1994b). Rhynchophorus species usually fly into the vicinity of a trap to land on leaves,
the tree trunk, or ground and then walk into the trap (Oehlschlager et al. 1993b). The
proportionally smaller M. hemipterus is agile in flight and, unlike large species of
Rhynchophorus, may not require a large landing surface or trap placement on the
ground (Giblin-Davis et al. 1995a). A lethal pitfall trap (Giblin-Davis et al. 1994a) ap-
pears to work best for P. distortus (Chinchilla et al. unpublished data), suggesting
that these weevils may occupy pruned or fallen leaves or petioles at the base of the
stem at or below the soil surface.
In trap height studies with R. palmarum (Oehlschlager et al. 1993a,b) and M. hemi-
pterus (Giblin-Davis et al. 1995a), traps associated with possible landing areas were su-
perior. Traps on the ground, e.g., captured moreR.palmarum than those pole-suspended
1.7 or 3.3 m above ground. In contrast, traps attached to palm trunks at 0, 1.7, and 3.3
m heights were equally effective (Oehlschlager et al. 1993a,b). UnlikeR. palmarum, M.
hemipterus was captured equally well in ground traps and pole-suspended traps (1 m)
(Giblin-Davis et al. 1995a). Open "ground" traps can be problematic in areas where rac-
coons (Procyon sp.) occur and eat captured weevils (Giblin-Davis et al. 1995a).
Trap silhouette and color as potential visual cues for foraging palm weevils have
not yet been intensively studied. In field tests using sugarcane and synthetic phero-
mone as trap bait, all trap colors tested equally attracted M. hemipterus to ground-
mounted bucket traps (Giblin-Davis et al. 1995a) and R. palmarum to tree-mounted
bucket traps (Oehlschlager et al. 1993b). These data suggest that color in the human-
perceived spectrum is not a critical parameter for an optimal trap design.


Semiochemical-based mass trapping of beetles may offer a viable pest manage-
ment alternative (Borden 1993, Hardee 1982). Rhynchophorus weevils are suitable
for mass trapping because of their long life cycle and adult longevity, low fecundity,
and reliance on aggregation pheromones and host kairomones.
The viability of semiochemical-based mass trapping for management of palm wee-
vils has been evaluated in a commercial African oil palm (Elaeis guineensis) planta-
tion in Costa Rica infested with R. palmarum and red ring nematode (Chinchilla et al.
1993, Oehlschlager et al. 1995a). Continuous mass trapping over 17 months at densi-
ties of about 6 traps per ha, together with good phytosanitation practices, reduced R.
palmarum trap counts over time and lowered RRD incidence (Chinchilla et al. 1993,
Oehlschlager et al. 1995a). At mass trapping onset, most R. palmarum were captured
in "border" traps of the test site, suggesting removal of potential immigrants into the
study area (Oehlschlager et al. 1995a). A combination of perimeter and "internal"
traps is most effective for mass trapping (Chinchilla et al. 1993). More than 62,500
weevils (about 94 weevils per ha per month) were captured during the study, with
RRD incidence decreasing by a factor of> 2 (Oehlschlager et al. 1995a).
Success of semiochemical-based R. palmarum management in Costa Rica is based
on: 1) autocratic control over a large R. palmarum-infested area (> 100 ha), 2) mono-
culture of African oil palm that is not as suitable and susceptible for R. palmarum and

June, 1996

Behavioral Ecology Symposium '95: Giblin-Davis et al. 163

red ring nematode as is coconut palm, and 3) implementation of mass trapping in con-
cert with intense phytosanitation (aggressive removal of any palm with early symp-
toms of RRD, little leaf, or palm weevil damage). The same criteria are applicable for
R. ferrugineus in date palm, P. dactylifera, in the United Arab Emirates (Hallett et al.
1993b), R. cruentatus in P. canariensis in Florida (Giblin-Davis et al. 1994a), and R.
phoenicis in African oil palm in Africa (Gries et al. 1993), except that RRD does not oc-
cur in these locations.
In the Neotropics, small (< 5 ha) coconut palm plantings may be less suitable for mass
trapping of R. palmarum and reduction of RRD incidence. Coconut is more suitable and
susceptible to R. palmarum and red ring nematode than is African oil palm. Moreover,
small farm holders may not be able to purchase the pheromones for mass trapping and
are not easily organized to consistently rogue trees at the onset of RRD symptoms.
More research is needed to demonstrate the potential of semiochemical-based mass
trapping for other palm weevils. Semiochemical-based monitoring can already be im-
plemented to track population dynamics of palm weevils and to facilitate pest manage-
ment decisions (Giblin-Davis et al. 1995B). In ports of entry, pheromone-baited survey
traps may help to detect and possibly intercept introduction of foreign weevils.


We thank F. W. Howard and R. H. Scheffrahn for critically reviewing the manu-
script. The research was supported in part by the U.S. Department of Agriculture Spe-
cial Grant in Tropical and Subtropical Agriculture, CRSR-95-34135-1763, to R. M. G-
D. and J. E. P. This manuscript is Florida Agricultural Experiment Station Journal
Series R-04941.


ARN, H., E. STADLER, AND S. RAUSCHER 1975. The electroantennographic detector a
selective and sensitive tool in the gas chromatographic analysis of insect pher-
omones. Z. Naturforsch. 30: 722-725.
BARE, C. 0. 1929. Rhynchophorus cruentatus Fab., the palmetto weevil, attracted to
automobile paint. J. Econ. Entomol. 22: 986.
Pheromone combination lures for Carpophilus (Coleoptera: Nitidulidae) spe-
cies. J. Econ. Entomol. 88: 864-869.
BORDEN, J. H. 1993. Strategies and tactics for the use of semiochemicals against for-
est pests in North America, in R. D. Lumsden and J. L. Vaughn [eds.], Pest
management: biologically based technologies. American Chem. Soc. Conf. Proc.
Ser. American Chem. Soc., Washington, D. C.
BUNDENBERG, W. J., I. O. NDIEGE, AND F. W. KARAGO. 1993. Evidence for volatile
male-produced pheromone in banana weevil Cosmopolites sordidus. J. Chem.
Ecol. 19: 1905-1916.
CHANG, V. C. S., AND G. A. CURTIS. 1972. Pheromone production by the New Guinea
sugarcane weevil. Environ. Entomol. 1: 476-481.
CHANG, V. C. S., G. A. CURTIS, AND A. K. OTA. 1971. Insects. Hawaiian Sugar Planters'
Association Annual Report. Pp. 43-44.
CHINCHILLA, C. M., R. MENJIVAR, AND E. ARIAS. 1990. Picudo de la palma y enferme-
dad del anillo rojo/hoja pequefia en una plantaci6n commercial en Honduras. Tu-
rrialba 40: 471-477.
mone-based traps for the management of the American palm weevil, Rhyn-
chophorus palmarum in a commercial oil palm plantation, in Proceedings,
International Oil Palm Congress, 20-25 September 1993. Palm Oil Research In-
stitute of Malaysia, Kuala Lumpur, Malaysia.

Florida Entomologist 79(2)

CHITTENDEN, F. H. 1902. The palm and palmetto weevils. U.S. Dept. Agr. Entomol.
Bull. 38: 23-28.
DIEGADO, H. V., AND F. 0. MORENO. 1986. Evaluaci6n de atrayentes vegetables y un
sistema de trampa para la capture de adults de "gualpa" (Rhynchophorus
palmarum) insecto-plaga de palma Africana y cocotero. Inst. Nac. Invest.
Agropecu. Bol. Tec. 63, Estaci6n Experimental Santo Domingo, Ecuador.
DUSENBERY, D. B. 1992. Sensory Ecology. W. H. Freeman and Co., New York, New
York. 558 pp.
GERBER, K., AND R. M. GIBLIN-DAVIS. 1990. Association of the red ring nematode,
Rhadinaphelenchus cocophilus, and other nematode species with
Rhynchophorus palmarum (Coleoptera: Curculionidae). J. Nematol. 22: 143-
GERBER, K., R. M. GIBLIN-DAVIS, AND J. ESCOBAR-GOYES. 1990. Association of the red
ring nematode, Rhadinaphelenchus cocophilus, with weevils from Ecuador and
Trinidad. Nematropica 20: 39-49.
GIBLIN-DAVIS, R. M. 1993. Interactions of nematodes with insects, pp. 302-344 in W.
Khan, [ed.] Nematode Interactions. Chapman & Hall, London.
GIBLIN-DAVIS, R. M., K. GERBER, AND R. GRIFFITH. 1989. Laboratory rearing of Rhyn-
chophorus cruentatus and R. palmarum (Coleoptera: Curculionidae). Florida
Entomol. 72: 480-488.
GIBLIN-DAVIS, R. M., AND F. W. HOWARD. 1989. Vulnerability of stressed palms to at-
tack by Rhynchophorus cruentatus (Coleoptera: Curculionidae). J. Econ. Ento-
mol. 82: 1185-1190.
GIBLIN-DAVIS, R. M., J. E. PENA, AND R. E. DUNCAN. 1994a. Lethal trap for evaluation
of semiochemical mediated attraction of Metamasius hemipterus sericeus (Oliv-
ier) (Coleoptera: Curculionidae). Florida Entomol. 77: 247-255.
1994b. Field response of Rhynchophorus cruentatus (Coleoptera: Curculion-
idae) to its aggregation pheromone and fermenting plant volatiles. Florida En-
tomol. 77: 164-177.
timization of semiochemical-based trapping of Metamasius hemipterus sericeus
(Olivier) (Coleoptera: Curculionidae). J. Chem. Ecol. 21:(in press).
GIBLIN-DAVIS, R. M., J. E. PENA, AND R. E. DUNCAN. 1995b. Evaluation of an ento-
mopathogenic nematode and chemical insecticides for control of Metamasius
hemipterus sericeus (Coleoptera: Curculionidae). J. Entomol. Sci. 31:(in press).
PIERCE, M. KOUDA, M. ZEBEYOU, AND N. NANOU. 1993. Aggregation pheromone
of the African palm weevil, Rhynchophorus phoenicis F. Naturwissenschaften
80: 90-91.
onate: synergistic kairomone for African palm weevil, Rhynchophorus
phoenicis L. (Coleoptera: Curculionidae). J. Chem. Ecol. 20: 889-897.
PIERCE, JR., M. ZEBEYOU, AND B. KOUAME. 1994b. Aggregation pheromone of
the African rhinoceros beetle, Oryctes monoceros (Olivier) (Coleoptera: Scara-
baeidae). Z. Naturforsch. 49: 363-366.
GRIFFITH, R. 1987. Red ring disease of coconut palm. Plant Dis. 71: 193-196.
GUNATILAKE, R., AND N. E. GUNAWARDENA. 1986. Ethyl alcohol: a major attractant of
red weevil Rhynchophorus ferrugineus. Proc., Sri Lanka Assoc. Adv. Sci. 42nd
Ann. Sess. p. 70.
HAGLEY, E. C. 1963. The role of the palm i h'. .... "*.*1 ." rus palmarum, as a vec-
tor of red ring disease of coconuts. I. Results of preliminary investigations. J.
Econ. Entomol. 56: 375-380.
HAGLEY, E. C. 1965. Tests of attractants for the palm weevil. J. Econ. Entomol. 58:

