Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00026
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 2003
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: VID00026
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Full Text

Mihelcic et al.: Pathogenic Microorganisms of Coneworms


Department of Entomology, University of Georgia, Athens, GA 30602

USDA Forest Service, Southern Research Station, 320 Green St. Athens, GA 30602-2044

1Current address: The University of Guelph, Dept. of Plant Agriculture, Guelph, ON, N1G 2W1, Canada

2Author to whom correspondence should be addressed (Email: jhanula@fs.fed.us)



Larvae of the Southern pine coneworm, Dioryctria amatella (Hulst) (Lepidoptera: Pyral-
idae), were collected monthly during the growing seasons of 1996 and 1997 from loblolly
pine, Pinus taeda L., seed orchards in Alabama, Florida, Georgia, South Carolina, and Vir-
ginia, and examined for pathogenic microorganisms. One fungus, Beauueria bassiana (Bals.)
Vuill, a granulosis virus (Baculoviridae: Eubaculovirinae), and a protozoan (phylum Mi-
crospora) were found. Five larvae from three localities were infected with B. bassiana, 37
larvae from six localities were infected with the granulosis virus, and 69 larvae from 5 loca-
tions were infected with the microsporidian. Laboratory trials confirmed that B. bassiana
and the granulosis virus caused coneworm mortality. B. bassiana isolates from all three lo-
cations were equally virulent to late instar larvae. Spores of the unidentified microsporidian
are free, elongate oval, binucleate and contain 13-14 turns of an isofilar polar filament. The
primary sites of infection were the Malpighian tubules and the silk glands. The microsporid-
ian was found in 2 to 51% of larvae sampled. It caused 100% mortality in early instar larvae
allowed to feed on artificial diet contaminated with 3 x 103 or 4.5 x 103 spores. More work is
needed to determine the importance of these pathogens in regulating populations of south-
ern pine coneworms or their potential utility in an IPM program.

Key Words: loblolly pine, Pinus taeda, Dioryctria amatella, pathogens, microsporidia, gran-
ulosis virus, Beauveria bassiana


Larvas del gusano de las bellotas de pino sureio, Dioryctria amatella (Hulst) (Lepidoptera:
Pyralidae) fueron recolectadas mensualmente durante las estaciones de crecimiento de 1996
y 1997 en huertos de semillas del pino, Pinus taeda L., en Alabama, Florida, Georgia, South
Carolina, y Virginia, y fueron examinadas por microorganisms pat6genos. Se encontraron
un hongo, Beauueria bassiana (Bals.) Vuill, un virus granuloso (Baculoviridae: Eubaculovi-
rinae), y un protozoario (Phylum Microspora). Cinco larvas de tres localidades fueron infec-
tadas con B. bassiana, 37 larvas de seis localidades fueron infectadas con el virus granuloso,
y 69 larvas de 5 localidades fueron infectadas con el microsporidio. Pruebas de laboratorio
confirmaron que el B. bassiana y el virus granoloso causaron mortalidad en el gusano de las
bellotas de pino. Los aislados de B. bassiana de las tres localidades fueron igualmente viru-
lentos para las larvas en sus ultimos estadios. Las esporas de un microsporidio no identifi-
cado son libres, alargadas, binucleadas, y contienen 13-14 vueltas de un filamento isofilar
polar. Los sitios principles de infecci6n fueron los tubulos de Malpighi y las glandulas de
seda. El microsporidio fu6 encontrado en un porcentaje del 2 al 51% de las larvas muestra-
das. Se caus6 el 100% de la mortalidad en las larvas en los primeros estadios que se permi-
tieron alimentar de una dieta artificial contaminada con 3 x 10' o 4.5 x 10' de esporas. Se
necesita hacer mas trabajo para determinar la importancia de estos pat6genos para regular
la poblaci6nes de Dioryctria amatella o su uso potential en un program de Manejo Inte-
grado de Plagas.

Florida Entomologist 86(1)


Several coneworm species cause severe seed
losses in loblolly pine (Pinus taeda L.) seed or-
chards. These include the southern pine cone-
worm, Dioryctria amatella (Hulst); the blister
coneworm, D. clarioralis Walker; the webbing
coneworm, D. disclusa Heinrich; and the loblolly
pine coneworm, D. merkeli Mutuura and Monroe
(Yates & Ebel 1975). Dioryctria amatella is the
most serious of these pests.
Dioryctria amatella feeds on first and second-
year cones, rust-infected conelets and terminals
of southern pines (Coulson & Franklin 1970,
Hedlin et al. 1980). It also attacks fusiform rust
galls, Cronartium quercuum (Berkely) Miyabe ex
Shirai F. sp. fusiforme, on stems and branches,
and the cambium of tree wounds as alternate
hosts necessary for overwintering (Neunzig et al.
1964, Coulson & Franklin 1970, Hedlin et al.
1980). The southern pine coneworm has up to
four generations per year in the southern United
States (Ebel 1965, Merkel & Fatzinger 1971,
Coulson & Franklin 1970, Fatzinger 1981), but
most of the damage is caused by progeny of
adults that emerge in the spring (Chatelain &
Goyer 1980).
Females lay eggs on or near second-year cones
(Coulson & Franklin 1970). Once in a cone, larvae
develop through five instars (Fatzinger 1970) that
feed throughout the cones where they eventually
pupate (Neunzig et al. 1964). Dioryctria clariora-
lis, D. merkeli and D. amatella often occur in sec-
ond year cones in the same orchard and several
instars of the same species may be present in a
single cone (Neunzig et al. 1964, Hanula et al.
Despite the importance of D. amatella and
other Dioryctria species in limiting seed produc-
tion, no pathogens have been reported from
southern pine coneworms. However, there are
records of pathogens associated with other Dio-
ryctria species. These include a granulosis virus
from D. abietella (D. and S.) (Zhimerikin & Gulii
1972) and a nuclear polyhedrosis virus from D.
pseudotsugella Munroe (Martignoni & Iwai
1986) and D. sylvestrella (Kunimi 1993). The
fungi Beauveria bassiana and Hirsutella satu-
maensis have been recovered from D. splen-
didella H. and L. and D. syluestrella, (Humber &
Hansen 2001) and Metarhizium anisopliae has
been recovered from D. sylvestrella in Japan
(Kunimi 1993). An unidentified microsporidium
was reported from D. splendidella in Russia
(Sprague 1977).
The objectives of this study were to identify
pathogenic microorganisms present in immature
stages of southern pine coneworms attacking
loblolly pine, their prevalence in natural host pop-
ulations, and their pathogenicity under labora-
tory conditions.


Cones were collected monthly during the sum-
mer (July-September) from six seed orchards in
1996 and from seven seed orchards in 1997. In
1996, samples were collected from loblolly pine
seed orchards in Escambia and Nassau Counties,
Florida; Bibb and Toombs Co., Georgia; Rapides
Parish, Louisiana and York Co., South Carolina.
In 1997, samples were obtained from orchards in
Choctaw Co., Alabama; Escambia Co., Florida;
Toombs Co., Georgia; Rapides Parish, Louisiana;
York and Dorchester Co., South Carolina; and Al-
bemarle Co., Virginia. Cooperating seed orchard
managers collected 50-150 cones per sample from
trees scattered throughout their orchards. Sam-
ples included almost equal numbers of newly at-
tacked cones (green), older infested cones (brown-
green), and old infested cones that were almost
dried (brown). The cones were cut open and all
larvae and pupae removed. Larvae with external
or internal parasitoids were noted along with ca-
davers. Larvae that appeared healthy were
placed individually in small cups containing an
artificial diet (Fedde 1982), held at room temper-
ature until they completed development or died,
and checked periodically for signs of disease.
Those that died were dissected in Ringer's solu-
tion (Poinar & Thomas 1978) and examined for
disease if no external symptoms were evident.
Larvae that appeared unhealthy, or were dam-
aged while being extracted from cones, were dis-
sected immediately in Ringer's solution. Larval
tissues that appeared abnormal during dissection
were prepared in wet mount slides and examined
with a phase contrast microscope (100-1000x) for
Suspected fungal pathogens were isolated ei-
ther by culturing them from hyphae or spores
scraped from the cuticle of cadavers, or by placing
whole, surface sterilized cadavers on growth me-
dia. Surface sterilization was done by immersing
them in a 5% solution of sodium hypochlorite (Na-
C10) followed by three rinses in sterile water (Poi-
nar & Thomas 1978). After pure cultures were
obtained, additional plates were prepared for use
in experiments.
We used Sabouraud dextrose agar (SDA) with
yeast extract for fungal isolation since it is effec-
tive for many entomogenous fungi and the acid
reaction (pH 5.6) retards bacterial growth (Poinar
& Thomas 1978). We added streptomycin sulfate
(0.03g/l) to further inhibit bacterial growth. Cul-
tures were held in a dark growth chamber at 20-
25'C for 1-2 weeks for growth and sporulation.
Fungi that sporulated in culture were stored for
up to 1 month in a refrigerator at 5C after which
new isolates were prepared.
Preparations for microscopic examination
were made by growing fungi on cellulose mem-
branes placed on water agar (Alexopoulos &

March 2003

Mihelcic et al.: Pathogenic Microorganisms of Coneworms

Beneke 1962). Cellulose membranes (dialysis
tubes, Fisher Scientific) were then removed and
examined in wet mount preparations after 24
hours, and every 2-3 hours (for 12 hours) thereaf-
ter, for diagnostic characteristics (Samson et al.
Occasionally we were unable to determine a
cause of death through dissection and light mi-
croscopy. In those cases, the larvae were ground
up in Ringer's solution and a drop of the homoge-
nate was placed on a carbon coated grid, stained
for 5 minutes with 5% uranyl acetate, and exam-
ined by transmission electron microscopy for viral
particles. Forty larvae were examined in this way.
Several experiments were conducted to deter-
mine the virulence of pathogens recovered from
field-infected larvae under laboratory conditions.
Coneworms for these experiments were obtain
from a laboratory colony of D. amatella.
A fungus suspected of causing mortality was
tested in one trial to insure that it was the causal
agent. Thirty late instar D. amatella larvae were
inoculated per inoculum density and we tested
densities of 0, 6.0 x 104, 1.0 x 106 and 1.6 x 105
spores/pl in sterile water. One 2pl droplet of each
spore suspension was placed on the cuticle of the
larvae which were then placed in individual cups
of artificial diet and held in a growth chamber at
26C + 1PC and 92% + 3% RH until they died or
completely developed. Fungi were reisolated from
cadavers which exhibited active fungal growth.
Pure cultures were then examined and compared
to the original cultures to confirm that the same
fungus was present and the cause of mortality.
In a second trial, conducted under the same
conditions, we compared strains of the same fun-
gus isolated from Dioryctria spp. collected from
three different localities at a single inoculum den-
sity of 4.0 x 104 spores/pl. We also compared mor-
tality following inoculation on the cuticle or per
os. Per os inoculations were done by placing 2pl
droplets of 4.0 x 104 spores/pl in a 5% sucrose so-
lution on the mouthparts of CO2 anesthetized lar-
vae. Cuticular inoculations were the same as
To test the granulosis virus, infected, field col-
lected larvae were homogenized in a tissue
grinder, diluted with a 5% sucrose solution, and 21
droplets of this inoculum were placed on the
mouthparts of anesthetized test larvae. Treated
larvae were observed daily and left to develop
completely or die. Observations of shape and size
of virus particles (virus body) and inclusion bod-
ies through TEM were used to determine the type
of virus (Smith 1967).
Microsporidia infections were diagnosed by
dissection and light microscopic examination of
fresh tissue or stained tissue smears. The latter
consist of tissue smeared on slides, air-dried, fixed
in methanol (5min), and stained with a 10% Gi-
emsa stain solution (pH 7.4). Larvae anesthetized

with CO2were inoculated per os with 21l of 1 x 104
microsporidia spores and dissected at 2 day inter-
vals to find which tissues were the primary site of
infection. Geimsa stained smears of various tis-
sues were prepared and examined for infection.
Since the primary site of infection appeared to be
the silk glands, infected portions of that tissue
were prepared for electron microscopy. For rapid
initial fixation of fresh silk glands, infected tissue
was submerged for one hour at 25C in 2.5% (w/v)
glutaraldehyde buffered with 0.1M Na-cacodylate
(pH 7.5) to which 5% (w/v) sucrose and 0.5% (w/v)
CaC12 were added. Tissues were post-fixed with
4% (w/v) OsO4 in cacodylate buffer (pH 7.5) for 1
hour, dehydrated through an ethanol series and
then tissues were submerged in propylene oxide
twice for 20 min. Tissues were embedded in 812
Epon plastic (Sabatini et al. 1963). Sections
were post-stained with 2% (w/v) aqueous uranyl
acetate followed by lead citrate (Harris 1997).
A trial was conducted to determine if the mi-
crosporidian caused mortality in D. amatella lar-
vae allowed to feed on artificial diet contaminated
with spores. Second instar larvae were placed in
1.5 ml microcentrifuge tubes containing a small
quantity of artificial diet contaminated with 0, 5 x
102, 2 x 103, 3 x 103 or 4.5 x 103 spores obtained
from infected tissues of field collected larvae.
Fifty to 60 larvae were treated per dose. The cen-
trifuge tubes were plugged with cotton and the
larvae were allowed to feed for 7 d. After 7 d they
were transferred to 30 ml capacity diet cups with
fresh uncontaminated food. Larvae were moni-
tored every other day until all of the control group
pupated. Those that died were dissected and ex-
amined for the presence of microsporidian spores.
At the end of the experiment all surviving larvae
and pupae were examined for microsporidium in-
Data from laboratory trials were analyzed us-
ing the SAS procedure FREQ to analyze contin-
gency tables (SAS Institute Inc. 1987).