June, 1996

Behavioral Ecology Symposium '95: Giblin-Davis et al. 165

HALFPAPP, K. H., AND R. I. STOREY. 1991. Cane weevil borer, Rhabdoscelus obscurus
(Coleoptera: Curculionidae), a pest of palms in Northern Queensland, Austra-
lia. Principes 35: 199-207.
gation pheromone of two Asian palm .. i h'i ... 1 -.. t""rus ferrugineus and
R. vulneratus.Naturwissenschaften 80: 328-331.
M. S. GASOUMA, AND J. H. BORDEN. 1993b. Field testing of aggregation phero-
mones of two Asian palm weevils. PORIM International Oil Palm Congress,
Kuala Lumpur, Malaysia, September. 661-668.
pheromone of coconut rhinoceros beetle, Oryctes rhinoceros (L.) (Coleoptera:
Scarabaeidae). J. Chem. Ecol. 21: 1549-1570.
HARDEE, D. D. 1982. Mass trapping and trap cropping of the boll weevil,Anthonomus
grandis Boheman, in A. F. Kydonieus and M. Beroza [eds], Insect suppression
with controlled released pheromone systems, V. 2, CRC Press, Inc. Boca Raton,
GUERRA, R. MARTINEZ, AND B. MIRAS. 1993. Chemical ecology of the palm wee-
vil Rhynchophorus palmarum (L.) (Coleoptera: Curculionidae): attraction to
host plants and to a male-produced aggregation pheromone. J. Chem. Ecol. 19:
MORALES, J. L., AND C. M. CHINCHILLA. 1990. Picudo de la palma y enfermedad del
anillo rojo/hoja pequefia en una plantaci6n commercial en Costa Rica. Turrialba
40: 478-485.
MORI, K., K. HIROMASA, AND D. ROCHAT. 1993a. Synthesis of the stereoisomers of 3-
methyl-4-octanol to determine the absolute configuration of the naturally oc-
curring (3S,4S)-isomer isolated as the male-produced aggregation pheromone
of the African palm weevil, Rhynchophorus phoenicis. Liebigs Ann. Chem.
1993: 865-870.
MORI, K., H. KIYOTA, C. MALOSSE, AND D. ROCHAT. 1993b. Synthesis of the enanti-
omers of syn-4-methyl-5-nonanol to determine the absolute configuration of the
naturally occurring (4S,5S)-isomer isolated as the male-produced pheromone
compound of Rhynchophorus vulneratus and Metamasius hemipterus. Liebigs
Ann. Chem. 1993: 1201-1204.
do comportamento olfativo de Rhynchophorus palmarum (L.) (Coleoptera: Cur-
culionidae) no campo. An. Soc. Entomol. Brasil 18: 267-274.
NAGNAN, P., A. H. CAIN, AND D. ROCHAT. 1992. Extraction and identification of vola-
tile compounds of fermented oil palm sap (palm wine), candidate attractants
for the palm weevil. Ol1agineaux 47: 135-142.
NAPOMPETH, B., T. NISHIDA, AND W. C. MITCHELL. 1972. Biology and rearing methods
of the New Guinea sugarcane weevil, Rhabdoscelus obscurus. Hawaii Agricul-
tural Experiment Station College of Tropical Agriculture. University of Hawaii
Technical Bulletin No. 85.
O'BRIEN, C. W., AND R. E. WOODRUFF. 1986. First records in the United States and
South America of the African oil palm weevils, Elaeidobius subvittatus (Faust)
and E. kamerunicus (Faust) (Coleoptera: Curculionidae). Florida Dept. Agric.
and Consumer Serv. Entomology circular No. 284.
Chemical communication in cucujid grain beetles. J. Chem. Ecol. 14: 2069-
CHINCHILLA, AND R. G. MEXZON. 1992. Chirality and field activity of Rhyn-

Florida Entomologist 79(2)

chophorol, the aggregation pheromone of the American palm weevil. Naturwis-
senschaften 79: 134-135.
of a pheromone-based trap for the American palm weevil, Rhynchophorus pal-
marum, in Proceedings, International Oil Palm Congress, 20-25 September
1993. Palm Oil Research Institute of Malaysia, Kuala Lumpur, Malaysia.
ZON, AND B. MORGAN. 1993b. Development of a pheromone-based trapping sys-
tem for Rhynchophorus palmarum (Coleoptera: Curculionidae). J. Econ.
Entomol. 86: 1381-1392.
Influence of a pheromone-based trapping system on the distribution of Rhyn-
chophorus palmarum (Coleoptera: Curculionidae) in oil palm. Environ. Ento-
mol. 24: 1005-1012.
JR., AND S. LAUP. 1995b. Structure, chirality, and field testing of a male-pro-
duced aggregation pheromone of Asian palm weevil Rhynchophorus bilineatus
(Montr.) (Coleoptera: Curculionidae). J. Chem. Ecol. 21: 1619-1629.
PENA, J. E., R. M. GIBLIN-DAVIS, AND R. E. DUNCAN. 1995. Impact of indigenous Beau-
veria bassiana (Balsamo) Vuillemin on banana weevil and rotten sugarcane
weevil (Coleoptera: Curculionidae) populations in banana in Florida. J. Agric.
Entomol. 12: 163-167.
Pheromone chirality of the African palm I i Ri ..-... ,.,.t... rus phoenicis (F.)
and the palmetto weevil, Rhynchophorus cruentatus (F.) (Coleoptera: Curcu-
lionidae). J. Chem. Ecol. 20: 2653-2671.
gation pheromones and host kairomones of the West Indian sugarcane weevil,
Metamasius hemipterus sericeus (Oliv.) (Coleoptera: Curculionidae). J. Chem.
Ecol. 21:(in press).
A. C. OEHLSCHLAGER. 1995b. Pheromone chirality of the Asian palm weevils,
Rhynchophorus ferrugineus (Oliv.) and Rhynchophorus vulneratus (Panz.), (Co-
leoptera: Curculionidae). J. Chem. Ecol. 21: (in press).
ROCHAT, D. 1987. Etude de la communication chimique chaz un ColeoptBre Curculio-
nidae: Rhynchophorus palmarum L. Diplome d'Etudes Aprrofondies Report,
University Paris VI.
ROCHAT, D., AND P. ZAGATTI. 1993. Mise en evidence d'une pheromone cuticulaire
chez les femelles de Rhynchophorus palmarum (L.) (Coleoptera: Curculioni-
dae). C. R. Acad. Sci. Paris Ser. III. (in press).
Evidence for male-produced aggregation pheromone in American palm weevil,
Rhynchophorus palmarum (L.) (Coleoptera: Curculionidae). J. Chem. Ecol. 17:
DESCOINS. 1991b. Male-produced aggregation pheromone of the American
palm weevil, Rhynchophorus palmarum (L.) (Coleoptera: Curculionidae): col-
lection, identification, electrophysiological activity, and laboratory bioassay. J.
Chem. Ecol. 17: 2127-2140.
MARIAU. 1993a. Ecologie chimique des charanqons des palmiers, Rhynchopho-
rus spp. (Coleoptera). Oleagineux 48: 225-236.
Identification of new pheromone-related compounds from volatiles produced by

June, 1996

Behavioral Ecology Symposium '95: Giblin-Davis et al. 167

males of four Rhynchophorinae weevils (Coleoptera, Curculionidae). C. R.
Acad. Sci. Paris Ser. II. 316: 1737-1742.
ROCHAT, D., F. AKAMOU, A. SANGARE, D. MARIAU, AND K. MORI. 1995. Field trapping
of Rhynchophorus phoenicis with stereoisomers of the synthetic aggregation
pheromone. C. R. Acad. Sci. Paris Ser. III. 318: 183-190.
SAMARAJEEWA, U., M. R. ADAMS, AND J. M. ROBINSON. 1981. Major volatiles in Sri
Lanka arrack, a palm wine distillate. J. Food Technol. 16: 437-444.
SCHUILING, M., AND J. B. M. VAN DINTHER. 1981.'Red ring disease' in the Paricatuba
oilpalm estate, Para, Brazil. Zeit. angew. Entomol. 91: 154-161.
SEYBOLD, S. J. 1993. Role of chirality in olfactory-directed behavior: aggregation of
pine engraver beetles in the genus Ips (Coleoptera: Scolytidae). J. Chem. Ecol.
19: 1809-1831.
TOQUICA, L. S. M. 1993. Insectos diseminadores de Rhadinaphelenchus cocophilus
(Cobb) Goodey (Tylenchida: Aphelenchoididae) agent causal del anillo rojo-
hoja corta en palma de aceite en San Carlos de Guaroa-meta. M. Sc. disserta-
tion. Universidad Nacional de Colombia. 68 pp.
UHL, N. W., AND J. DRANSFIELD. 1987. Genera Palmarum. Allen Press, Inc.,
Lawrence, Kansas. 610 pp.
VAURIE, P. 1966. A revision of the Neotropical genus Metamasius (Coleoptera, Curcu-
lionidae, Rhynchophorinae). Species groups I and II. Bull. American Mus. Nat.
Hist.131: 213-337.
VAURIE, P. 1970. Weevils of the tribe Sipalini (Coleoptera, Curculionidae, Rhyn-
chophorinae) Part 1. The genera Rhinostomus and Yuccaborus. Am. Mus. Novi-
tates No. 2419: 1-57.
VAURIE, P. 1973. The weevil genera Homalinotus and Ozopherus of the neotropical
cholinae (Coleoptera: Curculionidae). Bull. American Mus. Nat. Hist. 152: 1-50.
Male-produced aggregation pheromone of the maize weevil, Sitophilus zea-
mais, and interspecific attraction between three Sitophilus species. J. Chem.
Ecol. 9: 831-841.
SOR, AND K. MORI. 1987. Determination of chirality in 5-hydroxy-4-methyl-3-
heptanone, the aggregation pheromone of Sitophilus oryzae (L.) and S. zeamais
Motschulsky. J. Chem. Ecol. 13: 2159-2169.
WATTANAPONGSIRI, A. 1966. A revision of the genera Rhynchophorus and Dynamis
(Coleoptera: Curculionidae). Dep. Agr. Sci. Bull., Bangkok 1: 1-328.
WEISSLING, T. J., AND R. M. GIBLIN-DAVIS. 1993. Water loss dynamics and humidity
preference of Rhynchophorus cruentatus (Coleoptera: Curculionidae) adults.
Environ. Entomol. 22: 94-98.
WEISSLING T. J., AND R. M. GIBLIN-DAVIS. 1994. Fecundity and fertility of Rhyn-
chophorus cruentatus (Coleoptera: Curculionidae). Florida Entomol. 77: 75-78.
WEISSLING, T. J., AND R. M. GIBLIN-DAVIS. 1995. Oligidic diets for culture of the pal-
metto weevil, Rhynchophorus cruentatus. Florida Entomol. 78: 225-234.
DOZA. 1992. Trap for capturing and retaining Rhynchophorus cruentatus (Co-
leoptera: Curculionidae) adults using Sabal palmetto as bait. Florida Entomol.
75: 212-221.
WEISSLING, T. J., R. M. GIBLIN-DAVIS, AND R. H. SCHEFFRAHN. 1993. Laboratory and
field evidence for male-produced aggregation pheromone in Rhynchophorus
cruentatus (F.) (Coleoptera: Curculionidae). J. Chem. Ecol. 19: 1195-1203.
behavior and seasonal trapping ofRhynchophorus cruentatus (Coleoptera: Cur-
culionidae). Ann. Entomol. Soc. America 87: 641-647.
JR., AND A. C. OEHLSCHLAGER 1994b. An aggregation pheromone of the pal-
metto weevil, Rhynchophorus cruentatus (F.) (Coleoptera: Curculionidae). J.
Chem. Ecol. 20: 505-515.