We examined 1626 coneworm larvae from six
locations during the summers of 1996 and 1997.
Of those, 306 (18.8%) were parasitized by other
insects, 32 (2.4%) were infected with a granulosis
virus, 5 (0.4%) were infected with the fungus B.
bassiana, and 69 (5.2%) were infected with a mi-
crosporidium (Table 1).
The virus was identified as a granulosis type
(Baculoviridae: Eubaculovirinae) based the pres-
ence of virions occluded individually in granules
(Federici 1997). A total of 32 field collected larvae
were infected with a granulosis virus at six loca-
tions during 1996 and 1997. In 1998, we found 5
larvae infected in a Baldwin Co., Georgia seed or-
chard. Prevalence ranged from 1 to 14% in field
populations (Table 1). Nineteen of 30 late instar

Florida Entomologist 86(1)


Occurrence of pathogens

Sample location No. of larvae Microsporidia Granulosis virus* B. bassiana
Counties and states year examined No. (%) No. (%) No. (%)

York, SC 1996 88 0 0 0
Escambia, FL 1996 106 7 (6.6) 1(0.9) 0
Nassau, FL 1996 89 0 0 0
Bibb, GA 1996 89 0 0 0
Pineville, LA 1996 114 0 0 0
Toombs, GA 1996 410 10 (2.4) 0 0
Toombs, GA 1997 99 12 (12.1) 13 (13.1) 1(1.0)
Choctaw, AL 1997 74 1(1.3) 7(9.5) 1(1.4)
Escambia, FL 1997 116 17 (14.7) 4 (3.5) 3 (2.6)
York, SC 1997 35 18 (51.4) 5 (14.3) 0
Albemarle, VA 1997 45 0 0 0
Dorchester, SC 1997 34 0 0 0
Rapides, LA 1997 21 4 (19.1) 2 (9.5) 0

Note: Larvae parasitized by insects were not examined for diseases.
*In 1998, 5 additional larvae were found infected with virus in Baldwin CO., GA.

larvae inoculated per os with the virus in labora-
tory trials died, while all of the control group sur-
vived to the adult stage. Infected larvae exhibited
a variety of symptoms including prolonged devel-
opment, cessation of feeding, discoloration of the
integument (light gray or brown) and a milky
white appearance of the hemolymph. The latter
was the most reliable diagnostic characteristic for
determining infection.
The fungus B. bassiana was recovered from Di-
oryctria spp. larvae at three locations, although
prevalence was low (Table 1). In laboratory trials,
B. bassiana (Florida isolate) caused approxi-
mately equal mortality at all inoculum densities
tested (Table 2). In addition, per os inoculations
did not increase mortality over cuticular inocula-
tions with the Alabama isolate and we detected no
differences in mortality among the three isolates
we tested.

Sixty-nine field collected larvae from six loca-
tions were infected with a microsporidia (Table 1).
Prevalence of this disease organism ranged from
2% in Lyons, Georgia to over 51% at Catawba,
South Carolina.
Uninucleate and diplokaryotic vegetative
stages were observed in Geimsa stained smears
using light microscopy (1000x), but only diplokay-
otic stages were detected with the electron micro-
scope. Meronts and/or sporonts of the
microsporidium were diplokaryotic (Fig. 1A).
Spores of the microsporidium observed through
light and electron microscopy, were oval, binucle-
ate, contained a single coiled isofilar polar fila-
ment with 13-14 turns (Fig. 1B), and measured
5.87 x 2.85 pm (n = 20) in fresh preparations.
Spores were not enclosed in any type of sporo-
phorous vesicle and they always occurred individ-
ually. Malpighian tubules and silk glands were


Isolate/inoculation dose spores/pl N No. dead Percent mortality

Controls/cuticle 0 30 4 13.3 a
Controls/per os 0 30 3 10.0 a
Florida/cuticle 6.00 x 104 30 20 66.7 b
1.00 x 105 30 23 76.7 b
1.65 x 105 30 22 73.3 b
Louisiana/cuticle 4.10 x 104 30 21 70.0 b
Alabama/cuticle 4.10 x 104 30 19 63.3 b
Alabama/per os 4.10 x 104 30 20 66.6 b

Note: Percent mortality followed by the same letter are not significantly different (chi-square test, p < 0.01).

March 2003

Mihelcic et al.: Pathogenic Microorganisms of Coneworms



Fig. 1. Electron-micrographs of a microsporidian meront (A) from Dioryctria amatella silk gland containing two
nuclei (n) in a diplokaryotic arrangement and a binucleate spore (B) showing the polar filament (p) with 13-14
turns. Magnification ofA = 40,000x and B = 20,000x.

the primary sites of infection in lightly or newly
infected individuals, but we were unable to deter-
mine which tissue was infected first. In advanced
infections, the microsporidium could be found
throughout the host's tissues including the fat
body, midgut and epidermis. Larvae reared on ar-
tificial diet were successfully infected per os.
Other means of infection or transmission are un-
In laboratory trials, the highest two doses (3x
103 or 4.5x 103 spores) caused 100% mortality in
larvae exposed in the second instar. Only 45% of
the control group survived to pupation but none
was infected with the microsporidium.

We recovered three pathogenic microorgan-
isms from widely separated populations of the
southern pine coneworm. It is likely that these
pathogens are found throughout the range of D.
The fungus, B. bassiana, is a common patho-
gen with an extensive insect host list (Tanada &
Kaya 1993). Although we only isolated it from five
larvae at three locations, laboratory trials showed

that it was capable of causing mortality. The low
occurrence ofB. bassiana in field populations may
be due to the protected habitat of coneworms.
Most cones occur near the tops of pine trees and
once D. amatella larvae enter cones they rarely
leave them, so they are protected from wind-borne
pathogens. Vandenberg & Soper (1978) suggested
that increased frequency of fungal diseases was
due to greater host exposure, favorable physical
conditions for fungal spores and the potential for
spore accumulation in lower canopy areas. It may
be that naturally occurring epizootics of fungal
pathogens in coneworm populations are unlikely
because larvae occupy protected microhabitats,
the majority of loblolly pine cones occur in the
tops of trees, and southern pine forests are ex-
posed to frequent periods of hot, dry weather un-
favorable to spore longevity outside the soil
This is the first granulosis virus reported from
Dioryctria spp. in North America, although one
was reported from D. abietella in Siberia (Zhi-
merikin & Guli 1972). Baculoviridae are the most
commonly observed viral infections in insects
with 80% occurring in the Lepidoptera (Evans &
Entwistte 1987) and all granulosis virus infec-

Florida Entomologist 86(1)

tions have been recorded from Lepidoptera (Fe-
derici 1997). It is unclear why the granulosis virus
is more common than B. bassiana in the southern
pine coneworm. It may be the virus occlusion bod-
ies or capsules are more persistent in the environ-
ment than B. bassiana spores. In addition, under
laboratory conditions the granulosis virus pro-
longed larval development while B. bassiana
killed larvae within a few days. The extended de-
velopment of viral infected individuals may allow
greater viral reproduction (Tanada & Kaya 1993)
or it may have increased the likelihood that we
encountered infected individuals in our samples.
The microsporidian was the most prevalent
pathogen we encountered with up to 51% of the
larvae from the York Co., SC population infected
with this protozoan. The spores, observed
through light and electron microscopy, were oval,
diplokaryotic, and had a long, flexible polar fila-
ment. Based on characteristics of the spores and
the presence of uninucleate and binucleate
stages consistent with the description of the type
species N. bombycis (Sprague et al. 1992) we
thought the microsporidian was a member of the
genus Nosema (Nosematidida: Nosematidae).
However, analysis of small subunit ribosomal
DNA and comparison to an extensive database of
other microsporidia suggests that the microspo-
ridian we found is not closely related to other
Nosema spp. and is not likely a member of that
genus (C. R. Vossbrinck, personal communica-
tion). Further genetic analyses, and light and
electron microscopy studies are underway to de-
termine the identity and phylogenetic relation-
ships of this species.
Although the microsporidium can kill it's host
when they are treated with very high doses, it is
unclear what effect this microsporidium has on
hosts or host populations under natural condi-
tions. Onstad & Maddox (1989) modeled the ef-
fects of N. pyrausta (Paillot) on the population
dynamics of its pyralid host the European corn
borer, Ostrinia nubilalis (Hiibner). Their results
suggest that N. pyrausta can regulate 0. nubilalis
populations well below the carrying capacity of its
environment. The timing of infection is important
in determining what effect a microsporidian has
on its host (Onstad & Carruthers 1990). For ex-
ample, Sajap & Lewis (1992) found that early in-
stars of 0. nubilalis infected with N. pyrausta
formed abnormal pupae or adults. Infections in
later instars resulted in reduced adult longevity
and up to 50% reduction in fecundity. We found
that the microsporidium caused mortality at high
doses, but we ended our experiment at pupation
so we are uncertain if the surviving infected indi-
viduals would have developed normally or experi-
enced normal adult longevity. If the nosema-like
microsporidian we found has similar effects on its
pyralid host, D. amatella, then it may be an im-
portant factor limiting coneworm populations.

The potential utility of the pathogens found in-
fecting D. amatella as biological control agents
needs to be determined. The granulosis virus and
B. bassiana kill their host quickly, and therefore
might have potential as biopesticides. Beauveria
bassiana has the added advantage of being easily
cultured. The microsporidian may be important
in regulating host populations but additional
work is needed to determine its effect. In addi-
tion, these pathogens may be useful in biological
control programs of other Dioryctria spp. Of the
three pathogens reported here, only B. bassiana
was recovered from coneworms in surveys of 5
populations in California and Oregon (Hanula,
unpublished data).

We thank W. Pepper for statistical advice; M. Cody, C.
Crowe, F. Brantley, J. Clark, H. Gresham, E. McCall, M.
Young, A. Mangini, D. Fleming, T. Tigner, M. Powell, B.
Catrett, M. Wilson, V. Hicks and W. Lowe for providing
cone samples; R. Hanlin for determination and confir-
mation of fungal species; Y. Berisford, C. Mims, D.
Dwinell, B. Otrosina and S. Frederich for advice on my-
cology techniques; and M. Farmer for advice on EM tech-
niques. Also, we thank M. Adang, T. Andreadis, and C. W.
Berisford for reviewing early drafts of this manuscript.

ALEXOPOULOS C. J., AND E. S. BENEKE. 1962. Labora-
tory Manual for Introductory Mycology. Minneapo-
lis, MN. Burgess Publishing Company
CHATELAIN M. P., AND R. A. GOYER 1980. Seasonal at-
tack periods of cone-feeding insects of loblolly pine
cones. Ann. Entomol. Soc. Am. 73: 49-53.
COULSON R. N., AND R. T. FRANKLIN. 1970. The biology
of Dioryctria amatella (Lepidoptera: Phycitidae).
Can. Entomol. 102: 679-684.
EBEL, B. H. 1965. The Dioryctria coneworms of north
Florida pines (Lepidoptera: Phycitidae). Ann. Ento-
mol. Soc. Am. 58: 623-30.
EVANS, H. F., AND P. F. ENTWISTLE. 1987. Viral dis-
eases. in J. R. Fuxa, Y. Tanada. [eds.] Epizootiology of
Insect Diseases. John Wiley & Sons, New York.
FATZINGER, C. W. 1970. The number of instars of Dioryc-
tria amatella (Hulst) in north Florida.
USDA FOREST SERVICE RES. Note SE-142. Southeast-
ern Forest Experiment Station, Asheville, North
FATZINGER, C. W. 1981. Activity rhythms, influence of
host plant on mating and oviposition, and rearing of
the southern pine coneworm (Lepidoptera: Pyral-
idae). Fl. Entomol. 64: 506-515.
FEDDE V. H. 1982. Rearing the southern pine coneworm,
Dioryctria amatella (Hulst), on artificial diet. J. Ga.
Entomol. Soc. 17: 237-240.
Federici, B. A. 1997. Baculovirus pathogenesis. in
Miller, L. K. [ed.] The Baculoviruses. Plenum Press,
New York.
1985. Adult activity of Dioryctria amatella (Lepi-
doptera: Pyralidae) in relation to development of im-
mature stages in loblolly pine cones. Environ.
Entomol. 14: 842-845.

March 2003

Mihelcic et al.: Pathogenic Microorganisms of Coneworms

HARRIS J. R. 1997. Negative Staining and Cryoelectron
Microscopy: the Thin Film Techniques. BIOS Scien-
tific Publishing, Herndon, VA.
T. W. KOERBER, AND E. P. MERKEL. 1980. Cone and
Seed Insects of North American Conifers. Canadian
Forestry Service, Ottawa, Ontario, United States
Forest Service, Washington, DC, Secretaria de Agri-
cultura y Recursos Hidraulicos, Mexico.
Collection of Entomopathogenic Fungal Cultures:
Catalog of Isolates. USDA Agricultural Research
Service, Plant Protection Research Unit, Ithaca, NY.
KUNIMI, Y. 1993. A list of insect diseases of Japanese in-
sects. In H. Iwahan, M. Okada, Y. Kunimi and M.
Shimazu [eds.], Manual for the Research on Insect
Pathogens. Nihon Shokubutuboeki-kyoukai, Tokyo,
MARTIGNONI, M. E., AND P. J. IWAI. 1986. A catalogue of
viral diseases of insects, mites and ticks. USDA For-
est Service Gen. Tech. Rep. PNW-195, Pacific North-
west Research Station, Corvallis, Oregon.
MERKEL E. P., AND C. W. FATZINGER 1971. Periodic
abundance of pine cone-infesting Lepidoptera in
black light traps and sleeve cages in north Florida.
Fl. Entomol. 54: 53-61.
1964. Observations on the biology of four species of
Dioryctria in North Carolina (Lepidoptera: Phyciti-
dae). Ann. Entomol. Soc. Am. 57: 317-321.
ONSTAD D. W., AND J. V. MADDOX. 1989. Modeling the
effects of the Microsporidium, Nosema pyrausta, on
the population dynamics of the insect, Ostrinia nubi-
lalis. J. Invert. Pathol. 53: 410-421.
ONSTAD D. W., AND R. I. CARRUTHERS. 1990. Epizootio-
logical models of insect diseases. Ann. Rev. Entomol.
35: 399-419.
POINAR JR., G. O., AND G. M. THOMAS. 1978. Laboratory
Guide to Insect Pathogens and Parasites. Plenum
Press, New York.

The preservation of cellular ultrastructure and enzy-
matic activity by aldehyde fixation. J. Cell Biol. 17:
SAMSON, R. A., H. C. EVANS AND J. P. LATGE. 1988. At-
las of Entomopathogenic Fungi. Springer-Verlag,
New York.
SAJAP, A. S., AND L. C. LEWIS. 1992. Chronology of infec-
tion of European corn borer (Lepidoptera: Pyralidae)
with the microsporidiumNosema pyrausta: effect on
development and vertical transmission. Environ.
Entomol. 21: 178-182.
SAS INSTITUTE INC. 1987. Guide for personal comput-
ers, version 6 edition. SAS Institute Inc., Cary, NC.
SMITH, K. M. 1967. Insect Virology. Academic Press,
New York.
SPRAGUE, V. 1977. Annotated list of species ofmicrospo-
ridia. In L. A. Bulla, Jr. and T. C. Cheng. Compara-
tive Pathobiology: Volume 2, Systematics of the
Microsporidia. Plenum Press, New York.
onomy of phylum Microspora. Critical Rev. Micro-
biol. 18: 285-395.
TANADA Y., AND H. K. KAYA. 1993. Insect Pathology. Ac-
ademic Press, San Diego, CA.
VANDENBERGJ. D., AND R. S. SOPER 1978. Prevalence of
Entomophthorales mycosis in populations of spruce
budworm, Choristoneura fumiferana. Environ. En-
tomol. 7: 847-853.
YATES III H. O., AND B. H. EBEL. 1975. A light trapping
guide to seasonal occurrence of pine seed- and cone-
damaging moths of the Georgia Piedmont. USDA
Forest Service Res. Note SE-210. Southeastern
Forest Experiment Station, Asheville, North Caro-
ZHIMERIKIN V. N., AND V. V. GULII. 1972. Granulosis of
Dioryctria abietella Schiff. Izvestiya Sibirskogo Ot-
deleniya Akademii Nauk SSSR, Seriya Biolog-
icheskikh Nauk 5: 68-73.

Florida Entomologist 86(1)

March 2003


1Dillard Drive Middle School, 5200 Dillard Drive, Raleigh, North Carolina 27606

2University of Florida, Gulf Coast Research and Education Center, Bradenton, Florida 34203

3University of Florida, Department of Agricultural and Biological Engineering, Gainesville, Florida 32611


Liriomyza trifolii (Burgess) is an important leafmining pest of numerous ornamental and
vegetable crops. The pest is attacked by many species of parasitoids which inflict heavy
mortality in the absence of insecticides. The functional response, as well as parasitoid-in-
duced mortality, of one major parasitoid, Diglyphus intermedius (Girault), was estimated
over a range of densities of third instar L. trifolii at 25-27 C in the laboratory. The func-
tional response of D. intermedius was given by the equation Y = Kp-n2-(l-exp(-n,/Kh))
where Y, the rate of parasitism, is the number of hosts parasitized per day; Kp is a con-
stant of 7.3908 hosts parasitized per parasitoid-day; n, and n2 are the densities of leaf-
miner larvae and parasitoid adults (numbers/cm2 leaf area), respectively, and Kh is a
constant of 0.0144 leafminer larvae/cm2 leaf area. The relationship between host and par-
asitoid density and parasitoid-induced host mortality was given by Z = Cp.n.-(l-exp(-n/
Ch)) where Z, the rate of parasitoid-induced mortality, is the number of leafminer larvae
killed per day; Cp is a constant of 9.2064 hosts killed per parasitoid-day and Ch is a con-
stant of 0.0165 leafminer larvae/cm2 leaf area. The observed rate of parasitism at a par-
ticular parasitoid density was always lower than the observed rate of parasitoid-induced
host mortality at that density. Lower leafminer larval densities resulted in increased
multiple oviposition by D. intermedius. When eggs of the parasitoid were placed at in-
creasing densities on leafminer larvae in artificial mines, the number of parasitoid eggs
surviving to adulthood decreased while the number of individuals surviving per host
tended to remain at about one.

Key Words: Liriomyza, Diglyphus, biological control, leafminer, parasitism, host-killing,
functional response


El minador de hojas, Liriomyza trifolii (Burgess) es una plaga important de numerosas
plants ornamentales y de hortalizas. Varias especias de parasitoides atacan esta plaga in-
flijiendo una mortalidad alta en la ausencia de insecticides. Se estimaron la respuesta fun-
cional y la mortalidad inducida por parasitoides, en el parasitoide principal, Diglyphus
intermedius (Girault), en varias densidades del L. trifolii en el tercer estadio a 25-27 C en el
laboratorio. La respuesta funcional de D. intermedius fue dada por la ecuaci6n Y = Kp-n.-(1-
exp(-n,/Kh)) donde Y (la tasa de parasitismo) es el numero de hospederos parasitados por
dia; Kp es un constant de 7.3908 hospederos parasitados por dia-parasitoid; n, y n, son las
densidades de las larvas del minador y los adults de parasitoides (numeros/area cm2 de la
hoja), respectivamente, y Kh es una constant de 0.0144 larvas de minador /area cm2 de la
hoja. La relaci6n entire la densidad del hospedero y la densidad del parasitoide y la mortali-
dad inducida por el parasitoide ha sido dada por Z = Cp.n2.(l-exp(-n/Ch)) donde Z (la tasa de
mortalidad inducida por el parasitoid) es el numero de larvas de minador matadas por dia;
Cp es una constant de 9.2064 hospederos matados por dia-parasitoide y Ch es una cons-
tante de 0.0165 larvas de minador/area cm2 de la hoja. La tasa de parasitismo observada en
una densidad especifica del parasitoide siempre fu6 mas baja que la tasa de mortalidad ob-
servada inducida por el parasitoide de la misma densidad. Las densidades menores de lar-
vas de minador resultaron en un aumento de oviposiciones multiples por D. intermedius.
Cuando los huevos del parasitoide fueron colocados en densidades crecientes sobre las larvas
de minador en minas artificiales, el numero de huevos parasitados que sobrevivieron hasta
el estado adulto diminuy6 mientras que el numero de individuos quesobrevivieron por hos-
pedero tendieron a permanecer aproximadamente uno por hospedero.

Patel et al.: Parasitism and Host-Killing by Diglyphus

Liriomyza trifolii (Burgess) is an important
pest of many ornamental and vegetable crops in-
cluding tomato. Over 40 species of parasitoids
have been recovered from Liriomyza spp. leafmin-
ers (Waterhouse & Norris 1987), including 20 in
Florida (Schuster et al. 1991, Schuster & Wharton
1993) where, in the absence of insecticides, para-
sitism of the leafminer has ranged from about 65
to 75%. Applications of broad spectrum insecti-
cides like methomyl have resulted in a decline in
parasitism followed by an increase in leafminer
density (Oatman & Kennedy 1976). Action
thresholds for timing insecticide applications
have been established for tomato (Pernezny et al.
1996, Schuster et al. 1996). While the extent of
parasitism might be taken into account during
the sampling process (Schuster et al. 1996), the
relationship between leafminer and parasitoid
density and subsequent parasitism and host
death is not known.
Diglyphus intermedius (Girault) was one of the
most abundant parasitoids found attacking L. tri-
folii on tomato in Florida (Schuster & Wharton
1993). The parasitoid is ectoparasitic and prefers
third instar leafminers for oviposition. The life-
time fecundity (F) and the lifelong total number of
hosts killed (Hm) were found to be functions of
temperature (T) and were represented by F = -
196.11 + 42.65T 1.1T2 and Hm = 721.97 19.1T,
respectively (Patel & Schuster 1991). The temper-
ature range used in these experiments was 15.6 to
31.1C, which should be the limits for making in-
terpretations and predictions. The peak 3 day
moving averages of the number of eggs deposited
and the number of hosts killed were highest at
23.3 and 26.7C. The relationship between host
and parasitoid densities was not addressed in
these experiments. This relationship is often re-
ferred to as the functional response and can be ei-
ther curvilinear (type 2 response) or sigmoid (type
3 response) (as summarized by Price 1997). The
plateau of the type 2 response results from the
limitation of prey handling at higher prey densi-
ties and is characteristic of invertebrate parasit-
ism or predation. With the type 3 sigmoid
response, the efficiency or rate of capture in-
creases as the predator learns to find and recog-
nize prey, with a resulting rapid increase in
predation. Eventually, a plateau in the number of
prey captured is reached as the limitation in prey
handling is reached. The type 3 response is char-
acteristic of vertebrate predation, although inver-
tebrates can also demonstrate this response.
The leafminer-parasitoid interaction was a fo-
cus of the leafminer population dynamics model
proposed by Smerage et al. (1980), which provides
an excellent framework for the elucidation of the
role of natural enemies in regulating leafminer
populations. The model pertained to within-field
populations of eggs, larvae, pupae and adults of
leafminers and their parasitoids, broadly ex-

pressed as processes that contribute to the overall
dynamics of the population. The rate of parasit-
ism in the model was a function of host and para-
sitoid densities and the relationship between the
rate of parasitism and host and parasitoid densi-
ties at constant temperature was hypothesized to
be an exponential equation generating families of
curves at different host and parasitoid densities.
The purpose of the present investigation was
to obtain a mathematical description of the func-
tional response ofD. intermedius at various para-
sitoid densities when using L. trifolii as a host at
a constant temperature.


Several limitations were encountered in the
experimental design. Ideally, observations on par-
asitoids should be made in a large arena which al-
lows the parasitoids to move naturally in and out
of the arena. This would enable the parasitoids to
behave naturally and to have random access to
hosts. Also, the likelihood of encountering the
same host again would not be increased due to
confinement to a restricted arena. Arena size
would certainly be important if the period of ob-
servation was long and parasitoids were confined
to a small region. Because of the small size of
adult D. intermedius, it is not possible to monitor
adult parasitoid densities in a large arena with
confinement, nor is it possible to maintain unifor-
mity in leafminer and parasitoid densities be-
tween observation plots and to maintain constant
environmental conditions. To overcome these lim-
itations, a compromise was made. Female D. in-
termedius were confined in 67 x 67 x 67 cm cages
within a controlled environment room for 12 h ob-
servation periods. There were several advantages
to using cages. D. intermedius density on a per
cage basis was easy to record and manipulate, and
other parasitoid species could be excluded. The
relatively small cage size permitted them to be
maintained in a controlled environment room.
Tomato plants, Lycopersicon esculentum Mill.
cv Hayslip, used in the experiments were approx-
imately 30 days post transplanting, 50 to 60 cm
tall with 10 to 13 leaves, and were just beginning
to flower. Plants of this size just fit into the above
cages and allowed the maximum amount of leaf
area possible on one plant confined within a cage.
The plants selected had approximately 2,000 to
3,000 cm2 leaf area, although on occasion some
plants were smaller or larger. Three selected
plants were placed in each of two 67 x 67 x 67 cm
cages. The number of leafminer adults released
into the cages ranged from as few as 10 to as many
as 50 and the duration of exposure ranged from as
little as one minute to as long as 4 h. By manipu-
lating the number of adults and the exposure pe-
riod, leafminer larval densities ranging from 0 to
0.06 larvae/cm2 were obtained. The range then