Florida Entomologist 79(2)


Subtropical Agricultural Research Laboratory,
2301 S. International Blvd.,
Weslaco, TX 78596


Tephritid fruit fly larval mortality due to heat has been investigated by immersing
immature fruit flies into heated water. It was postulated that this information could
be used to model the relationship between heat and fruit fly kill in fruits subjected to
heat quarantine treatments, such as hot water immersion, vapor heat, and forced hot
air. Third instar Caribbean fruit flies, Anastrepha suspense (Loew), were reared on a
semi-artificial diet or grapefruits and immersed in: a) tap water, b) grapefruit juice, or
c) tapwater with the same pH as grapefruit juice (using citric acid), all at 43.0+0.05 C.
LT,, estimates were 20-31% lower for puparia and adults of third instars immersed in
grapefruit juice than those immersed in heated water, with or without acid. Mortality
of third instars immersed in either heated grapefruit juice or heated water was not af-
fected by either diet.

Key Words: Quarantine treatment, hot-water immersion, modeling, Anastrepha sus-


Ha sido investigada la mortalidad de larvas de moscas tefritidas de la fruta me-
diante su inmersi6n en agua caliente. Se ha postulado que esta informaci6n podria ser
usada para modelar la relaci6n entire el calor y la muerte de las larvas en frutas so-
metidas a tratamientos cuarentenarios de calor tales como inmersi6n en agua ca-
liente, calentamiento con vapor, y aire caliente forzado. Moscas fruteras del Caribe,
Anastrepha suspense (Loew), del tercer estadio fueron criadas en una dieta semiarti-
ficial o en toronjas e inmersas en: a) agua, b) jugo de toronja, o c) agua con el mismo
pH que el jugo de toronja (usando acido citrico), todos a 43.00.05 C. Los tiempos le-
tales medios estimados fueron 20-30% mas bajos para los puparios y adults obteni-
dos de los terceros estadios sumergidos enjugo de toronja que en los inmersos en agua
caliente, con o sin acido. La mortalidad de los terceros estadios sumergidos enjugo de
toronja o agua calente no fue afectada por ninguna dieta.

More research is conducted on heat quarantine treatments (hot water immersion,
vapor heat, and forced hot air) by the U.S. Department of Agriculture, Agricultural
Research Service, than any other type of disinfestation treatment for tephritid fruit
flies (Anonymous 1992). Hot water immersion is used to disinfest mangoes shipped to
the United States (U.S.) of various fruit fly species (Sharp 1994). Forced hot air is used
as a fruit fly quarantine treatment for papayas grown in Hawaii and shipped to the
continental U.S., and heated air treatments are preferred quarantine procedures by

June, 1996

Hallman: Fruit Fly Mortality in Water vs. Fruit Juice

Japan (Hallman & Armstrong 1994). Research on the mortality rate of fruit fly third
instars by immersion in hot tap water has been carried out with the goal of using this
information to model quarantine treatments of fruits infested with tephritids (Jang
1986, Sharp & Chew 1987, Rodriguez et al. 1989). However, the mortality response of
insects immersed in hot water must be very similar to their response inside fruits for
the data to be valid for modeling quarantine treatments. There are many differences
between immersion of fruit fly third instars in hot water and heating of third instars
inside of fruits exposed to a heat disinfestation treatment. For example, larvae used
in mortality studies are usually from a colony which may be genetically different from
feral flies (Chambers 1977). Fiducial limits (95%) for probit 9 estimates of time neces-
sary to kill feral versus colony-reared West Indian fruit fly,Anastrepha obliqua (Mac-
quart), in mangoes (mean weight = 625 g) immersed in 46.1C water overlapped little,
indicating possible differences in heat susceptibility (Sharp 1988). Larvae reared to
third instar and removed from diet undergo a behavioral change as they cease feeding
and seek a pupation site. Concurrently, their mortality response to heat may change.
Larvae heated inside of fruits may not yet be physiologically ready to emerge from the
fruit. Host may affect heat-induced mortality; larvae used in hot water immersion
studies were reared on semi-artificial diets. Larvae immersed in a liquid medium are
exposed to the turbulence and raised temperature immediately, while larvae in a fruit
are gradually exposed to increased temperature.
Specific rearing conditions could influence mortality. Hallman (1994) found a pos-
itive relationship between rearing temperature and resistance of third instar Carib-
bean fruit fly,Anastrepha suspense (Loew), to hot water immersion. Fruit flies used in
laboratory experiments are usually reared at a high population density, whereas pop-
ulation densities in the field are relatively low.
Larvae heated inside of fruits are exposed to fruit pulp substances, some created
because of the fruit fly infestation and waste products of the developing larvae in a
semi-solid matrix, while those heated in water are in a simple liquid medium devoid
of other substances. Post-treatment handling could conceivably affect survival of lar-
vae. Larvae within a fruit would still have to complete development and emerge from
the fruit, while those dipped in hot water are taken out and placed in a pupation me-
dium, thus likely enhancing the latter's probability of survival. Therefore, there are
many opportunities for the mortality response of third instars in each system to vary.
The significance of each of these differences must be measured and appropriate alter-
ations in the model be made before models can be used to accurately predict fruit fly
mortality in fruits subjected to heat disinfestation treatments.
The objective of this research was to compare the heat mortality response of third
instar Caribbean fruit fly reared on a semi-artificial diet or grapefruits and immersed
in tap water, grapefruit juice, or tap water with the same pH as grapefruit juice at 43C.


The Caribbean fruit flies used in this research were from a colony reared on a semi-
artificial diet since 1971 at the USDA Subtropical Horticulture Research Station in
Miami, Florida (Hennessey 1994). Third instars were placed in a stainless-steel wire
mesh strainer and rinsed with water to remove the diet. Third instars reared on grape-
fruits were obtained by exposing 'Marsh' white grapefruits to oviposition in a fly cage
for 4 days and then placing the grapefruits on racks where third instars would emerge
and drop into a bin containing sand beneath the rack (Hallman 1994). Grapefruits
were cut open once larvae began emerging, and larvae were removed directly from
grapefruit pulp. Depending on the number of larvae available on any day, 25-200 lar-

Florida Entomologist 79(2)

vae were placed in round stainless steel-mesh (0.5 mm between threads) tea infusers
(40-mm diam). A set of five infusers each was immersed in: a) tap water (pH 7.1-9.3),
b) grapefruit juice, or c) tapwater with the same pH (3.2-3.7) as grapefruit juice (using
citric acid), in electrically-heated circulating water baths (11-liter capacity, Gaffney
Engineering Co., Gainesville, FL) at 43.0-0.05 C. One infuser was removed after 6,
12, 18, 24, and 30 min. Immediately upon removal from the heated liquids, each in-
fuser was cooled in a separate container of the same liquid at 241C for 1 min. A sixth
infuser with larvae was immersed in tap water, grapefruit juice, or acidic water at
24+1C for 30 min as a control. The larvae were removed from the infusers and placed
in 70 ml of moist No. 4 vermiculite in 0.5-liter plastic containers with screen lids.
The numbers of larvae that formed puparia and emerged as adults with functional
wings and legs were counted. There were ten replicates. Numbers of puparia and
adults were corrected for natural mortality (Abbott 1925) and subjected to probit
analysis (LeOra Software 1987). Additionally, percentage adult mortality at 24 min
immersion in 43.00.05 C was subjected to analysis of variance of mortality among
the treatments. Percentage adult emergence from the controls was examined with
analysis of variance to determine if the immersion medium itself (without heat)
caused differential mortality (SAS Institute 1988). In both of these analyses of vari-
ance, diet and heating medium were considered fixed, independent factors. Percent-
ages were transformed by the arcsine transformation prior to analysis of variance.
Differences between means were tested for statistical significance (95% confidence
level) with the Ryan-Einot-Gabriel-Welsh multiple range test.

Probit analysis of development of Caribbean fruit fly puparia and adults heated as
third instars is presented in Table 1. Fiducial limits (95% confidence level) for the im-
mersion times estimated to kill 95% of the population (LT,,) did not overlap for numbers
of abnormal-looking puparia reared on agar-based diet or grapefruits and heated as
third instars in grapefruit juice versus water or water with the same pH as grapefruit
juice. The LT,, for puparia immersed as third instars in grapefruit juice was lower than
that for puparia immersed as third instars in water or water with the same pH as
grapefruit juice. For adults, fiducial limits (95% confidence level) for LT,, did not over-
lap for flies reared in grapefruits, although numerically the LT,, for adults immersed as
third instars in grapefruit juice was lower than that for puparia immersed as third in-
stars in water or water with the same pH as grapefruit juice in both rearing media.
Highly significant differences occurred among heating media for adult mortality
after 24 min at 43.00.05 C (F probability = 0.0006; df = 2, 33). MeanSEM mortali-
ties for Caribbean fruit fly adults were 98.1+0.67, 83.93.9, and 82.34.9 for third in-
stars immersed in grapefruit juice, water, and water with the same pH as grapefruit
juice, respectively, with significant differences only between the first versus the latter
two immersion media. There were no significant differences between rearing diets (F
probability = 0.93; df = 1, 33) nor any significant diet by immersion medium interac-
tion (F probability = 0.84; df = 2, 33).
There were no significant differences in percentage adult emergence among the
controls for diet (F probability = 0.85; df = 1, 51), immersion medium (F probability =
0.22; df = 2, 51), or diet by medium interaction (F probability = 0.35; df = 2, 51).