Florida Entomologist 86(1)

was divided into six classes of equal size for exper-
imentation. After exposure to leafminer adults,
the plants were removed, examined for leafminer
adults (which were removed, if present) and
moved to cages of similar size in a temperature-
controlled room at 25-27 C and a photoperiod of
14:10 (L:D) h. After five d, the five plants most
similar in height, quality and density of third in-
star leafminers were removed at 0800 h and
placed individually into each of five 67 x 67 x 67
cm cages.
The parasitoid adults used for the experiments
were collected from a laboratory colony main-
tained on L. trifolii on tomato (Patel 1987). Ten fe-
males and five males were confined in each of four,
150 x 15 mm Petri dishes. Each dish was provi-
sioned daily with 40 to 60 third instar leafminers
in tomato leaflets. On the fourth day, female par-
asitoids were isolated singly in 00 gelatin cap-
sules and the males were discarded. Thus, the
females were at the age of peak oviposition and
peak host mortality inducement (Patel &
Schuster 1991). One, two, three, four, or five par-
asitoid females were then released into each of
the five cages holding the leafminer- infested
plants. After 12 h (2000 h) the plants were re-
moved from the cages, shaken to dislodge parasi-
toids and transferred to the laboratory where
plant height and age, and the numbers of leaves
and leaflets were recorded. The total leaf area, the
area of leaflets containing leafmines and the area
of leaflets containing parasitized leafminer larvae
were measured (LI-3000, LiCor, Lincoln, NE).
The numbers of live and paralyzed (or dead) lar-
vae were recorded in each of three categories: old
leaves (usually the first three leaves which
showed signs of yellowing), fully expanded leaves
(the majority of leaves) and non-expanded leaves
(usually the top three to four leaves, although oc-
casionally there may have been several more in a
small, tight bundle at the plant apex). Leafmines
containing paralyzed (or dead) larvae were dis-
sected and the number of parasitoid eggs depos-
ited on each leafminer larva was recorded.
Paralyzed or dead leafminer larvae on which no
parasitoid eggs were deposited were categorized
as parasitoid-induced mortality because in previ-
ous studies (Patel & Schuster 1991) no mortality
of leafminer larvae was observed in the absence of
parasitoid females.
The experiment was repeated four times. The
assignment of parasitoid density to the cages was
rotated so that the same cage had the same para-
sitoid density every fifth time. Because the plants
varied in leaf area, and leafminer oviposition
could not be made uniform on each plant, it was
not possible to regulate leafminer density as uni-
formly as parasitoid density. The number of leaf-
miner larvae on each plant, and, hence, in each
cage, could have been kept the same by destroying
some larvae on each plant; however, this would

not have been appropriate for two reasons. Para-
sitoids searching a greater leaf area with the
same number of leafminers per cage would take
longer than for the same number of leafminers on
a smaller leaf area. Also, destroying leafminer lar-
vae may elicit its own response from the parasi-
toid adults, if the adults cue in to host plant
damage in locating leafminer larvae.
To study the survival of parasitoid larvae to
adulthood as influenced by the initial density of
parasitoid eggs per third instar leafminer, para-
sitized third instar leafminers and associated par-
asitoid eggs were dissected from real leafmines
obtained from the D. intermedius colony and were
placed in artificial mines (Patel & Schuster 1983).
The artificial mines consisted of a piece of con-
struction paper with a 7 mm hole and a piece of fil-
ter paper, both equal in dimensions to a
microscope slide, sandwiched between a glass mi-
croscope slide and a cover slip. Parasitoid eggs,
less than 24 h old, were dissected from leafmines
containing parasitized leafminer larvae and
placed next to the larvae in the artificial mines at
densities of one, two, three, or four eggs per larva.
Each artificial mine was sealed with sticky tape
and kept in a 100 x 15 mm Petri dish containing
a water moistened filter paper to provide rela-
tively high humidity in order to prevent desica-
tion of the leafminer larva and the parasitoid
egg(s). The number of host larvae utilized was 12
with one egg per larva, six with two eggs per
larva, four with three eggs per larva, and three
with four eggs per larva. Thus, there were 12 eggs
at each egg density. The Petri dishes were main-
tained at the same conditions as the leafminer
per parasitoid density study and the filter paper
in each Petri dish was moistened daily until no
more parasitoid adults emerged. The experiment
was replicated five times and the number of par-
asitoids emerging from each mine was recorded.
The NLIN procedure (SAS Institute 1982) was
utilized to determine the parameters for the hy-
pothesized exponential equations describing the
relationships between leafminer larval density
and the rate of parasitism and the rate of parasi-
toid-induced mortality at various parasitoid den-
sities. Chi square tests were performed to
determine if there was any effect of altering either
leafminer or parasitoid density on the numbers of
parasitoid eggs on paralyzed host larvae. The
ANOVA procedure (SAS Institute 1982) was uti-
lized to determine if any relationship existed be-
tween the initial total number of parasitoid eggs
placed at each egg density and the total number of
individuals reaching the adult stage at each egg
density in the artificial mines.


The exponential equation obtained using the
nonlinear regression to express the relationship

March 2003

Patel et al.: Parasitism and Host-Killing by Diglyphus

between host and parasitoid density and parasit-
ism was

Y = Kp n, (1-exp(-n,/Kh)) (r2= 0.79)

where Y, the rate of parasitism, is the number of
hosts parasitized per day. Kp is a constant of
7.3908 hosts parasitized per parasitoid-day and n1
and n2 are the densities of leafminer larvae and
parasitoid adults (numbers/cm2 leaf area), respec-
tively. Kh is a constant of 0.0144 leafminer larvae/
cm2 leaf area. The exponential equation express-
ing the relationship between host and parasitoid
density and parasitoid-induced host mortality

Z = Cp.n2.(1-exp(-n/Ch)) (r 2= 0.78)

where Z, the rate of parasitoid-induced mortality,
is the number of leafminer larvae killed per day.
Cp is a constant of 9.2064 hosts killed per parasi-
toid-day and Ch is a constant of 0.0165 leafminer
larvae/cm2 leaf area.
The hypothesized relationship between host
density, parasitoid density and the rate of parasit-
ism utilizing the above equation is depicted in Fig.
1 and the hypothesized relationship between the
rate of parasitoid-induced mortality and host and
parasitoid density is depicted in Fig. 2. The ob-
served rate of parasitism at a particular parasi-
toid density was always lower than the observed
rate of parasitoid-induced mortality. These re-
sults confirmed previous findings that D. interme-
dius killed more hosts than it parasitized (Patel &
Schuster 1991); however, the proportion of hosts
killed that were also parasitized was not the same
at all leafminer densities. In the 0-0.0099 host lar-
vae/cm2 range, 24 of 38 observations (63%) had a
100% parasitization rate while only 5 of 19 (26%)
and 3 of 33 (10%) observations had 100% parasit-
ization of the killed hosts when leafminer larval
densities ranged from 0.01 0.0199 and 0.02 0.06
larva/cm2 leaf area, respectively.

, 35

S20 n=
6 5 1

0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 0.055 0.06
No. leafminer larvae/cm2, ni

Fig. 1. The relationship between density of Liri-
omyza trifolii larvae (n,) and parasitism rate at differ-
ent densities of the parasitoid Diglyphus intermedius
(n2, number of females/plant).

40 nl=5
S 35
I n2=4
0 30
is n,=2

0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 0.055 0.06
No. leafminer larvae/cm2, nl
Fig. 2. The relationship between density of Liri-
omyza trifolii larvae (n,) and rate of parasitoid- induced
mortality at different densities (n2) of the parasitoid
Diglyphus intermedius (n2, number of females/plant).

D. intermedius did not always deposit a single
egg per host as was assumed by Smerage et al.
(1980) (Table 1). A x2 test of the data in Table 1
suggested a significant relationship between the
number of parasitoid eggs per host and host den-
sity (x2 = 67.62; df= 25; P < 0.001), while grouping
the data according to parasitoid density sug-
gested no significant relationship between the
number of parasitoid eggs and parasitoid density
(X2 = 15.6; df = 20; P > 0.70). Increasing the num-
ber of artificially established parasitoid eggs per
host larva from one to four resulted in decreasing
numbers of individuals (of 12) surviving to adult-
hood from about 10 to three (Table 2). The differ-
ence in the number of parasitoids surviving per
leafminer larva was significantly different be-
tween initial egg densities of one and three eggs
per leafminer larva; however, there was little dif-
ference biologically in the numbers surviving per
larva. Regardless of the initial density of parasi-
toid eggs per leafminer larva, generally one para-
sitoid survived per leafminer larva.


The results of this investigation clearly demon-
strate that both parasitism and parasitoid-in-
duced mortality of L. trifolii by D. intermedius are
representative of the type 2 functional response.
This suggests that handling time becomes a limi-
tation to parasitization and host-killing as host
density increases. The curvilinear description of
the functional response further suggests that in-
creased experience ofD. intermedius with its host
did not result in an increased rate of host discov-
ery or decreased rate of host handling, i.e. learn-
ing. However, the female parasitoids used in these
experiments already had been exposed to leaf-
miner larvae for four days when the experiments
were initiated. It is possible that naive females
would exhibit a type 3 functional response. Even
if this were so, the effect would be expected to di-

Florida Entomologist 86(1)


No. parasitoid eggs/killed leafminer host larva

Leafminer larvae/cm2 leaf area 0 1 2 3 4 4+ Total

0.00 0.0099 19 108 60 19 11 16 233
0.01 -0.0199 42 189 74 31 8 6 350
0.02 0.0299 49 163 50 11 10 3 286
0.03 0.0399 45 180 46 10 2 2 285
0.04 0.0499 43 135 43 9 1 0 231
0.05 0.0600 16 40 13 5 0 1 75
0.00 0.0600 214 815 286 85 32 28 1460

minish within at least four days, if not sooner. The
functional response was studied in a room main-
tained at 25-27C, using four- day-old parasitoids.
These conditions of temperature and parasitoid
age are ideal for estimating the maximum rate of
parasitism. Both parasitoid age and temperature
affect the rate of parasitism (Patel & Schuster
1991). Altering temperature and utilizing parasi-
toid females younger or older than four days old,
unless naive females behave differently, would
likely affect the observed rates of parasitism but
not the exponential nature of the relationship be-
tween host and parasitoid density.
The term host-feeding was not used in this
study to describe parasitoid-induced mortality be-
cause, as was shown in a previous study (Patel &
Schuster 1991), D. intermedius kills more hosts
than it parasitizes and because killed larvae can
be used for oviposition or host-feeding or can be
rejected, as was the case for D. begin (Heinz &
Parrella 1989). In this latter study, the proportion
of stung hosts in each category varied depending
upon the host size distribution which the parasi-
toids were provided. When a large host size distri-
bution (third instars) was encountered, 35% of the
larvae were killed without oviposition (20% for
host-feeding), while when a small host size distri-
bution (late second and early third instars) was
encountered, 62% were killed (18% for host feed-
ing). Thus, the proportion of larvae used for host-

feeding remained fairly constant, while the pro-
portion rejected increased as the frequency of
small larvae increased. In a later study, Heinz &
Parrella (1990a) observed that D. begin killed 1.3
L. trifolii larvae for every larva used for oviposi-
tion (about 23%) and made no distinction as to
whether the excess larvae were used for host-
feeding. Minkenberg (1989) determined that the
proportion of L. bryoniae (Kaltenbach) larvae at-
tacked by Diglyphus isaea (Walker) and used for
host-feeding varied from 15 to 40%; however, he
considered any stung host without oviposition to
have been fed upon. In the present study, all L. tri-
folii larvae were large (third instars); neverthe-
less, the number of larvae killed per day always
exceeded the number of larvae parasitized. About
15% of the larvae killed in the experiment were
not used for oviposition (Table 1), which is lower
than the total host-kill observed by Heinz & Par-
rella (1989) and Minkenberg (1989) but is about
the same as the total host-kill observed by Heinz
& Parrella (1990a) and about the same as the per-
centage used for host feeding observed by Heinz &
Parrella (1989). The differences could be due to
the larger arena used in the present study (67 x
67 x 67 cm cages) compared to the smaller arenas
used by Heinz & Parrella (1989) (9 cm diam Petri
dishes and 11.5 cm diam by 13.2 cm high cylindri-
cal cages) and Minkenberg (1989) (7.5 cm diam by
61 cm high cylindrical cages). The cage size used


Initial egg density/host larva Mean no. eggs Survivorship
(12 eggs at each density) surviving to adulthood of eggs/larva

1 9.6 a 0.80 b
2 6.6 b 1.10 ab
3 6.0 b 1.35 a
4 3.0 c 0.98 ab

Means in a column followed by the same letter are not significantly different at the P = 0.05 level, Duncan's multiple range tests.