Heating Caribbean fruit fly third instars in grapefruit juice resulted in higher levels
of mortality compared with the more typical practice of heating the insects in water.
This increased mortality was not due to the acidity of the juice because larvae heated

June, 1996

Hallman: Fruit Fly Mortality in Water vs. Fruit Juice


Larval Diet Immersion Medium SlopeSEM LT,, (95% FL)

Abnormal-looking puparia
Semi-artificial Water 4. 10.85 43.2 (37.2-52.9)
Semi-artificial Grapefruit juice 4.90.90 30.0 (26.7-35.1)
Semi-artificial Acidic water 3.80.10 43.7 (35.1-61.7)
Grapefruit Water 4.30.18 40.3 (34.4-50.3)
Grapefruit Grapefruit juice 4.80.18 28.1 (25.2-32.1)
Grapefruit Acidic water 5.00.21 37.4 (35.0-61.7)

Adult mortality
Semi-artificial Water 3.60.75 34.3 (28.5-44.9)
Semi-artificial Grapefruit juice 3.90.77 26.9 (23.2-33.2)
Semi-artificial Acidic water 3.40.93 33.6 (25.8-54.6)
Grapefruit Water 4.30.17 33.0 (28.3-40.8)
Grapefruit Grapefruit juice 4.80.18 23.9 (21.5-27.5)
Grapefruit Acidic water 5.10.20 30.1 (27.7-33.1)

in water with the same pH as grapefruit juice, using citric acid, showed the same rate
of mortality as larvae heated in water alone. Quarantine treatments affect larvae in
fruits, hence, it might be expected that heating larvae in fruit juice would give a more
accurate simulation of mortality inside heated fruits than heating larvae in water. Nev-
ertheless, because of the additional and abundant dissimilarities in response between
in vivo (fruit) and in vitro heat treatment of fruit fly immatures, some of which were
enumerated earlier, in vitro studies may not be useful in achieving precise estimates of
heat dose required for quarantine treatments. In vitro studies are useful, however, in
determining relative differences among various factors affecting insect mortality.
There were no significant differences in mortality of third instars reared on grape-
fruit versus diet and immersed in 43.00.05 C media indicating that diet was an ad-
equate substitute for grapefruit in this case.
In two of 81 controls it was noted that adult emergence exceeded formation of vi-
able-looking puparia. Examination of the puparia revealed that normal-looking
adults emerged from puparia that were considered non-viable. Normal-looking adults
were also noted to emerge from abnormal puparia of some of the heated larvae.
Thomas & Mangan (1995) found that normal-looking adults sometimes emerged from
abnormal-looking puparia of the Mexican and West Indian fruit flies, Anastrepha
ludens (Loew) andA. obliqua (Macquart), respectively, following heat treatment and
cautioned against the practice of considering abnormal puparia dead when research-
ing quarantine treatments. This study provides evidence that this same phenomenon
occurs with the Caribbean fruit fly.

I thank Wilhelmina Wasik, USDA-ARS, Miami, FL, for technical assistance and
Daniel A. Wolfenbarger, USDA-ARS, Weslaco, TX, and Lisa G. Neven, USDA-ARS,
Yakima, WA, for helpful reviews.

Florida Entomologist 79(2)

June, 1996


ABBOTT, W. S. 1925. A method of computing the effectiveness of an insecticide. J.
Econ. Entomol. 18: 265-267.
ANONYMOUS. 1992. Quarantine workshop for horticultural commodities: final report.
U.S. Dept. Agr, Agric. Research Service.
CHAMBERS, D. L. 1977. Quality control in mass rearing. Annual Rev. Entomol. 22: 289-
HALLMAN, G. J. 1994. Mortality of third-instar Caribbean fruit fly (Diptera: Tephriti-
dae) reared at three temperatures and exposed to hot water immersion or cold
storage. J. Econ. Entomol. 87: 405-408.
HALLMAN, G. J., AND J. W. ARMSTRONG. 1994. Heated air treatments, pp. 149-163 in
J. L. Sharp and G. J. Hallman [eds.], Quarantine Treatments for Pests of Food
Plants. Westview Press, Boulder, Colorado.
HENNESSEY, M. K. 1994. Depth of pupation of Caribbean fruit fly (Diptera: Tephriti-
dae) in soils in the laboratory. Environ. Entomol. 23: 1119-1123.
JANG, E. B. 1986. Kinetics of thermal death in eggs and first instars of three species
of fruit flies (Diptera: Tephritidae). J. Econ. Entomol. 79: 700-705.
LEORA SOFTWARE. 1987. POLO-PC: a user's guide to Probit Or Logit analysis. LeOra
Software, Berkeley, California.
fruit quarantine heat treatments. Paper no. 89-6053. 1989 Summer Meeting,
American Soc. Agric. Engineers, Quebec.
SAS INSTITUTE. 1988. SAS technical report: P-179, additional SAS/STAT procedures,
release 6.03. SAS Institute, Cary, NC.
SHARP, J. L. 1988. Status of hot water immersion quarantine treatment for tephriti-
dae in mangos. Proc. Florida State Hort. Soc. 101: 195-197.
SHARP, J. L. 1994. Hot water immersion, pp. 133-147 in J. L. Sharp and G. J. Hallman
[eds.], Quarantine Treatments for Pests of Food Plants. Westview Press, Boul-
der, Colorado.
SHARP, J. L., AND V. CHEW. 1987. Time/mortality relationships for Anastrepha sus-
pensa (Diptera: Tephritidae) eggs and larvae submerged in hot water. J. Econ.
Entomol. 80: 646-649.
THOMAS, D. B., AND R. L. MANGAN. 1995. Morbidity of the pupal stage of the Mexican
and West Indian fruit flies (Diptera: Tephritidae) induced by hot-water immer-
sion in the larval stage. Florida Entomol. 78: 235-246.

Reynolds et al.: Rhagoletis pomonella traps


'Department of Entomology, University of Massachusetts, Amherst, MA 01003

'Pest Management Supply, Inc., 311 River Dr., Hadley, MA 01035


Two traps, designed to kill with pesticide instead of a sticky coating, were evalu-
ated for their potential to control apple maggot flies, Rhagoletis pomonella (Walsh).
Both traps were designed to protect feeding stimulant and pesticide from rainfall, a
problem with previous trap designs. The first trap was a perforated hollow red sphere,
the interior of which contained odor attractants, feeding stimulant and pesticide. In
field tests, internally baited spheres were slightly less attractive to R. pomonella than
externally baited spheres, and alighting flies were highly reluctant to enter openings
into the sphere interior. The second trap was a sphere in which odor attractants, feed-
ing stimulant and pesticide were contained in a liquid inside the trap and released
through a sponge at the surface. Few of the flies alighting on these traps were induced
to feed. The post-alighting behavior ofR. pomonella on both trap types tested suggests
that neither type (as tested) holds much promise to replace existing designs.

Key Words: Rhagoletis pomonella, pesticide treated traps, odor attractants, Tangle-


El potential para controlar la mosca de la manzana, Rhagoletis pomonella (Walsh),
fue evaluado en dos tipos de trampa diseiados para matar con pesticide en lugar de
con una cubierta adhesive. Ambos tipos de trampa fueron diseiados para proteger el
fagoestimulante y el pesticide de la lluvia, lo cual habia sido un problema en los dise-
nos de trampas anteriores. La primera trampa fue una esfera hueca roja perforada, el
interior de la trampa contenia olores atrayentes, fagoestimulante y pesticide. En
pruebas de campo, las esferas cebadas interiormente fueron ligeramente menos atrac-
tivas a R. pomonella que las esferas cebadas externamente y las moscas que se posa-
ban en ellas no entraban a trav6s de los agujeros. La segunda trampa fue una esfera
que contenia los olores atrayentes, el fagoestimulante y el pesticide en un liquid den-
tro de la trampa que pasaba a la superficie de esta mediante una esponja. Pocas de las
moscas que se posaban en estas trampas se alimentaban. El comportamiento de R. po-
monella despues de posarse en ambas trampas sugiere que ninguno de los tipos pro-
bados es lo suficientemente prometedor para reemplazar las trampas ya existentes.

The apple maggot, Rhagoletis pomonella (Walsh), is a major pest of apples in east-
ern and central North America. Recently, odor-baited sticky traps have been used as
a substitute for pesticide in controlling apple maggot in several commercial orchards
(MacCollom 1987, Prokopy et al. 1990). To date, the most economically effective trap
has proven to be an 8 cm red sphere coated with Tangletrap adhesive and baited with
synthetic food and/or fruit odor (Duan & Prokopy 1992). One of the impediments to
greater use of such spheres by apple growers is the laborious process of coating the

Florida Entomologist 79(2)

spheres with a sticky adhesive and cleaning them of insects and debris every two
weeks to maintain capturing effectiveness (Duan & Prokopy 1995b).
In concept, pesticide applied to spheres could be an effective substitute for adhe-
sive in killing R. pomonella. Toward this end, Duan & Prokopy (1995a) showed that
spheres coated with a mixture containing dimethoate, sucrose as a feeding stimulant
eliciting fly ingestion of pesticide, and latex paint as a residue-extending agent killed
a large majority of alighting R. pomonella before exposure to rainfall. After rainfall,
however, the spheres were less effective, largely due to loss of feeding stimulant. An
analogous trap for the olive fruit fly, Dacus oleae (Gmelin), consisting of a plywood
rectangle soaked in deltamethrin and sucrose, provided effective control in Greece
(Haniotakis et al. 1991). However, no rain fell during the trapping period due to the
dry climate.
We have envisioned two principal approaches to eliminating the negative effects of
rainfall on the residual activity of pesticide and feeding stimulant: (1) using a protec-
tive cover to prevent rainfall from contacting the spheres, and (2) finding a more ef-
fective residue-extending polymer to combine with, or substitute for, latex paint
(Prokopy et al. 1995). Regarding the former, to date we have found that all tested vari-
ants of protective covers placed above spheres reduce alightment of R. pomonella by
at least 50 percent, an unacceptable level (Duan & Prokopy 1992). A possible alterna-
tive to a protective cover is the placement of pesticide, feeding stimulant, and syn-
thetic food and fruit odor within a hollow, perforated sphere. The wall of the sphere
would serve to protect the interior from rainfall. A similar perforated, cylindrical trap
baited with food odor is being developed for the Mediterranean fruit fly, Ceratitis cap-
itata (Weidmann) (Health & Epsky 1995). However, to our knowledge, spheres of this
type have not yet been evaluated against R. pomonella or any other tephritid flies.
Previously, Reissig (1974, 1975) evaluated a yellow hollow rectangular box with a hole
on each side and food odor and pesticide on the interior as a potential trap for R.
pomonella. Initially, it appeared to be an effective trap in trees harboring hungry
adults, but subsequently it proved ineffective when evaluated under a broader range
of field conditions.
Here, we first evaluated R. pomonella response to internally and externally-baited
red spheres perforated with holes and to internally-baited spheres with varying num-
bers of holes. We then observed post-alighting behavior on internally-baited spheres
with varying numbers of holes. Finally, we evaluated commercially available red
sphere traps designed so that both feeding attractant and pesticide are contained in
a liquid inside the trap and are released through a sponge on the underside of the
sphere, protected from rainfall.


In the first experiment, internally and externally-baited red spheres were evalu-
ated for propensity to capture R. pomonella in a commercial orchard. The spheres (ob-
tained from Pest Management Supply Co., Hadley, MA) consisted of two separate,
hollow halves (10 cm diam), which allowed odor baits to be placed inside the trap.
Odor baits consisted of one unit each of synthetic fruit odor (butyl hexanoate, dis-
pensed from a capped 15 ml polyethylene vial) and synthetic food odor (ammonium
carbonate packet, purchased from R. Heath, Gainesville, FL). Spheres were perfo-
rated with three 2.4 cm holes. Four treatments were set up: (1) internally-baited
spheres with two cardboard strips containing dimethoate (also placed inside) as the
killing agent, (2) internally-baited spheres coated on the exterior with a layer of
Tangletrap (from the Tanglefoot Co., Grand Rapids, MI), (3) externally-baited

June, 1996

Reynolds et al.: Rhagoletis pomonella traps

spheres (odors placed about 10 cm from the traps) with Tangletrap, and (4) externally-
baited, non-perforated spheres with Tangletrap. The test was conducted in an orchard
block of about 30 Gravenstein apple trees. Traps were positioned one per tree, accord-
ing to methods described by Duan & Prokopy (1992). After one week, captured flies
were counted and removed, and the trap types were rotated. Capture data were ana-
lyzed using a two way ANOVA, in which columns consisted of trap type and rows con-
sisted of replicates.
In the second experiment, sticky 0-, 3-, 6-, 12-, and 24-hole spheres were evaluated
for propensity to capture R. pomonella. Holes were 2.4 cm diam except for the 24-hole
spheres, which were 1.4 cm. The odor baits used in this test were the same as in the
first experiment. All traps were coated with Tangletrap and internally-baited (except
for the 0-hole sphere, which was externally-baited). One trap of each type was hung
in a large hawthorn tree known to contain a substantial R. pomonella population.
Once daily, the traps were checked for R. pomonella captures, cleaned, and rotated.
This was done for one complete rotation (5 days). For each day, the total number of fly
captures over all trap types was summed and a percentage of that total was then cal-
culated for each trap type. By using this approach, any day-to-day fluctuations in R.
pomonella population size and activity were negated. Data were analyzed using a two
way ANOVA, in which columns consisted of trap type and rows consisted of test days
(trap position).
In the third experiment, post-alighting behavior of R. pomonella was observed on
internally-baited red spheres with 3, 6, 12, or 24 holes. We wanted to determine fly in-
clination to enter holes to the interior of the trap (where feeding stimulant and pesti-
cide could potentially be located). The same hawthorn tree and traps used in the
second experiment were used in this test. Three traps of each type were hung and
monitored for R. pomonella alightment, flies entering trap holes and time spent on the
sphere. Residence times of R. pomonella on the spheres were analyzed by a one way
In the fourth experiment, an alternative trap type (Fruitect trap, mfd. by RonPal
Ltd., Rishpon, Israel) and red wooden spheres were evaluated forR. pomonella post-
alighting behavior. The Fruitect trap consisted of a red plastic sphere (12.5 cm diam)
in which a feeding attractant (protein hydrolysate) and feeding stimulant (sucrose)
were dispensed from the interior to the exterior via a sponge that formed a 1.0 cm
band on the underside of the sphere. Red wooden spheres (8.0 cm diam) were dipped
in an aqueous solution of 20% sugar prior to testing. The test was conducted in an in-
door field cage by hanging four spheres (Fruitect and wooden spheres were tested sep-
arately) in a potted fig tree. For each trial, 40 female R. pomonella were released and
allowed to forage freely for up to 1 h. Test flies were of wild origin, aged 3-4 weeks, and
were either starved of all protein or continually fed protein since eclosion. Alighting
flies were monitored for total time on the sphere and time spent feeding. Data on res-
idence time, percent feeding, and feeding time were analyzed using two sample t-tests.


In Experiment 1 (Table 1), approximately 30-40% fewer R. pomonella were caught
on 3-hole sticky traps internally-baited than on externally-baited sticky traps with or
without 3 holes. Internally-baited 3-hole traps with pesticide instead of Tangletrap
as the killing agent failed to trap a single fly over the entire experiment.
In Experiment 2 (Table 2), externally-baited traps with no holes captured the
highest number of flies and had the highest daily percentage of fly captures. Daily per-
cent fly captures were about 15-40% less on the internally-baited spheres, although

Florida Entomologist 79(2)

June, 1996


Mean No. Flies
Captured Per
Trap Type Killing Agent Odor Position ReplicateSE1

3-hole Dimethoate Internal 0.00.0c
3-hole Tangletrap Internal 24.35.8b
3-hole Tangletrap External 41.37.6a
0-hole Tangletrap External 36.15.8ab

'Column values with different letters are significantly different according to two way ANOVA and LSD crite-
rion at the 0.05 level.

two way ANOVA showed that differences among all five trap types were not
In Experiment 3 (Table 3), 0, 0, 0 and 16% of alighting flies, respectively, entered
holes in 3-, 6-, 12-, and 24-hole spheres. Flies spent significantly more time on 3- and
24-hole spheres than on 6- and 12-hole spheres.
In Experiment 4 (Table 4), a significantly greater proportion of alighting flies fed
on red wooden spheres than on Fruitect traps. This was true for protein-fed flies (90
vs. 2%) and protein-starved flies (75 vs. 23%). Protein-fed flies on red wooden spheres
fed much longer than flies on Fruitect traps (although the sample size feeding on Frui-
tect traps consisted of only one fly). Protein-starved flies on Fruitect traps and red
wooden spheres showed no significant difference in mean time feeding.

Our findings indicate that the trap designs tested here are not an effective alter-
native to prototype pesticide-coated spheres described by Duan & Prokopy (1995a). To

RIODS (N=5).

Mean Percent of Total
Trap Type Total No. Trap Captures Daily CapturesSE'

0-hole 347 25.94.9
3-hole 285 21.43.3
6-hole 203 15.13.2
12-hole 193 15.31.9
24-hole 286 22.32.8

Differences in percentage captures between trap types were not significant according to two way ANOVA at
the 0.05 level.

Reynolds et al.: Rhagoletis pomonella traps


Mean Time Per Fly % Alighting Flies
Spent on Trap Entering Trap
Trap Type No. Flies Alighting (s)SE' HolesSE

3-hole 23 163.271.4a 0.00.0
6-hole 25 18.43.0b 0.00.0
12-hole 25 38.911.2b 0.00.0
24-hole 25 110.227.5a 16.37.5

'Column values with different letters are significantly different according to two way ANOVA and LSD crite-
rion at the 0.05 level.

kill R. pomonella alighting on a sphere using pesticide instead of Tangletrap flies
must remain on the sphere long enough to acquire a lethal dose of toxicant. This is
best accomplished when toxicant is combined with a feeding stimulant (such as su-
crose) and a high percentage of alighting flies contact the pesticide/sucrose mixture
(Duan & Prokopy 1995a). The trap designs tested here failed in this regard.
The perforated hollow red spheres were constructed to protect both feeding stimu-
lant and pesticide from rainfall by encasing them within the sphere. Success, however,
is contingent upon the notion that alightingR. pomonella will readily enter trap holes
to gain access to feeding stimulant and pesticide. This did not prove to be the case. In
Experiment 3, only a very small percentage (no more than 16%) of flies alighting on
perforated spheres actually entered a hole, regardless of the number of holes per
sphere. Clearly, this is inadequate, as the vast majority of flies alighting on spheres
would never come into contact with the killing agent. Reluctance to enter openings
into traps has been shown in other tephritid flies as well. Reissig (1976) showed that


Mean Time Per Mean Feeding
Fly type No. Flies Fly Spent on % Feeding Time Per
-Trap Type Alighting Trap (s)SE' SE' FlySE1

Protein Fed
-Fruitect 51 204.341.5a 2.02.0b 5.0--
-Wooden sphere 30 204.846.2a 90.05.6a 212.051.2

Protein Starved
-Fruitect 52 240.227.5a 23.15.9b 162.735.4a
-Wooden sphere 40 161.530.2a 75.06.9a 172.032.3a

Protein fed and protein hungry flies were analyzed separately. For each fly type, column values with different
letters are significantly different according to a two sample t-test at the 0.05 level.
Sample size (n) = 1.

Florida Entomologist 79(2)

with the cherry fruit flies, Rhagoletis fausta (Osten Sacken) and R. cingulata (Loew),
traps requiring the flies to enter constricted openings were not effective. Prokopy &
Economopoulos (1975) showed that non-sticky McPhail traps (which require flies to
enter a port for capture) captured less than half of arriving olive flies. Similarly, Aluja
et al. (1989) found that only 31% ofAnastrepha spp. flies alighting on McPhail traps
were ultimately captured. However, tests have shown that perforated cylindrical traps
baited internally with food odor have promise for capturing both female and male
Mediterranean fruit flies, although the percent of alighting flies that ultimately is cap-
tured is unknown (Heath & Epsky 1995). The reason why most R. pomonella in this
study and most tephritid flies in other studies were not inclined to enter holes in traps
containing bait on the interior is uncertain. Possibly most alighting flies do not come
into contact with plumes of attractive odor emanating from trap holes.
The Fruitect red spheres tested here also failed to elicit a sufficient level of fly feed-
ing to be effective. As was the case with hollow perforated spheres, most R. pomonella
alighting on Fruitect traps departed without ever contacting the site of feeding stim-
ulant and potential killing agent. The problem with Fruitect spheres may be that the
sponge containing the feeding stimulant represents only a small part of the total sur-
face area of the sphere. Conversely, flies that alighted on sucrose-coated red wooden
spheres were exposed to feeding stimulant almost immediately upon tarsal contact
with the sphere surface.
An additional factor to consider is trap attractiveness to foraging flies. We found in
Experiments 1 and 2 that internally-baited red spheres were consistently slightly less
attractive to R. pomonella than externally-baited spheres. A possible explanation for
this is that the amount of odor released may have been reduced by positioning odors
inside the sphere as opposed to outside.
To date, three approaches towards the development of a pesticide-treated sphere
for controlling R. pomonella have been evaluated. The first of these is coating the ex-
terior of the sphere with both feeding stimulant and pesticide. This approach is rep-
resented by the 8 cm wooden spheres described by Duan & Prokopy (1995a). Under
dry conditions, these traps have been shown to be as effective as traditional red sticky
spheres in managing R. pomonella in commercial orchard blocks, with the major
drawback being negative effects of rainfall (Duan & Prokopy 1995b). The second and
third approaches (tested here) attempted to modify sphere design so that feeding
stimulant and pesticide could be protected from rainfall. The second approach places
feeding stimulant and pesticide within the trap interior, thus protecting it from rain.
The third approach places feeding stimulant on the interior which is dispensed to the
surface of the trap through a sponge. Neither of these two designs showed promise as
an alternative to the first approach. Future research efforts will be directed at in-
creasing the residual effectiveness of exterior-coated pesticide spheres using residue
extending agents.