March 2003

Patel et al.: Parasitism and Host-Killing by Diglyphus

by Heinz & Parrella (1990a) (50 x 50 x 50 cm) was
similar to the cage size in the present study. Heinz
& Parrella (1990a) also believed that the lower
percentage of host-killing that they observed rel-
ative to other studies was due to the use of a
larger cage. In the present study, the proportion of
larvae killed in the absence of oviposition in-
creased from 8 to 20% as the host density in-
creased (Table 1). It is clear from the present
study with D. intermdius and with the above
studies with other species of Diglyphus that more
hosts are killed than are needed for oviposition,
that at least some are used for host-feeding, and
that some may be rejected. While host-feeding can
provide parasitoid females with necessary nutri-
ents for egg development, the benefit of host rejec-
tion is less apparent, particularly in light of the
energy and time expended in the process. Perhaps
host rejection is a mechanism for managing the
density of leafminer larvae on individual leaflets,
thus ensuring that a leaflet containing parasit-
ized larvae will not be lost to the leafmining of
surviving, non-parasitized larvae on the same
leaflet. Excessive leafmining can cause desicca-
tion, necrosis and abscission of tomato leaflets,
thus potentially resulting in reduced survival of
parasitoid larvae.
Little is known regarding the ability of Digly-
phus species to locate infested leaflets relative to
non-infested leaflets; however, other species of
parasitoids attacking Liriomyza spp. have been
shown to have discriminatory behaviors. In flight
tunnel and olfactometer studies, Petitt et al.
(1992) demonstrated that females of Opius dissi-
tus Musebeck responded preferentially to olfac-
tory cues emanating from foliage infested with
larvae of L. sativae Blanchard. A greater propor-
tion of Dacnusa sibirica Telenga flew upwind in
no-choice flight tunnel experiments when leaf-
miner infested plants were placed upwind (Dicke
& Minkenberg 1991). When given a choice, female
D. begin landed on leaves mined by L. trifolii
more than on leaves not mined (Heinz & Parrella
1990a). Although the present cage experiments
were not designed specifically to study host-find-
ing behavior, some deductions regarding the effect
of spatial heterogeneity of host larvae on parasit-
ism rate can be made. In 68 of 79 observations, the
leaflet with the greatest number of leafmines was
encountered by parasitoid adults as indicated by
at least one leafminer larva being parasitized on
that leaflet. In another four observations, the par-
asitoids similarly found and oviposited on larvae
in the leaflet with the next highest leafminer den-
sity. In the remaining seven observations, the
maximum number of leafminer larvae on the leaf-
lets was four. These observations suggest that ei-
ther the parasitoids were more attracted to
leaflets that were more heavily mined or that,
once landing on a leaflet, the parasitoids were
more likely to encounter a host if the host density

on that leaflet was higher. Which behavior, or
maybe both, that was exhibited cannot be deter-
mined because the area of each leaflet and the dis-
tance between individual leafminer larvae were
not measured.
Heinz & Parrella (1990b) observed rates of su-
perparasitism of 0 and 3.1% when D. begin was
released in greenhouses for control of L. trifolii.
These rates are much lower than the 35% super-
parasitism observed in the present study
(Table 1). This much higher rate of superparasit-
ism could have resulted from the confinement of
D. intermedius in cages. Parrella et al. (1989) ob-
served that superparasitism of L. trifolii by D. be-
gini would occur in the rearing method they
developed but that the extent was not known.
They further observed that, if two parasitoid eggs
were deposited adjacent to a third instar leaf-
miner, two adults would be produced, although
they would be smaller. In the present study, only
one parasitoid generally completed development
per leafminer larva, regardless of the initial para-
sitoid egg number per host (Table 2). Further-
more, the number of eggs surviving to adulthood
declined as the initial parasitoid egg numbers per
host increased. Thus, superparasitism by D. inter-
medius represents a waste of resources and an im-
pediment to parasitoid population increase and,
ultimately, to biological control ofL. trifolii.

The authors wish to thank John Petti and Emily
Vasquez for their assistance in preparing this manu-
script. This research was supported by the Florida Agri-
cultural Experiment Station, and approved for
publication as Journal Series No. R-08473.

DICKE, M., AND O. P. J. M. MINKENBERG. 1991. Role of
volatile infochemicals in foraging behavior of the
leafminer parasitoid Dacnusa sibirica Telenga. J. In-
sect Behavior 4: 489-500.
HEINZ, K. M., AND M. P. PARRELLA. 1989. Attack behav-
ior and host size selection by Diglyphus begin on Li-
riomyza trifolii in chrysanthemum. Entomol. Exp.
Appl. 53: 147-156.
HEINZ, K. M., AND M. P. PARRELLA. 1990a. Holarctic dis-
tribution of the leafminer parasitoid Diglyphus be-
gini (Hymenoptera: Eulophidae) and notes on its life
history attacking Liriomyza trifolii (Diptera:
Agromyzidae) in chrysanthemum. Ann. Entomol.
Soc. Am. 83: 916-924.
HEINZ, K. M., AND M. P. PARRELLA. 1990b. The influence
of host size on sex ratios in the parasitoid Diglyphus
begin (Hymenoptera: Eulophidae). Ecol. Entomol.
15: 391-399.
MINKENBERG, O. P. J. M. 1989. Temperature effects on
the life history of the eulophid wasp Diglyphus isaea,
an ectoparasitoid of leafminers (Liriomyza spp.), on
tomatoes. Ann. Appl. Biol. 115: 381-397.
OATMAN, E. R., AND G. G. KENNEDY. 1976. Methomyl in-
duced outbreak of Liriomyza sativae on tomato.
J. Econ. Entomol. 69: 667-668.

FERRENTINO. 1989. Mass rearing ofDiglyphus begin
(Hymenoptera: Eulophidae) for biological control of
Liriomyza trifolii (Diptera: Agromyzidae). 82: 420-
PATEL, K. J. 1987. Parasitization of Liriomyza trifolii
(Burgess) by Diglyphus intermedius (Girault). Ph.D.
dissertation, University of Florida, Gainesville.
PATEL, K. J., AND D. J. SCHUSTER 1983. Influence of
temperature on the rate of development of Digly-
phus intermedius (Girault) (Hymenoptera: Eu-
lophidae), a parasite of Liriomyza spp. leafminers
(Diptera: Agromyzidae). Environ. Entomol. 12: 885-
PATEL, K. J., AND D. J. SCHUSTER 1991. Temperature-
dependent fecundity, longevity, and host-killing ac-
tivity of Diglyphus intermedius (Hymenoptera: Eu-
lophidae) on third instars of Liriomyza trifolii
(Burgess) (Diptera: Agromyzidae). Environ. Ento-
mol. 20: 1195-1199.
TINI, AND J. CASTNER 1996. Florida tomato scouting
guide with insect and disease identification keys.
Univ. of. Fla., IFAS, Coop. Ext. Serv., Gainesville, SP-
Adult experience modifies attraction of the leaf-
miner parasitoid Opius dissitus (Hymenoptera: Bra-

March 2003

conidae) to volatile semiochemicals. J. Insect
Behavior 5: 623-634.
PRICE, P. W. 1997. Insect Ecology Third Edition. John
Wiley & Sons, Inc., New York, NY.
SAS INSTITUTE, INC. 1982. SAS user's guide: statistics,
1982 edition. Cary, NC.
menopterous parasitoids of leaf-mining Liriomyza
spp. (Diptera: Agromyzidae) on tomato in Florida.
Environ. Entomol. 22: 1188- 1191.
P. R. SEYMOUR 1991. Agromyzidae (Diptera) leaf-
miners and their parasitoids in weeds associated
with tomato in Florida. Environ. Entomol. 20: 720-
1996. IPM in tomatoes, pp. 387-411. In D. Rosen, F.
D. Bennett and J. L. Capinera [eds.], Pest Manage-
ment in the Subtropics, Integrated Pest Manage-
ment-A Florida Perspective. Intercept. Ltd,
Andover, Hants, UK.
ESHLEMAN. 1980. Systems analysis of insect popula-
tion dynamics. Univ. of Fla., IFAS, Gainesville, IPM-
3, 92 pp.
WATERHOUSE, D. F., AND K. R. NORRIS. 1987. Liriomyza
species (Diptera: Agromyzidae) leafminers, pp. 159-
176. In Biological Control: Pacific Prospects. Inaka
Press, Melbourne, Australia.

Florida Entomologist 86(1)

Hall & Simms: Damage to Citrus by Texas Citrus Mite


Research Department, United States Sugar Corporation, P.O. 1207, Clewiston, Florida 33440


Studies were conducted during 1996-1999 to evaluate damage to citrus leaves by the Texas
citrus mite, Eutetranychus banksi (McGregor), and its impact on leaf longevity in irrigated
citrus. Natural mite infestations were followed in a citrus orchard (Rhode Red Valencia') un-
der irrigation management, and damage to leaves and leaf abscission were assessed period-
ically. The number of feeding stipples per cm2 on the upper leaf surface was used as an index
of feeding damage. A variable 'mite-days' [average number of mites per leaf multiplied by the
number of days of infestation] was used to characterize infestation densities over time. In-
creases in average numbers of stipples per cm2 per leaf (Y) across different mite-day values
(X) were described by the equation Y = 44.08 + 0.59X (r2 = 0.57). A model including temper-
ature was marginally better. The final mean density of feeding stipples on infested leaves for
the 1996, 1998 and 1999 evaluation periods averaged 327, 134 and 873 per cm2, respectively,
with an overall mean of 470. Leaf life from the date of full expansion until abscission aver-
aged 443, 387 and 380 days for the respective periods, with an overall average of 399 days.
The observed life of the leaves was typical to what has been observed in Florida citrus. Over-
all, no significant negative relationship was found between leaf life and mite damage. The
study indicated that damage by Texas citrus mites to'Valencia' citrus leaves promoted little
or no premature leaf abscission in irrigated trees.

Key Words: Citrus red mite, Panonychus citri, leaf abscission, damage assessment, Florida


Estos studios fueron conducidos durante 1996-1999 para evaluar el dano en las hojas de ci-
tricos causado por el acaro t6jano de citricos, Eutetranychus banksi (McGregor), y su impact
sobre la longevidad de las hojas de citricos en huertos irrigados. Las infestaciones naturales
de los acaros fueron observadas en un huerto de citricos de la variedad'Rhode Red Valencia'
bajo el sistema de irrigaci6n, y se evaluaron peri6dicamente el dano y la desprendimiento de
las hojas. El numero de picaduras por cm2 sobre la superficie superior de la hoja fu6 usado
como un indice del dano causado por la alimentaci6n. Una variable 'dias-de acaros' [el pro-
medio del numero de acaros por hoja multiplicado por el numero de dias de infestaci6n] fu6
usada para caracterizar la densidad de la infestaci6n sobre el tiempo. El aumento en el nu-
mero promedio de las picaduras por cm2 por hoja (Y) a trav6z de diferentes valores de "dias-
de acaros" (X) fu6 descrito por la ecuaci6nY = 44.08 + 0.59X (r2= 0.57). Un modelo incluyendo
la temperature fu6 ligeramente mejor. El promedio final de la densidad de las picaduras en
las hojas infestadas en los aios 1996, 1997 y 1999 fu6 327, 134 y 873 por cm2, respectiva-
mente, con un promedio total de 470. La vida de la hoja desde la expansion complete hasta
el desprendimiento dur6 un promedio de 443, 387 y 380 dias para los periods respectivos,
con un promedio total de 399 dias. La longevidad de las hojas observadas fu6 tipica de lo que
fu6 observado en los citricos de Florida. Sobretodo, no se encontr6 una relaci6n negative sig-
nificante entire la longevidad de la hoja y el dano hecho por los acaros. El studio indic6 que
el dano causado por el acaro t6jano de citricos a las hojas de citricos 'Valencia' promovi6 poco
o nada el desprendimiento premature de las hojas en arboles irrigados.

Of the four spider mite species (Acari: Tet- may live and (presumably) feed on either leaf
ranychidae) reported to infest Florida citrus, the surface (Muma 1961, Jones & Parrella 1984).
Texas citrus mite (Eutetranychus banksi Feeding by these spider mites on the upper leaf
(McGregor)) and the citrus red mite (Panonychus surface results in small, whitish or light-colored
citri (McGregor)) are the most prevalent and im- stipples within the palisade leaf layer where cy-
portant spider mite pests (Childers 1994). The toplasmic contents including chlorophyll are re-
Texas citrus mite lives and (presumably) feeds moved (Albrigo et al. 1981). When leaves are
almost exclusively on the upper leaf surface heavily damaged by spider mites, mesophyll col-
(Childers et al. 1991) while the citrus red mite lapse may occur and leaves may abscise prema-

Florida Entomologist 86(1)

turely, particularly during dry, windy conditions
(Browning et al. 1995, Hare & Youngman 1987).
Whether mite damage and/or premature leaf ab-
scission promotes economic losses in Florida cit-
rus has never been documented but considered
The stipple damage associated with spider
mite injury to the upper surface of leaves serves
as an index of the intensity of mite damage. Indi-
vidual stipples are small, ranging from around
0.04 to 0.52mm in diameter (mean 0.172mm,
SEM 0.025;'Valencia' leaves; damage by E. banksi
and/or P citri) (Hall, unpublished). Variation in
the size of individual stipples may be a function of
how long a mite feeds and the developmental
stage of a mite as well as other factors including
leaf age and leaf tissue characteristics. Whether
or not each individual stipple is always the result
of a single feeding wound is not known. Individual
stipples sometimes are so close to each other that
they coalesce, and as the density of stipples in-
creases, incidences of coalescence increase. Spider
mite damage to a leaf may not be apparent to the
naked eye until stipple densities exceed densities
of 100 to 150 stipples per cm2. In rating damage by
the naked eye, visual damage ratings of faint,
light, moderate and heavy damage may generally
be associated with stipple densities of around 200,
400, 1,000 and 1,800 stipples per cm2, respec-
Little is known regarding the quantitative re-
lationship between Texas citrus mite or citrus red
mite infestation densities over time and resulting
amount of leaf stippling damage to Florida citrus,
information which could be helpful in establish-
ing management guidelines. Economic thresholds
for the citrus red mite in California during the
1980s (2 to 4 adult female mites per leaf depend-
ing on the time of year and density of predatory
mites) were based largely on preventing excessive
stipple damage in the absence of more appropri-
ate information on the relationship between dam-
age and economic losses (Pehrson et al. 1984,
Hare & Youngman 1987, Hare et al. 1990).
Whether these thresholds are appropriate for pre-
venting excessive damage by infestations of Texas
citrus mites or citrus red mites in Florida citrus is
not known. Although premature leaf abscission
may be a major concern with damage by these
mites, particularly if trees with damaged leaves
are subjected to adverse environmental condi-
tions (e.g., drought and hot windy weather), quan-
titative data are lacking on the relationship
between mite damage and premature leaf abscis-
sion in Florida citrus.
Presented here are the results of quantitative
assessments of (1) the relationship between Texas
citrus mite infestations over time and resulting
damage to 'Valencia' citrus leaves and (2) the in-
fluence of Texas citrus mite damage on leaf abscis-
sion in 'Valencia' in irrigated citrus.