We thank Wayne Rice and Tom Clark for the use of their orchards during this study
and Eric Quist and Joseph Laventure for technical assistance. This work was sup-
ported by the Sustainable Agricultural Research and Education project (SARE) and
the Massachusetts Agricultural Experiment Station under Hatch grant 6-35107 (679).


Anastrepha ludens, A. obliqua, and A. serpentina (Dipter: Tephritidae) on a

June, 1996

Reynolds et al.: Rhagoletis pomonella traps

wild mango tree (Mangifera indica) harbouring three McPhail traps. Insect Sci.
Applic. 10: 309-318.
DUAN, J. J., AND R. J. PROKOPY. 1992. Visual and odor stimuli influencing effective-
ness of sticky spheres for trapping apple maggot flies Rhagoletis pomonella
(Walsh) (Dipt., Tephritidae). J. Appl. Entomol. 113: 271-279.
DUAN, J. J., AND R. J. PROKOPY. 1995(a). Toward developing pesticide-treated spheres
for controlling apple maggot flies (Diptera: Tephritidae): pesticides and resi-
due-extending agents. J. Econ. Entomol. 88: 117-126.
DUAN, J. J., AND R. J. PROKOPY. 1995(b). Control of apple maggot flies (Diptera: Te-
phritidae) with pesticide-treated red spheres. J. Econ. Entomol. (in press).
mass trapping method for the control of Dacus oleae (Diptera: Tephritidae). J.
Econ.Entomol. 84: 564-569.
HEATH, R. R., AND N. D. EPSKY. 1995. Mediterranean fruit fly trap methodology. Proc.
The Medfly in California: Defining Critical Research (Riverside, Calif.).
MACCOLLOM, G. B. 1987. Decrease pesticide costs by using traps. Amer. Fruit Grower
(April): 42-43.
PROKOPY, R. J., AND A. P. ECONOMOPOULOS. 1975. Attraction of laboratory-cultured
and wild Dacus oleae flies to sticky-coated McPhail traps of different colors and
odors. Environ. Entomol. 4: 187-192.
PROKOPY, R. J., S. A. JOHNSON, AND M. T. O'BRIEN. 1990. Second-stage integrated
management of apple arthropod pests. Entomol. Exp. Appl. 54: 9-19.
PROKOPY, R. J., J. J. DUAN, AND X. P. HU. 1995. Progress towards pesticide treated
spheres for controlling apple maggot flies. Proc. 1995 New England Fruit Meet-
ings (Sturbridge, Mass.) (in press).
REISSIG, W. H. 1974. Field tests of traps and lures for the apple maggot. J. Econ. En-
tomol. 67: 484-486.
REISSIG, W. H. 1975. Evaluation of traps for apple maggot in unsprayed and commer-
cial apple orchards. J. Econ. Entomol. 68: 445-448.
REISSIG, W. H. 1976. Comparison of traps and lures for Rhagoletis pomonella and R.
cingulata. J. Econ. Entomol. 69: 639-643.

Florida Entomologist 79(2)

June, 1996


'Fort Lauderdale Research and Education Center
University of Florida, Institute of Food and Agricultural Sciences
3205 College Avenue, Ft. Lauderdale, FL 33314

'Research Associate, National Fund for Scientific Research,
Belgium; Universit6 Libre de Bruxelles, CP 160/12
Laboratoire de Biologie Animale et Cellulaire
Avenue F. D. Roosevelt 50, B-1050 Brussels, Belgium


The soldier and worker castes of Constrictotermes guantanamensis n. sp., are de-
scribed. Known distribution of C. guantanamensis is confined to the arid extreme
southeastern coastal zone of Cuba. This is the fourth described species of this endemic
neotropical genus and the first island record.

Key Words: Nasute, soldier, worker, taxonomy, West Indies, Caribbean.


Se described las castas soldado y obrero de Constrictotermes guantanamensis n.sp.
La distribuci6n conocida de C. guantanamensis esta restringida a la arida zona cos-
tera del extreme sureste de Cuba. Esta es la cuarta especie descrita de este g6nero
neotropical end6mico y el primer registro islefio.

The termite genus Constrictotermes Holmgren (1910) is one of 37 nasutitermitine
genera found in the New World and, except for the pantropicalNasutitermes Dudley and
the southern nearctic limits ofTenuirostritermes Holmgren, all remaining genera are re-
stricted solely to the Neotropics (Pearce & Waite 1994, Constantino 1994, Roisin et al.
1996). Three species of Constrictotermes, all from mainland South America, are cur-
rently recognized (Silvestri 1903, Holmgren 1910, Araujo 1977). The descriptions ofC.
cyphergaster (Silvestri) and C. cavifrons (Holmgren) are based on the soldier and imago
caste. Both occur in Bolivia and Brazil, while the former is also reported from Paraguay
and the latter from Venezuela, Guyana, and Surinam (Araujo 1977, Fontes 1983). No
soldiers have been described from the lone Ecuadorian species, C. latinotus (Holmgren).
Having originally reported the species as Constrictotermes? n. sp. (Scheffrahn et
al., 1994), we now confirm the suspected generic status of this new Cuban termite,
based primarily on the soldier head capsule shape and on worker mandible and diges-
tive tube characters. We herein describe the soldier and worker caste of C. guantana-
mensis n. sp.


Field-collected foraging groups composed of soldiers and workers were preserved
in 75% aqueous ethanol. Two different colony samples of C. guantanamensis were col-

Krecek et al.: New Cuban Constrictotermes

elected under stones in Loma de Macambo (2009'N, 75'28'W) on 1-XII-1971 by Jorge
de la Cruz (collection reference nos. C233 and C236), and one was collected from an
epigeal nest on 24-V-1978 (no. C991) in Tortuguilla (19'58'N, 75'03'W) by J. Krecek.
Both sites are in Guantanamo Province, Cuba. The latter sample, containing many
workers and soldiers, could not be used for descriptive work because alcohol had ac-
cidentally evaporated.
The holotype soldier from Loma de Macambo, labeled C236B, will be deposited in
the termite collection of National Museum of Natural History in Washington, D.C.
Paratype soldiers (from colony samples nos. C233, 11 soldiers, and C236, 9 soldiers)
will be placed at the Institute of Systematics and Ecology of Cuban Academy of Sci-
ences, Havana, in the Florida State Collection of Arthropods, Florida Dept. Agric.
Cons. Serv., Division of Plant Industries, Gainesville, Florida, and at their respective
Measurements of specimens were made with a calibrated ocular micrometer and
follow those defined by Sands (1965) and Roonwal (1970), except soldier maximum
head height (Table 1). Terms used to describe soldier and worker morphology and
color follow those of Sands (1965). Left mandible index of workers equals the distance
between the apical and first marginal tooth divided by the distance between first and
third marginals (Emerson 1960). Scanning electron micrographs of a soldier head and
right worker mandible dehydrated by the method of Nation (1983) were made with a
Hitachi S-4000 field emission microscope at 10kV.
The worker digestive tube was exposed by removal of the abdominal wall and fat
tissue under a dissecting microscope. The digestive tube configuration of C. guantan-
amensis was drawn with the aid of a camera lucida at 50X and described using the ter-
minology of Noirot & Noirot-Timothee (1969) and Kovoor (1969). The enteric valve
was split longitudinally with a fine scalpel. Mandibles and the enteric valve were de-
hydrated and mounted in toto for microphotography. Although our description relies
on older material, the cuticular pigmentation of the specimens appears to have re-
mained essentially intact.


Measurements in mm (n=20) Range Mean+SD Holotype

Head length with nasus 1.52-1.61 1.560.026 1.55
Head length without nasus 0.99-1.05 1.010.024 1.00
Head width, maximum 0.88-0.96 0.920.021 0.93
Forehead width, maximum 0.59-0.65 0.620.015 0.62
Head width, minimum 0.53-0.59 0.580.017 0.55
Nasus width at base 0.14-0.15 0.150.006 0.15
Nasus width at middle 0.11-0.12 0.120.006 0.12
Head height, maximum' 0.68-0.71 0.690.009 0.70
Pronotum width 0.47-0.55 0.520.018 0.52
Pronotum length, maximum 0.23-0.30 0.260.020 0.24
Hind tibia length 1.62-1.78 1.700.042 1.72
Hind tibia width, maximum 0.08-0.09 0.080.005 0.08
Total body length 3.00-4.25 3.520.32 3.50

IMeasured at constriction, postmentum included.

Florida Entomologist 79(2)


IMAGO. Unknown.
SOLDIER (Figs. 1-2, Table 1). Apparently monomorphic. Head capsule brown;
clypeus, labrum, postmentum, tip of mandibular point, and area adjacent to antenna
socket pale brown. Antenna, palpi, mandible excluding tip, all sternites, and legs yel-
low-white. Nasus yellow brown apically, basally sepia brown. Pronotum brown ante-
riorly, remaining tergum pale brown.
Head conspicuously constricted near middle. Posterior lobe bulbous, abruptly
raised behind constriction. In profile, nasus and posterior lobe raised above vertex of
anterior lobe thereby forming concavity. Anterior lobe with inconspicuous but well-
marked hunch behind antennal sockets. Nasus rather long, cylindrical, and weakly
rugose on ventral base; raised about 35 above dorsal plane of head. Rear of head un-
usually high and relatively flat. Mandibles vestigial, but adorned with long points;
tips of points turned outward very slightly. Posterior lobe of head with 2-3 pairs of se-
tae; lateral pairs long (approximately a nasus middle diam), median setae shorter,
usually in pair, sometimes singular or absent. Anterior lobe with seta above each an-
tennal socket. Nasus usually with 1 (0-2) inconspicuous seta near middle.
Antenna very long, with 14 articles; articles of subequal length (about 0.2 mm) ex-
cept 2nd, penultimate and last which are shorter. Legs, including femora and tibiae,
long; hind legs longest. Pronotum saddle-shaped with roundly raised unnotched an-
terior lobe; margin of anterior lobe with about 6-8 long setae. Posterior margins of all
thoracic nota with several alternating short and long (2X) setae. Abdominal tergites
with row of long setae near posterior margins of same length as those on thoracic
nota, and accompanied in parallel by an additional row of shorter (about 2X) setae
with sinuous course; glabrous anteriorly. Abdominal sternites with setae of medium
length covering surface and several longer setae near posterior margins.
Comparisons. Head color of C. guantanamensis is primarily brown and its nasus
cylindrical, while head color in C. cavifrons and C. cyphergaster is very dark sepia
brown, and their nasi are conical. Most C. guantanamensis soldier measurements are
smaller than C. cavifrons and C. cyphergaster (Silvestri 1903, Emerson 1925,
Mathews 1977), namely head width, head height (no data for C. cavifrons), pronotum
width and hind tibia length. Soldier antenna of C. cavifrons is 15-segmented
(Holmgren 1910, Emerson 1925), compared with 14 segments in both C. guantana-
mensis and C. cyphergaster (Silvestri 1903, Mathews 1977).
Comparisons with C. cavifrons are based on Holmgren's (1910) and Emerson's
(1925) descriptions and Constantino's (1991) figures. Comparisons with C. cypher-
gaster are based on soldiers and workers collected from arboreal nest in cerrado for-
mation on X-1961 by Araujo in Brasilia, D.F., Brazil (MZ-USP sample no. 0155).
WORKER (Figs. 3-5, Table 2). Apparently monomorphic. Head pigmentation
mostly sepia brown anteriorly, grading to dark sepia brown posteriorly, with yellow-
white cranial suture. Corners between antennal sockets and postclypeus, postmen-
tum, postclypeus and adjacent frontal area very pale brown. Labrum, palpi, and an-
tenna white or yellow-white, including median part of mandibles. Pronotum pale
brown anteriorly, remaining dorsum yellow-brown, except pale brown posterior areas
of abdominal tergites. Coxa and femur very pale brown; tarsal claws yellow-brown.
Sternum white, yellow-white, or colorless.
Head with conspicuously inflated postclypeus, elevated at near right angle to
frons. Postclypeus about twice as broad as long. Head slightly longer than broad, pos-
teriorly broadly rounded and with Y-shaped spot, a conspicuous remnant of the cra-
nial suture.