Four cohorts of 100 flush citrus leaves ('Rhode
Red Valencia') were studied during 1996-1999 at a
well-managed, irrigated orange grove on a flat-
woods (sandy spodosol) soil in Hendry County,
Florida. The full-expansion dates for leaves of the
four cohorts were approximately 1 September
1996 (trees 4.3 years old); 1 October 1997 (trees
5.4 years old); 1 October 1998 (trees 6.0 years old);
and 1 March 1999 (trees 6.5 years old). The trees
were planted on two-row beds with a tree spacing
of 3.7m and row spacing of 7.6m. For each cohort,
50 newly-expanded flush leaves were tagged
along the bed side of one row of trees and 50 were
tagged along the bed side of the adjacent row
along the same bed (one or two tagged leaves per
tree; in cases where two were tagged per tree,
these were 0.6 to 0.9m apart) All tagged leaves
were 0.6 to 1.8m above the ground and near the
outside of the canopy. Leaves of the 1999 cohort
were tagged along a bed next to the bed where the
1998 cohort of leaves was tagged. The length of
each leaf from the base (excluding the petiole) to
the tip and the width at the widest point of each
leaf were measured (cm). Leaf area (one surface)
in cm2 (Y) was estimated from leaf length (x,) and
width (X,) using the following equation: Y = 1.88 +
0.195(x,)2 + 0.487(x2)2, r2 = 0.94, n =100 (Hall, un-
Leaves were examined weekly to identify spi-
der mites present on the upper leaf surface. For
each species present, the number of spider mites
(excluding eggs) was counted. Within each cohort
of leaves, mite damage to 20 leaves was limited by
periodically wiping mites off with a soft damp
cloth or by misting them with either a 5% petro-
leum oil (FC-435-66) in water solution or a fenb-
utatin-oxide 50W treatment (0.6 g per 500ml
water). For each cohort of leaves, numbers of
mites per leaf were studied for 3 to 4 months, after
which weekly mite counts were discontinued and
all leaves were individually treated with the fen-
butatin-oxide treatment to eliminate mites.
Thereafter until abscission, each leaf was period-
ically treated with either the oil or fenbutatin-ox-
ide treatment to prevent further mite
In addition to counting mites on the upper sur-
face of leaves, damage by mites to the upper leaf
surface was quantified weekly on 52,40 and 50 in-
fested leaves for the 1997, 1998 and 1999 cohorts.
Early during the development of mite infestations
on the leaves of each cohort, few infested leaves
were available for damage evaluations. When
numbers of infested leaves increased to more than
20, we split the leaves into two groups and evalu-
ated them biweekly, one group evaluated one
week and the second group the following week.
When the number of infested leaves exceeded 40,
we split the leaves into three groups and evalu-

March 2003

Hall & Simms: Damage to Citrus by Texas Citrus Mite

ated damage to each group every three weeks, one
group per week. The average density of stipples
per cm2 across the upper leaf surface was used as
the measure of mite damage. Damage by mites to
individual leaves was assessed beginning on the
first day mites were observed on the leaves and
continued until all leaves of a cohort were treated
to eliminate mites. After all leaves had been
treated with fenbutatin-oxide to eliminate mites,
a final estimate of the average stipple density was
made for every leaf within each cohort. For the
1996 cohort of leaves, the average density of stip-
ples on each leaf was estimated only after appre-
ciable damage had occurred to leaves and mites
had been controlled. For all cohorts, estimates of
average stipple densities per leaf were made by
counting the number of stipples per cm2 at each of
10 sites uniformly spaced across the upper sur-
face of each leaf. To count stipples, an Edmond
Scientific comparator (12X transparent base mag-
nifier with 27mm contact reticule adapter ring,
Kellner-type/AR coated lens, reticule with 1-cm2
grid of 100 squares, Edmond Optics, Barrington,
NJ USA) was placed against or just above the leaf
surface. The leaf and magnifier were held so that
as much light as possible illuminated the leaf sur-
face being examined. All stipples within the grid
were counted when there were less than approxi-
mately 100 stipples within the grid; however,
when there were more than approximately 100
stipples within the grid, the number of stipples
within a sub-sample often squares of the grid was
counted and multiplied by ten to estimate the to-
tal number per cm2. At high stipple densities per
cm2 (e.g., 1,500 or more), stipples sometimes coa-
lesced, making it difficult to estimate the actual
number present. In this case, the number of indi-
vidual stipples constituting an area of coalesced
stipples had to be estimated based on the average
diameter of surrounding or nearby individual
stipples. Leaves with dust or sooty mold were gen-
tly cleaned using a soft cloth dampened with wa-
ter before counting stipples (the entire surface
when no mites were present or 10 spots each ap-
proximately 1-cm2 in size if mites were present).
For final stipple counts, the entire upper surface
of each leaf was gently cleaned with water, after
which the magnifier was placed against the wet
leaf surface.
Records were maintained for each leaf of each
cohort on the incidence of injury by citrus leaf-
miner (Phyllocnistis citrella Stainton) and citrus
rust mites (Pyllocoptruta oleivora (Ashmead) and
Aculops pelekassi (Keifer)), infection by greasy
spot .I .... .I....... !!.' citri Whiteside), nutritional
disorders, mesophyll collapse, freeze damage and
hail injury. The percentage surface area infected
by greasy spot was estimated for 1998 leaves on
28 May 1999 and for 1999 leaves on 17 February
2000. Air temperature, rainfall, evaporation and
wind data during the study period were obtained

from the Corp of Engineer's Moore Haven Lock 1
weather station about 4.8 km north of the study
sites. Exceptional environmental events during
the study were noted.

Relationship Between Mite Density and Damage

The quantitative relationship between spider
mite densities per leaf (upper surface) and result-
ing damage to the upper surface of leaves was in-
vestigated by comparing the average mite density
on a leaf over a period of time to the increase in
average stipple density over the same period of
time (data from cohorts 1997, 1998 and 1999). For
each infestation period on each leaf, the average
number of mites per leaf was calculated by aver-
aging the numbers of mites observed on different
observation dates during the infestation period.
The duration of mite infestation (days) was deter-
mined by the number of days between the first
and last observation dates during the infestation
period. An infestation density/duration variable
'mite days' (see Allen 1976, Yang et al. 1995) was
calculated for each infestation on each leaf:'mite
days' = (average number of mites per leaf) (num-
ber of days). The resulting damage caused by
mites feeding during each infestation period on
each leaf was estimated by subtracting the mean
number of stipples per cm2 present at the begin-
ning of the period from the mean number present
at the end of the period. A linear regression anal-
ysis was conducted between the increase in mean
stipple densities per cm2 per leaf and 'mite days'
for leaves of each cohort and over all three co-
horts. Correlation analyses were conducted be-
tween the following variables: increases in stipple
densities; 'mite days'; leaf area; mean, maximum
and minimum daily air temperatures; daily rain-
fall; daily evaporation; and daily wind. Stepwise
regression analyses were then conducted using
variables significantly correlated (P < 0.05) with
increases in damage to select a multiple regres-
sion model for predicting damage.

Relationship Between Mite Damage and Leaf Abscis-

The leaves of three cohorts (1996, 1998 and
1999) were examined every 2 to 5 weeks (mean
20.7 days, SEM 2.1 days) after mite infestations
were controlled to determine when the leaves ab-
scised. The abscission date was estimated using
the mid date between the date abscission was dis-
covered and the date a leaf was last observed on a
tree.'Leaf life' was approximated as the period of
time from the date of full expansion until the ab-
scission date. 'Life after attack' by mites was ap-
proximated as the period of time between a leaf's
mean infestation date (weighted on infestation
densities across successive infestation dates) and
its abscission date.

Florida Entomologist 86(1)

March 2003

To determine if spider mite damage promoted
premature abscission, linear regression analyses
were conducted between 'leaf life' and damage
(average number of stipples per cm2); and be-
tween 'life after attack' and damage (average
number of stipples per cm2). Analyses on 'life after
attack' were restricted to leaves on which mites
were observed on at least three successive sample
dates. Leaves with disorders such as damage by
other arthropod pests, nutritional problems,
freeze damage and hail injury were excluded from
all analyses. Correlation analyses were conducted
to investigate the relationship between 'leaf life'
and each of the following variables: average num-
ber of stipples per cm2; mean surface area with
greasy spot infection; mean, maximum and mini-
mum daily air temperature; daily wind; daily
rainfall; daily evaporation; and interaction effects
between mite damage and each of the other inde-
pendent variables. For each date on which leaf ab-
scissions were discovered (i.e., an abscission
event had occurred), the percentage of abscised
leaves within each cohort was calculated. Corre-
lation analyses were then conducted to investi-
gate the relationship between percentage leaf
drop and the aforementioned variables. For all
correlation analyses, air temperature, wind, rain
and evaporation data were averaged over a period
of within 40 days prior to discovering leaf abscis-
sion. For variables which were significantly corre-
lated (Pr > I r I < 0.05) with 'leaf life' and 'life after
attack', stepwise regression analyses were con-
ducted to select regression models for predicting
the life of leaves damaged by spider mites.


Both Texas citrus mites and citrus red mites
were observed on leaves during this study, with
Texas citrus mites being the predominant species
(Table 1). Spider mites densities were greater on
the 1996 and 1999 leaf cohorts (e.g., means of 18.8
and 21.9 Texas citrus mites per leaf, respectively)
than on the 1997 and 1998 cohorts (e.g., means of
5.1 and 4.0 Texas citrus mites per leaf, respec-
tively) (Table 1). Among the three fall flush co-
horts, mite densities on leaves during the first
several months after leaf expansion were greater
during 1996 than either 1997 or 1998 and, based
on correlation analyses (not presented), these in-
festation density differences were attributed to
less rainfall during these months in 1996 based
on rainfall at the weather station. This is consis-
tent with Pratt & Thompson (1953), who previ-
ously reported a negative relationship between
rainfall and citrus spider mite levels. During each
of the four flushes investigated in this study, spi-
der mite infestations generally began to develop
within 1 to 3 months after leaves had fully ex-
panded (Fig. 1). Leaves of the 1998 cohort, which
were present on trees and being monitored when




i -

0 ^




-^ 0

0 co






n o

1-hi C11 1-i It cc X O X c C11 t_
C 1c CC. C C O -' 1 1 (
6^ C6 6^ 6- ci ^ CO cIO C

(M l hI (M (M
1_ C11^ W' ^ C11'C ^ O C11

CTi i o s co o o- c^ o o T-iT c -^
c C'i 00' (0i ( C- Ic CX C.1i t -

C.cfl l1f CIC

0j cc 0u

Zo g" Z
I~i I
a c0

m m
c o cu c o
+ C O
000 000-* O -

01 (0


o o C

Hall & Simms: Damage to Citrus by Texas Citrus Mite







200 -

150 -

100 -

50 -

'1/96 10/1/96 11/1/96 12/1/96

1/1/97 2/1/97 3/1/97

Fall flush 1997
Leaves fully expanded by 10/01/97
treated with miticide 3/24/98

. ..... .. ... .. .. ....... 1 . .....Q i .Q.i.i .i ...................i ..i. C -i ......... .i i .i n....... i....... ... ... ..

10/1/97 11/1/97 12/1/97

1/1/98 2/1/98 3/1/98

4/1/98 5/1/98

Fall flush 1998
DO Leaves fully expanded by 10/01/98
treated with miticide 4/15/99
50 -



10/1/98 11/1/98 12/1/98 1/1/99 2/1/99 3/1/99 4/1/99 5/1/99

0 -3/1

4/1/99 5/1/99 6/1/99 7/1/99 8/1/99 9/1/99 10/1/99

Fig. 1. Spider mite infestations (densities per leaf and dates) observed during the study (Texas citrus mite den-
sities solid data points, citrus red mite densities open data points).

Fall flush 1996
Leaves fully expanded by 9/01/96

A freeze on 1/19/97
terminated the infestation

iill .; /


Florida Entomologist 86(1)

research was initiated on the 1999 cohort, were
infested by lower densities of spider mites than
the leaves of the 1999 cohort during March and
April 1999. The reason mites were more abundant
on the younger leaves was not known but could
have been related to factors such as microclimate,
leaf nutrition or biological control.

Relationship Between Mite Density and Damage

Since only a few citrus red mites were observed
during this study, analyses on the quantitative re-
lationship between mite infestations and result-
ing damage were restricted to leaves known to be
infested solely by Texas citrus mites. Data for a to-
tal of 131 individual leaf infestations of Texas cit-
rus mites were subjected to analyses comparing
leaf infestation densities/durations to resulting
damage, with data for 58, 34 and 39 individual in-
festations from the 1997, 1998 and 1999 cohorts,
respectively. Overall, these infestations averaged
21.9 days in duration (SEM 0.5 days) at a mean
density of 7.8 Texas citrus mites per leaf(SEM 1.3
A statistically significant relationship was
found between infestations ('mite days') of Texas
citrus mites and resulting increases in densities
of stipples per cm2 for each of the 1997, 1998 and
1999 cohorts. Over all 3 cohorts, increases in the
mean density of stipples per cm2 per leaf (Y) were
related to'mite days' (X) by the equation Y = 44.08
+ 0.59 X (r2 = 0.57, F = 168.8, Pr > F = 0.0001, df
130, slope SEM = 0.045) (Fig. 2). The intercept pa-
rameter 44.08 (significantly greater than zero, t =
2.7, Pr > t 0.008) reflected the presence of stipples
which could not be attributed directly to mites ob-
served on leaves. This may have been a result of
mites moving from leaf to leaf or being subjected
to mortality factors prior to leaf observations.

0 200 400 600 800 1000 1200 1400 1600
Mitedays (Mean number of mites per day number of days)
Fig. 2. Relationship between increases in feeding
damage (mean stipple density per cm2 per leaf) and mite
days (mean number mites per leaf times the number of
days of infestation) on leaves infested by Texas citrus
mites, data from 1997-1999 combined.

Stepwise regressions indicated that increases in
the mean density of stipples per cm2 per leaf (Y)
were best described by a multiple regression
model based on 'mite days' (X,) and maximum
daily air temperatures at the weather station (X2):
Y = -414.6 + 0.516X,+ 17.9X,, r2 = 0.60, F = 94.4,
Pr > F = 0.0001, df 130. Texas citrus mites there-
fore caused more damage as temperature in-
creased. The correlations between observed and
estimated increases in damage were similar with
respect to the multiple regression model (R =
0.77, Pr > I RI = 0.0001) and the simple model
based only on 'mite days' (r = 0.75, Pr > Irl =
0.0001). A strong statistical relationship existed
between estimates from the simple model (Y) and
the multiple regression model (X): Y = 6.89 +
0.95X, r2 = 0.95, F = 2,512.8, Pr > F = 0.0001, df
130. Based on the slope 0.95, the inclusion of max-
imum daily temperatures only marginally im-
prove estimates across the temperatures
observed during our study. Over all three cohorts,
leaf area and increases in mite damage were neg-
atively correlated, but regression analyses indi-
cated leaf area was not a significant variable in
predicting damage. The leaves studied were fairly
uniform in size, with leaf area averaging 40.5 cm2
(SEM = 1.2, n = 127). It remained probable that a
given density of mites would cause more damage
over a given period of time to small leaves than to
large leaves.

Relationship Between Mite Damage and Leaf Abscis-

Leaf disorders observed during the study in-
cluded damage by citrus rust mites (species not
identified), leaf miners and some other leaf-feed-
ing insects, greasy spot disease, freeze damage,
and hail injury. Although little damage by rust
mites occurred during the study, two leaves of the
1996 cohort and one leaf of the 1999 cohort were
dropped from leaf life assessments due to rust
mite injury. Two leaves of the 1998 cohort were
dropped from life assessments due to nutritional
problems (yellowing), and 15 leaves of this cohort
were dropped from life assessments due to dam-
age by leaf-feeding insects. Among leaves within
the 1996 cohort, 62 were damaged by a freeze on
19 January 1997 (temperatures as low as around
-5C for several hours at the weather station,
probably colder at the study site); 48 of these
leaves abscised within several days following the
freeze and 14 others were rendered unfit for fur-
ther research, leaving 36 leaves for life assess-
ments. Among these 36 leaves, 13 had suffered
considerable spider mite damage (e.g., averages
in excess of 1,000 stipples per cm2) before the
freeze, suggesting that a freeze will not necessar-
ily promote immediate abscission of leaves with
mite damage. Among leaves of the 1998 and 1999
cohorts, 37 and 20 leaves, respectively, were ren-

March 2003

Hall & Simms: Damage to Citrus by Texas Citrus Mite

dered unfit for leaf life assessments due to hail
damage suffered on 28 May 1999. Twelve leaves of
the 1999 cohort abscised or were hedged off before
a final estimate of spider mite damage was made
and were thus not available for leaf life assess-
ments. Fifty-one of the 67 remaining leaves of the
1999 cohort were selected for leaf life assessment
studies. None of the leaves studied developed any
signs of mesophyll collapse. In spite of two stan-
dard summer treatments of copper and petroleum
oil used for greasy spot control each summer, low
infection levels of greasy spot developed on at
least some leaves in each cohort. Usually, only one
to several small infection sites could be found on
any leaf with greasy spot. All leaves with greasy
spot were therefore retained for leaf life assess-
Data on spider mite infestations and damage,
greasy spot infections, and leaf longevity are pre-
sented in Table 2. The life of all leaves studied,
from full expansion to the abscission date, is de-
picted in Fig. 3. A total of only four leaves studied
were known to have been infested solely by citrus
red mites (Table 2). Among 45 leaves known to
have been infested by both Texas citrus and citrus
red mites, averages of 5.7 (SEM = 1.2) Texas cit-
rus mites and 0.9 (SEM = 0.3) citrus red mites per
leaf were observed. Data from leaves with citrus
red mites were retained for all analyses, however,
since so few citrus red mites were observed, con-
clusions from the data regarding the influence of
mite damage on leaf abscission may only be appli-
cable to Texas citrus mites.
An average leaf life of 399 days was estimated
across all leaves (n = 133), with averages of 443,
387 and 380 days for the 1996, 1998 and 1999 co-
horts, respectively (Table 2). The maximum life
expectancy of the leaves appeared to be 18 to 20
months (Fig. 3). An average density of 470 stip-
ples per cm2 per leaf was estimated across all
leaves studied, with averages of 327, 134, and 873
stipples per cm2 per leaf for the 1996, 1998 and
1999 cohorts, respectively. Within the 1996 co-
hort, leaf life (Y) decreased as mite damage (X) in-
creased (Y = 489.7-0.144X; F = 9.47, Pr > F =
0.0041; df= 35), but the relationship between leaf
life and damage was statistically weak (e.g.,
r2 = 0.22) (Fig. 4). No negative relationship was
found between leaf life and spider mite damage
among leaves of either the 1998 or 1999 cohorts
(Fig. 4). Statistical analyses on the combined data
from the three cohorts indicated no significant re-
lationship between leaf life and mite damage. The
1996 cohort of leaves was subjected to a hard
freeze during January 1997 and, although the
leaves followed until abscission did not exhibit
any signs of freeze damage, this adverse temper-
ature event may have contributed to the short-
ened life of leaves with mite damage. Other
unknown factors may have contributed to the re-
duced life of these leaves with mite damage. Al-

though the life of leaves within the 1996 cohort
tended to shorten as mite damage increased, the
average life of these leaves was longer than the
life of leaves of the other two cohorts.
In addition to being subjected to a hard freeze
on January 19, 1997, the 1996 cohort of leaves
was subjected to near-freezing (1 to 4C at the
weather station) air temperatures for a short pe-
riod of time one day during April 1997 (no leaf
drop occurred during this month), on one day dur-
ing December 1997 (20% of this cohorts leaves
dropped during December), and on several days
during January, February and March 1998 (some
leaves of this cohort dropped during January and
February of 1998, and 25% dropped during March
1998). The 1998 cohort of leaves was subjected to
near-freezes on several days during December
1998, February 1999, March 1999, and January
2000 (none of the Fall 1998 cohort leaves dropped
during any of these 4 months). The 1999 cohort of
leaves was subjected to near-freezes on several
days during March 1999 and January 2000 (none
of this cohort's leaves dropped during March 1999
and few dropped during January 2000). Overall,
near-freezing temperatures for short periods of
time did not appear to promote premature abscis-
sion of leaves whether they were damaged by
mites or not.
The 1998 and 1999 cohorts of leaves were sub-
jected to two notable wind events, the first event
during October 1999 associated with Hurricane
Irene on 15 -16 October (23.7 and 25.5 mph aver-
age daily wind speeds, respectively, at the
weather station) and the second event on one day
during January 2000 (16.7 mph average daily
wind speed at the station). These wind events did
not result in an immediate drop of any leaves.
A total of 40 leaf abscission events (21, 11 and
8 for the 1996, 1998 and 1999 cohorts, respec-
tively) were subjected to correlation analyses be-
tween leaf life (days after fully expanded to the
abscission date), damage stipplee density per leaf)
and environmental variables. Only 5 events could
be subjected to correlation analyses with greasy
spot infections for 1999 data because three leaves
of this cohort abscised before greasy spot ratings
were made. Analyses over all data indicated that
leaf longevity was not correlated with the amount
of mite damage to a leaf (Table 3). A significant
negative correlation (i.e., Pr > I r I < 0.05) was
found between leaf life and mite damage for the
1996 cohort of leaves (r = -0.59, Pr > I r I = 0.01)
but not for either the 1998 or 1999 cohorts nor
over all 3 cohorts combined. A significant negative
correlation was found between leaf life and inci-
dence of greasy spot. No significant correlations
were found between leaf longevity and any of the
environmental variables. Analyses on 'leaf life af-
ter attack' indicated that, for the 1996 cohort, life
after damage decreased as mite damage in-
creased (Fig. 5). However, the relationship be-

Florida Entomologist 86(1)

March 2003

0 4 C6

L-- U, In q
CO ( (
C 0^ i U^ h 5
^ 101^ 0 O

'.-^ So'o -

SC ci10 (0 C"
co ci clCZ C cx

0 a-

0 -"


_ C


0 m




0. o

0 f

- 0







o ,^

^ o

ci u
'7 (0
'~ 0 C
c-3 c3 [

C6 C6 6 Ci

C~i C~i COCO
In CO cc x CO
cc U, x' C
x x It xC`

C) := c q ci 'c

00 Q00 ; D 00 CTi 0 C0O 0q (M 0 CT
L-- : Cq x00 C) C) I

I? Cq

ci CO CO C
Cq CqUCq

In 6d co
C" c
0 cm -
'"0 0

:= Co I ci cX

(0 0' (0 '
CO '-- (-l 0 CO
,- ,CI ci ,l


Oq Q
i i
I, 0
c-l '^ T-
cI CO [-1

co o~co i oci

C0O C -i
0(0 cc (0
In 1i0 10

(0 ^ C
'c 0 '
[-c [
dcd (

CT^ C^I .I -' (0 .I C^I In CO (0 .I 0 CT^ .-I .-I -' In CO CO

0-- -- CO -- -I -- C -l 1 0 0^ 0 ^
coh iCD ~ h s0 co c o co ~~u
'3 ^ '3 ^ 'i 3 us C ~'h 03

*a o ,- o ^" a ^ i^ "

m ~ ~ ~ c ' "c 'a m' S' a m "S
cac g s- cca g cca g ca
?.a E
ccm~ 0om~ o~m~ o~m~
$UF4~ ~U40 $F90 $F

0000o 0



S 0 0 o,
m- sA A^
ai V


~- C- 3-

- o ( '3'o
i I.I E c E
J) c *3 *

a g L s
E- C']

bt bt bt
^ .2a v

9I 58 3 -S-S

Hall & Simms: Damage to Citrus by Texas Citrus Mite

S55 Fall flush 1996
,. 50 -leaves fully expanded on 9/1/96
45 -36 leaves total
-o 40 -
w 35
> 30
5 25 Hard freeze on 1/19/97
20 -
,) 10-
0- 5 -

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 192021 22 23

Sep'96 Dec'96 Mar'97 Jun'97 Sep'97 Dec'97 Mar'98 Jun'98

co 55 -
' .
50 -
- 45-
- 40-
cn 35-
> 30-
i 25-
-- 20
e 15-
n 5-

0) 55
- 45i
S40 -
C, 35
> 30
j 25-
- 20-
0 15
) 10

Fall flush 1998
leaves fully expanded on 10/1/98
46 leaves total

11 I- .ii r

1 2 3 4 5 6 7 8 9 1011 12131415161718192021 2223

Oct'98 Jan'99 Apr'99 Jul'99 Oct'99 Jan'00 Apr'00 Jul'00

Spring flush 1999
leaves fully expanded on 3/1/99
51 leaves total


1 2 3 4 5 6 7 8 9 1011 12131415161718 192021 22 23
Mar'99 Jun'99 Sep'99 Dec'99 Mar'00 Jun'00 Sep'00 Dec'00

Month after leaf fully expanded

Fig. 3. Percentage abscission of citrus leaves over time.