June, 1996

Krecek et al.: New Cuban Constrictotermes

Figures 1-2. Scanning electron micrographs of lateral (1) and dorsal (2) views of
Constrictotermes guantanamensis n. sp. soldier head.

Florida Entomologist 79(2)

Figure 3. Scanning electron micrograph of C. guantanamensis right worker man
dible from molar plate view.

Figure 4. Dorsal (D), right (R), ventral (V), and left (L) configurations of the diges-
tive tube in situ of C. guantanamensis worker. CP, crop; M, mesenteron (stippled) in-
cluding mesenteric part of MS, mixed segment; 0, oesophagus; P1, first proctodeal
segment; P2, enteric valve; P3, paunch; R, rectum. Scale bar = 0.5 mm.

June, 1996

Krecek et al.: New Cuban Constrictotermes 185

Antenna with 15 articles (n=3); all elongated and of subequal length, except the
seventh to tenth somewhat longer. Pronotum shallowly saddle-shaped; anterior mar-
gin convex, with median line but lacking median notch. Femora and tibiae long. Ab-
domen voluminous, dorsally arched and laterally slightly flattened.
Head, including postclypeus, with about 8 longer setae and about 40-50 shorter se-
tae. Pronotum with one row of longer setae on anterior margin and two rows along the
posterior margin; meso- and metanotum with two posterior rows of setae only.
Mandibles of Constrictotermes type (Fig. 3). Apical and first marginal teeth of both
mandibles short and deflected downward. Left mandible index 0.32-0.33 (n=3). Molar
plates large, boat-like, and composed of about 10 ridges.
Digestive tube configuration (Fig. 4) with strikingly enlarged crop and moderately
long mesenteron; mixed segment short (length 1.5X width of mesenteron); paunch
large. Enteric valve armature composed of three weakly sclerotized swellings, each
bearing 5-10 stout, sharply-pointed spines measuring about 8mm long by 7pm wide
(Fig. 5).
Comparisons. Only large worker of C. cyphergaster (type species) was available
for comparison. Cranial suture in C. guantanamensis reduced to a Y-shaped spot,
while complete in large worker of C. cyphergaster. The latter condition is similar in the
small worker of C. cavifrons (Emerson 1926). Frons and vertex around cranial suture
bifurcation uneven in C. cyphergaster, while even in C. guantanamensis.
Pilosity of head including postclypeus is characterized by about 8 long setae in
both species, but only in C. guantanamensis complemented by about 40-50 additional
shorter setae. C. cyphergaster has about 10-15 more prominent setae on abdominal
tergites, while C. guantanamensis has more than 50 setae of different lengths on ab-

".." !


Figure 5. Section of enteric valve of C. guantanamensis worker showing one of
three spine-bearing swellings.

Florida Entomologist 79(2)


Measurements in mm (n=2) Range Mean

Head length to postclypeus 0.70-0.71 0.71
Head width, maximum 0.97-0.99 0.98
Postclypeus width 0.50-0.52 0.51
Postclypeus length 0.26-0.23 0.25
Pronotum width 0.49-0.55 0.52
Pronotum length, maximum 0.33-0.36 0.35
Hind tibia length 1.51-1.53 1.52
Hind tibia width, maximum 0.08-0.10 0.09
Total body length 3.90-3.50 3.70

dominal tergites. Worker antennae of C. cavifrons 16-segmented, compared with 15
segments in both C. guantanamensis and C. cyphergaster.
Etymology. Derived from name of type locality.


The new species described here bears all of the following characteristics of the ge-
nus Constrictotermes: soldier external morphology habitus including typically con-
stricted head capsule shape, worker mandibles with long molar areas and deflected
apical teeth, enlarged crop, short mixed segment (length about 1.5X the width of me-
senteron), armature of enteric valve consisting of 3 equal swellings with robust,
pointed spines.
Constrictotermes guantanamensis is the second recently described species in the
family Termitidae from Cuba with Parvitermes subtilis Scheffrahn and Krecek 1993.
The latter occurs also on Hispaniola. The existence of C. guantanamensis on Cuba is
especially remarkable, since we have recently shown (Roisin et al. 1996) that the
Greater Antillean nasute fauna is more endemic than previously suggested.
Constrictotermes guantanamensis is apparently restricted to a xeric and rocky
coastal habitat in extreme southeastern Cuba. The zone represents one of the most
arid and warm ecotypes on the island and contrasts with the moister habitats of
known congeners (Emerson 1925, Araujo 1970, Mathews 1977).
Observations regarding C. guantanamensis biology are limited. Only one epigeal
nest ofC. guantanamensis was observed by the first author. Both C. cavifrons and C.
cyphergaster, in contrast, are arboreal nesters (Araujo 1970). Biology of C. latinotus is
unknown (Holmgren 1910). The C. guantanamensis nest was constructed on rocky
ground, and stood a little broader than high, about 0.5 m in horizontal diameter and
rather pale grayish in color. Peripheral surface layers were not hardened or thick-
ened. Excluding the ubiquitous Nasutitermes, C. guantanamensis is the only species
in the Greater Antilles which constructs conspicuous nests.


We are indebted to the original collector of the described species, Dr. J. de la Cruz
from Institute of Ecology and Systematics, Cuban Academy of Sciences, Havana; for
use of the ICBR Electron Microscope Core Facility at the University of Florida,

June, 1996

Krecek et al.: New Cuban Constrictotermes

Gainesville to produce SEM photographs; and Dr. E. M. Cancello, Museu de Zoologia,
Universidade de Sao Paulo, Brazil, for the gift of Constrictotermes cyphergaster spec-
imens and N.-Y. Su and R. Giblin-Davis for critically reviewing and improving this
contribution. The senior author's early investigation was supported by a Smithsonian
short-term fellowship at the National Museum of Natural History, Washington, D.C.
This is article R-04844 of the Florida Agricultural Experiment Station Journal Series.

ARAUJO, R. L. 1970. Termites of the Neotropical Region. Chapter 12 in Krishna, K.
and F. M. Weesner [eds.] Biology of Termites, Vol. 2. Academic Press, New York.
ARAUJO, R. L. 1977.Cattlogo dos Isoptera do Novo Mundo. Acad. Brasileira de Cien-
cias, Rio de Janeiro, RJ. 92 pp.
CONSTANTINO, R. 1991. Termites (Isoptera) from the lower Japura river, Amazonas
State, Brazil. Bol. Mus. Paraense Emilio Goeldi, ser. Zool. 7: 189-224.
CONSTANTINO, R. 1994. A new genus of Nasutitermitinae with mandibulate soldiers
from tropical North America. Sociobiology 25: 285-294.
EMERSON, A. E. 1925. The termites of Kartabo Bartica District, British Guiana. Zoo-
logica 6: 291-459.
EMERSON, A. E. 1926. Development of a soldier of Nasutitermes (Constrictotermes)
cavifrons (Holmgren) and its phylogenetic significance. Zoologica 7: 69-100.
EMERSON, A. E. 1960. New genera of termites related to Subulitermes from the Ori-
ental, Malagasy, and Australian Regions (Isoptera, Termitidae, Nasutitermiti-
nae). American Mus. Nov. 1986: 1-28.
FONTES, L. R. 1983. Acrescimos e correqoes ao Catalogo dos Isoptera do Novo Mundo.
Revta. brasileira Entomol. 27: 137-145.
HOLMGREN, N. 1910. Versuch einer Monographie der Amerikanischen Eutermes-
Arten. Mitt. naturhist. Mus. Hamburg, 27, Jb. hamburger wiss. Anst. 2: 171-
KOVOOR, J. 1969. Anatomie comparee du tube digestif des termites II. Sous-Famille
des Nasutitermitinae. Insectes Soc. 16: 195-233.
MATHEWS, A. G. A. 1977. Studies on termites from the Mato Grosso State, Brazil.
Acad. Brasileira de Ciencias, Rio de Janeiro, RJ. 267 pp.
NATION, J. A. 1983. A new method using hexamethyldisilazane for the preparation of
soft insect tissue for scanning electron microscopy. Stain Technol. 55: 347-352.
NOIROT, C., AND C. NOIROT-TIMOTHEE. 1969. The digestive system. Chapter 3 in
Krishna, K. and F. M. Weesner [eds.] Biology of Termites, Vol. 1. Academic
Press, New York.
PEARCE, M. J., AND B. S. WAITE. 1994. A list of termite genera (Isoptera) with com-
ments on taxonomic changes and regional distribution. Sociobiology 23: 247-
ROISIN, Y., R. H. SCHEFFRAHN, AND J. KRECEK. 1996. Generic revision of the smaller
nasute termites of the Greater Antilles (Isoptera, Termitidae, Nasutitermiti-
nae). Ann. Entomol. Soc. America (submitted).
ROONWAL, M. L. 1970. Measurements of termites (Isoptera) for taxonomic purposes.
J. Zool. Soc. India 21: 9-66.
SANDS, W. A. 1965. A revision of the termite subfamily Nasutitermitinae (Isoptera,
Termitidae) from the Ethiopian Region. Bull. British Mus. Nat. Hist., Entomol.
Suppl. 4: 1-172.
SCHEFFRAHN, R. H., AND J. KRECEK. 1993. Parvitermes subtilis, a new subterranean
termite (Isoptera: Termitidae) from Cuba and the Dominican Republic. Florida
Entomol. 76: 603-607.
1994. Termites (Isoptera: Kalotermitidae, Rhinotermitidae, Termitidae) of the
West Indies. Sociobiology 24: 213-238.
SILVESTRI, F. 1903. Contribuzione alla conoscenza dei Termiti e Termitofili dell'Ame-
rica meridionale. Redia, Firenze, 1: 1-234.

Florida Entomologist 79(2)


Campo Experimental Rio Bravo, INIFAP-SAGAR
Apartado Postal 172, Rio Bravo, Tamaulipas, M6xico 88900


Susceptibility of eight bermudagrasses (Tifton 68, Tifton 78, Tifton 85, Gigante,
Brazos, Cruza 1, Callie, and NK 37) to the Mexican rice borer, Eoreuma loftini (Dyar)
was evaluated monthly during 1994 in northern Tamaulipas, northeastern Mexico.
Damage by E. loftini occurred year around, peaking in April and December. Borer at-
tack was reduced substantially after mowing the pastures to 3-5 cm height. Suscepti-
bility of the cultivars to borer damage was influenced by stolon width, an important
characteristic for E. loftini tunneling. This relationship between stolon width and
borer damage was explained closely (R = 0.86) by the curvilinear model y = (1.2 +
0.1757 x3)2, where y = borer damage and x = stolon diameter. The widest stemmed ber-
mudagrass, and hence the most susceptible to E. loftini, was Tifton 68. Tunneling be-
havior of E. loftini, as affected by stem width of host plants, is discussed.

Key Words: Mexican rice borer, host plant resistance, seasonality, tunneling behavior.


Se evalu6 mensualmente durante 1994 la susceptibilidad de ocho cultivares de
zacate bermuda (Tifton 68, Tifton 78, Tifton 85, Gigante, Brazos, Cruza 1, Callie y NK
37) al ataque del barrenador del tallo Eoreuma loftini (Dyar) en el norte de Tamauli-
pas, M6xico. Los daios de E. loftini ocurrieron durante todo el aio, aunque la mayor
incidencia se present en Abril y Diciembre. El ataque del barrenador se redujo subs-
tancialmente despues de cosechar los pastos a una altura de 3-5 cm sobre el nivel del
suelo. La susceptibilidad de los cultivares al dano del barrenador result asociada con
el grosor de los estolones, una caracteristica important en los habitos de dano de E.
loftini. Dicha relaci6n entire el grosor del tallo y el dano del barrenador fue explicada
(R2 = 0.86) por el modelo y = (1.2 + 0.175 x3)2, donde y = dano del barrenador y x= dia-
metro del estol6n. El cultivar Tifton 68 result tener los estolones mas gruesos, y por
lo tanto fue el mas susceptible al dano de E. loftini. Se discute el comportamiento de
dano de E. loftini en relaci6n al grosor de los tallos de las plants hospederas.

The Mexican rice borer, Eoreuma loftini (Dyar), is an important pest of sugarcane,
Saccharum officinarum L., field corn, Zea mays L., grain sorghum, Sorghum bicolor
(L.) Moench, and forage grasses in Mexico and southern Texas (Browning & Hussey
1987, Rodriguez-del-Bosque et al. 1988, Youm et al. 1988, Meagher et al. 1994). In ad-
dition, a number of wild grasses serve as hosts to E. loftini (Osborn & Phillips 1946,
Browning & Hussey 1987). Although earlier reports acknowledge the distribution of
E. loftini in southern California and southern Arizona (Dyar 1917, Osborn & Phillips
1946), there is no recent evidence on the pest status of this species in those areas.
Feeding by E. loftini on leaves and stems causes varied damage, including tunneling
near the plant growing point during the early growth stages, and a symptom called

June, 1996

Rodriguez et al.: Susceptibility of bermudagrass to E. loftini 189

"dead heart", which appears as a dead or necrotic center whorl on green shoots
(Browning et al. 1989).
Management ofE. loftini in sugarcane has focused on several tactics, including bi-
ological control (Smith et al. 1987), insecticidal suppression (Meagher et al. 1994),
male mating disruption technique (Shaver & Brown 1993), and host plant resistance
(Pfannenstiel & Meagher 1991). A major concern, particularly in Texas where E. lof-
tini was first detected in 1980, is the potential of this insect pest to expand its geo-
graphical range via transportation of infested plant material such as pasture grasses
(Browning & Hussey 1987). In northern Tamaulipas, a subtropical agricultural region
south of the Texas border, the main agronomic crops are corn and sorghum planted
over approximately one million hectares (30% irrigated; 70% dryland). However, re-
cent reductions in international grain prices and the increasing production costs for
corn and sorghum are forcing many producers to consider production alternatives,
such as forage grasses for feeding livestock. Several cultivars of bermudagrass, Cyn-
odon spp., are being evaluated in northern Tamaulipas for their establishment char-
acteristics, dry matter yield of forage, nutritional quality, and cattle performance
(Palomo & Mendez 1993, 1994). The objective of this investigation was to detect pos-
sible differences in susceptibility of eight bermudagrass cultivars to E. loftini.


This study was conducted in the Campo Experimental Rio Bravo (Rio Bravo Ex-
periment Station), near Rio Bravo, Tamaulipas. In April 1992, eight bermudagrass
cultivars (Tifton 68, Tifton 78, Tifton 85, Gigante, Brazos, Cruza 1, Callie, and NK 37)
were established in 3x4 m plots arranged in a randomized complete block design with
four replications. Plots were evaluated for E. loftini damage during the first week of
each month during 1994 by counting and removing at ground level stolons exhibiting
dead heart symptoms. Excised stolons were transported to the laboratory and exam-
ined for evidence of E. loftini tunneling or living larvae. A proportion (25-50%) of the
collected larvae was reared on artificial diet (Martinez et al. 1986) in the laboratory
for completion of development to corroborate the borer species by examining genitalia
of emergent moths (Bleszynski 1969). Stolons exhibiting dead heart symptoms caused
by factors other than E. loftini (<5%) were omitted from the evaluations. The diam of
10 undamaged stolons per plot was measured (about 5 cm above ground level) at each
sampling date. During 1994, plots were mowed to a height of 3-5 cm during the second
week of January, April, May, July, and September. After mowing, plots were fertilized
with urea (46 kg N/ha) and irrigated (10 cm). No insecticide was applied during the
Dead heart data were subjected to analysis of variance as a factorial design (A, cul-
tivars; B, months), and means separated by Tukey's studentized range test (p = 0.05)
(PROC ANOVA, SAS Institute 1988). Average dead hearts (over 12-mo) of each culti-
var (y) were fit to the curvilinear model = (a + bx3)f, where a and b are constants and
x the average (12-mo) stolon diam.


The only stalk borer attacking bermudagrass throughout the study was E. loftini.
The analysis of variance indicated damage was significantly different (p < 0.05)
among bermudagrass cultivars (A), and months (B), with no AxB interaction. Inci-
dence of E. loftini occurred throughout the year, with the greatest damage occurring
in April and December (Fig. 1). Mowing the pastures substantially reduced E. loftini

Florida Entomologist 79(2)

June, 1996

abc abe

4- bc

3 cd ed

de de


Figure 1. Seasonal damage by E. loftini to bermudagrass (average of eight culti-
vars and four replications in plots of 3x4 m) in northern Tamaulipas, Mexico. Arrows
indicate mowing practices. Means (bars) with the same letter are not significantly dif-
ferent (p < 0.05; Tukey's studentized range test).

attack, a result similar to the findings by Browning & Hussey (1987). Overall, the oc-
currence of dead hearts was 51% lower in those months when mowing was practiced
during the previous month (Fig. 1). The practice of periodically harvesting the grasses
removed or killed the larvae and prevented a continued recruitment of E. loftini lar-
vae throughout the period of study. The maximum incidence of dead hearts (22 per
plot) observed in Tifton 68 in April represented only about 0.2% of the total stolons in
the plot, a proportion that probably did not affect either forage yield or quality. How-
ever, these damage figures are much lower than those observed in larger semicom-
mercial plots planted with Tifton 68 in Rio Bravo in previous years (L.A.R.B.,
unpublished data) and in south Texas, where an average of 38% of the Tifton 68 sto-
lons exhibited E. loftini damage (Browning & Hussey 1987). One possible explanation
of the higher damage in Texas is that counts of damaged stolons included not only
dead hearts, but also other symptoms such as entry holes and discoloration at the
area of attack. Although dead hearts may represent a reasonable and practical pa-
rameter for measuring stolon attack by borers, this method likely underestimates the
actual damage and yield loss. In addition, in the study by Browning & Hussey (1987),
mowing was less frequent than in our study, and borers were probably removed less
frequently, allowing more borer attack and/or pest population buildup.
The most susceptible cultivar throughout the study was consistently Tifton 68 (Ta-
ble 1). Average dead hearts (12-mo) observed in Tifton 68 were 3.4 times greater than
the average of the remaining seven cultivars. In April, when E. loftini incidence
peaked, Tifton 68 had 6.6 times more dead hearts than the average for the other
grasses. A closer look at the data showed that borer damage was positively associated
with stolon diam (Table 1). This relation was explained (R' = 0.8632, df = 5, p < 0.01)

Rodriguez et al.: Susceptibility of bermudagrass to E. loftini 191


Dead Hearts Stolon Diam (mm)
Cultivar (Mean+SEM)' (MeanSEM)'

Tifton 68 7.961.47a 2.090.07a
Tifton 78 3.710.60b 1.260.05b
Brazos 3.310.59bc 1.290.03b
Cruza 1 2.380.47bc 1.320.05b
Tifton 85 2.330.49bc 1.420.04b
Gigante 2.250.41bc 0.890.03cd
NK 37 1.580.33bc 0.840.06d
Callie 0.980.25c 1.170.06bc

'Means within a column followed by the same letter are not significantly different (p<0.05; Tukey's studen-
tized range test). Mean values averaged over 12 months and 4 replications in 3x4 m plots.


0.0 0.2 0.4 0.6 0.8 1.0

1.2 1.4 1.6 1.8 2.0 2.2 2.4
1.2 1.4 1.6 1.8 2.0 2.2 2.4


Figure 2. Relation of dead hearts caused by E. loftini to stolon diam of eight ber-
mudagrass cultivars (average of 12 months and four replications in plots of 3x4 m) in
northern Tamaulipas, Mexico.

1 = TIFTON 68
2 = TIFTON 78
3 = TIFTON 85
6 = CRUZA 1
8 = NK 37


-4 6 3 y= (1.2 +
6 R = 0.8632


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