Florida Entomologist 86(1)

Fall flush 1
mean leaf I


350 1

200 -

life = 442 days
le density per cm2 per leaf = 327


L mean stipp

____* _

Fall flush 1998
mean leaf life = 386 days
mean stipple density per cm2 per leaf = 134


- **
Y = 368.7 + 0.132X, r2 = 0.03
F = 1.51, Pr >F = 0.226, d.f. = 45

Spring flush 1999
mean leaf life = 379 days
mean stipple density per cm2 per leaf = 873


* *0

Y = 363.4 + 0.018, r2 = 0.08
F = 4.50, Pr>F = 0.039, d.f. = 50





Average number stipples per cm2 per leaf

Fig. 4. Relationship between leaf life (days from full flush expansion until abscission) and spider mite damage
(mean number of stipples per cm2).

Y = 489.7 0.144X, r2 = 0.22
F = 9.47, Pr>F = 0.004, d.f. = 35


March 2003

Hall & Simms: Damage to Citrus by Texas Citrus Mite


Correlation between leaf life and
the indicated variable

Variable r Pr> IrI

Mean number (#) stipples per cm2 -0.13" 0.42
Mean percent surface area per leaf with greasy spot -0.58b 0.02
Mean daily air temperature ( C) -0.15" 0.37
Mean minimum daily air temperature ( C) -0.08" 0.63
Mean maximum daily air temperature ( C) -0.22" 0.17
Mean daily wind (k) 0.21 0.19
Mean daily evaporation (cm) 0.03" 0.88
Mean daily rainfall (cm) 0.07" 0.66
Mean # stipples per cm2 X mean pct area per leaf with greasy spot -0.34b 0.20
Mean # stipples per cm2 X mean daily air temperature ( C) -0.06 0.71
Mean # stipples per cm2 X mean minimum daily air temperature ( C) -0.02 0.91
Mean # stipples per cm2 X mean maximum daily air temperature ( C) -0.08 0.61
Mean # stipples per cm2 X mean daily wind (k) 0.04 0.81
Mean # stipples per cm2 X mean daily evaporation (cm) -0.02 0.92
Mean # stipples per cm2 X mean daily rainfall (cm) 0.10 0.56

n = 40.
bn = 16.

tween damage and 'life after attack' for the 1996
leaves was weak (r2 = 0.16), and no significant
negative relationship between these variables
was found among leaves of either the 1998 or
1999 cohorts. Further, leaves of the 1996 cohort
stayed on trees longer after being damaged by
mites than did leaves of the other two cohorts.
The longevity of leaves damaged by infestations
of the Texas citrus mite, or by infestations of Texas
citrus mites in combination with low levels of cit-
rus red mites, was similar regardless of the
amount of damage by mites (Fig. 6). The research
indicated that, over all leaves evaluated in this
study, Texas citrus mite damage promoted little or
no premature abscission of citrus leaves. Also,
damage resulting from low levels of citrus red
mites in combination with infestations of Texas cit-
rus mites did not decrease the longevity of citrus
leaves during this study. No conclusions could be
made from this study about the effect of extensive
citrus red mite injury on leaf longevity. Thompson
et al. (1954) observed mesophyll collapse in June
following large outbreaks of the citrus red mite in
April and May and reported that heavy leaf drop
by citrus trees during late winter may sometimes
be promoted in Florida by citrus red mite infesta-
tions (no data presented). Some scions may be
more sensitive to damage by Texas citrus mites
than 'Valencia' (e.g., 'Sunburst' mandarin, see Al-
brigo et al. 1987), and premature leaf abscission
associated with mite injury may be more likely to
occur in these scions even in an irrigated orchard.

Infestations of Texas citrus mites occur prima-
rily on the upper surface of leaves and, conse-
quently, feeding injury by these mites may occur
primarily on this leaf surface. The Texas citrus
mite could be a more important pest if it fed on
the lower leaf surface. McCoy (1976) found that
injury by citrus rust mites (species not indicated)
to the lower surface of leaves promoted more me-
sophyll collapse and leaf drop than damage by the
mite to the upper leaf surface. For this same rea-
son, the citrus red mite may be a more important
pest than the Texas citrus mite, as this mite in-
fests both the upper and lower leaf surfaces
(Jones & Parrella 1984). Rust mite damage to the
upper leaf surface may promote less water loss
from a leaf than damage to the lower leaf surface
because the upper surface lacks stomates, has a
highly developed waxy layer, and has a compact
palisade parenchyma layer of cells beneath the
epidermis that contribute to the prevention of wa-
ter loss (McCoy 1976). Therefore, the effect on wa-
ter loss of damage by Texas citrus mites to the
upper leaf surface may also be less. Based on re-
search by McCoy (1976), the ultimate cause of
premature defoliation of citrus leaves is water
loss. Although injury by mites may promote water
loss, a good water supply for trees may help pre-
vent premature abscission of leaves damaged by
mites. Working in trees with an overhead water-
ing system rarely used during the winter, McCoy
(1976) speculated that scant rainfall (0.5 cm per
week) promoted premature abscission of leaves

Florida Entomologist 86(1)







0 200 400 600 800 1000 1200 1400 1600 1800

o 500

| 300-

to 200 -

S 100

0 100












Average number stipples per cm2 per leaf

Fig. 5. Relationship between leaf life after mites damaged leaves (days from damage until abscission) and spider
mite damage (mean number of stipples per cm2).

Fall flush 1996, n=29
Y *= Y= 365.5-0.12X, r2=0.16
* -F=5.12, Pr>F=0.03


Fall flush 1998, n=13
Y = 207.2+0.14X, r2=0.03
F=0.37, Pr>F=0.56


* *

S 0

Spring flush 1999, n=48
Y = 319.5+0.02X, 2=0.06
F=2.99, Pr>F=0.09

March 2003



Hall & Simms: Damage to Citrus by Texas Citrus Mite

400 *
30 ** *

0 0*

Texas citrus mites, n=49
Y = 328.3-0.03X, r2=0.05
F=2.46, Pr>F=0.12

0 30 600 900 1200 1500 1800 2100 2400 2700 3000
0 300 600 900 1200 1500 1800 2100 2400 2700 3000

Texas and red mites, n=37
overall average of 6 Texas mites and 1 red mite per leaf
Y = 305.3+0.03X, r2=0.03
F=0.97, Pr>F=0.33


0 300 600 900 1200 1500 1800 2100 2400 2700

Average number stipples per cm2 per leaf

Fig. 6. Leaf life (days) after attack by mites and damage, leaves infested by Texas citrus mites or both Texas cit-
rus and citrus red mites.

damaged by rust mites and that increased water
loss from leaves through mite feeding damage to
the lower leaf surface may be enough during dry
periods to cause leaf abscission. The equivalent of
an average of 2.4, 2.5 and 1.2 cm of daily rain in
the general vicinity of our study site was associ-
ated with periods of time we observed leaf drop
among the 1996, 1998 and 1999 cohorts, respec-
tively, considerably more than 0.5 cm per week.
But for each cohort, there were one or two 40-day
periods during which weekly rainfall at the
weather station averaged less than 0.5 cm, and
for the 1999 cohort there was one 40-day period
during which no rainfall was recorded. Increases
in leaf abscission at the study site were not ob-
served during these dry periods. Irrigation during
dry periods may have helped prevent premature
abscissions of citrus leaves with mite damage

and, therefore, the results of our study may only
pertain to irrigated trees.
Healthy citrus leaves can remain on a tree for
2 to 3 years or longer (Kelley & Cummins 1920,
Davies & Albrigo 1994). Disease and pest pres-
sures as well as low light levels can significantly
reduce leaf longevity (Davies and Albrigo 1994).
In a California study, the majority of orange
leaves abscised by 17 months and almost all by 24
months (Wallace et al. 1954). Whiteside (1982)
speculated that, in the absence of freezes and
greasy spot disease, the expected life of citrus
leaves in Florida may be 1 to 2 years, similar to
what was observed in our study. This supports the
conclusion that damage by Texas citrus mites had
little influence on leaf longevity under our study
conditions. A study of mature 'Valencia' trees indi-
cated that leaf abscission may occur all year long,

500 -


100 *o
I .


- ~ ( 40. p
~-+--0 0

500 -

400 -


200 -




with higher abscission rates during late Septem-
ber-November and mid-April to May (Erickson &
Brannaman 1960). The greatest abscission rate in
citrus normally occurs during the spring flower-
ing period (Erickson 1968). Whiteside (1982) re-
ported that the major seasonal leaf drop from
Florida citrus trees begins after the spring growth
flush has emerged (usually in March) and may ex-
tend until late-May. During our study, abscission
of leaves of the spring 1999 cohort generally fol-
lowed Whiteside's seasonal leaf drop profile while
abscission of leaves of the fall 1996 and 1998 co-
horts generally did not (Fig. 3).
Based on the study, the amount of physical
damage to citrus leaves resulting from an infesta-
tion of Texas citrus mites can be projected based
on duration of mite densities. If the economic im-
portance of damage by the mite was known, such
projections could be useful in making mite control
decisions. Whether or not feeding injury by Texas
citrus mites to leaves results in economic losses
remains to be determined. Although premature
leaf abscission could be an important economic
problem associated with damage by some pests,
our study indicated it was not important with re-
spect to Texas citrus mites in an irrigated citrus


We thank the following individuals for their contri-
butions to this project: Carol J. Lovatt (University of
California at Riverside), Gene Albrigo, Carl Childers
and Clay McCoy (Citrus Research & Education Center,
University of Florida, IFAS, Lake Alfred).


1981. Structural damage to citrus leaves from spider
mite feeding. Proc. Int. Soc. Citriculture. 2: 649-652.
1987. Observations of cultural problems with the
'sunburst' mandarin. Proc. Florida State Hort. Soci-
ety. 100: 115-118.
ALLEN, Jon C. 1976. A model for predicting citrus rust
mite damage on Valencia orange fruit. Environ. En-
tomol. 5(6): 1083-1088.
V. CALVERT, AND W. F. WARDOWSKI. 1995. Florida
citrus diagnostic guide. Fla. Sci. Source, Inc., Lake
Alfred, FL 244 pp.
1991. Biology of Eutetranychus banksi: life tables on
'Marsh' grapefruit leaves at different temperatures
(Acari: Tetranychidae). Internat. J. Acarol. 17: 29-35.

March 2003

CHILDERS, C. C. 1994. Biological control of phytopha-
gous mites on Florida citrus utilizing predatory ar-
thropods. In Pest Management in the Subtropics,
Biological Control-A Florida Perspective. eds.
D. Rosen, F. D. Bennett and J. L. Capinera. Inter-
cept Ltd., U. K. 737 pp.
International, Oxford, UK. 242 pp.
ERICKSON, L. C. 1968. The general physiology of citrus.
In The Citrus Industry, Vol. II: Anatomy, Physiol-
ogy, Genetics and Reproduction. eds W. Reuther, L.
D. Batchelor and H. J. Webber. University of Califor-
nia. 398 pp.
ERICKSON, L. C., AND B. L. BRANNAMAN. 1960. Abscis-
sion of reproductive structures and leaves of orange
trees. Proc. Amer. Soc. Hort. Sci. 75: 222-229.
HARE, J. D., AND R. R. YOUNGMAN. 1987. Gas exchange
of orange (Citrus sinensis) leaves in response to feed-
ing injury by the citrus red mite (Acari: Tetrany-
chidae) on lemons in southern California. J. Econ.
Entomol. 80: 1249-1253.
MEYER. 1990. Effects of managing citrus red mite
(Acari: Tetranychidae) and cultural practices on to-
tal yield, fruit size and crop value of 'Navel' orange.
J. Econ. Entomol. 83: 976-984.
JONES, V. P., AND M. P. PARRELLA. 1984. Intra-tree re-
gression sampling plans for the citrus red mite (Ac-
ari: Tetranychidae) on lemons in southern
California. J. Econ. Entomol. 77: 810-813.
KELLY, W. P., AND A. B. CUMMINS. 1920. Composition of
normal and mottled citrus leaves. Journal of Agric.
Research. 20: 161-191.
MCCOY, C. W. 1976. Leaf injury and defoliation caused
by the citrus rust mite, Phyllocoptruta oleivora.
Florida Entomol. 59(4): 403-410.
MUMA, M. H. 1961. Mites associated with citrus in Flor-
ida. Unv. Fla. Agr. Exp. Sta., Bulletin 640. 39 pp.
P. A. PHILLIPS. 1984. Integrated pest management
for citrus. Unv. Calif. Publ. 3303. 144pp.
PRATT, R. M., AND W. L. THOMPSON. 1953. Spray pro-
grams, varieties and weather conditions in relation
to six-spotted mite and purple mite infestations.
Proc. Fla. Soc. Hort. Sci. 66: 65-69.
1954. The status of the purple mite and its control.
Proc. Fla. Soc. Hort. Sci. 67: 50-56.
NORTH. 1954. Translocation of nitrogen in citrus
trees. Proc. Amer. Soc. Hort. Sci. 64: 87-104.
WHITESIDE, J. O. 1982. Effect of temperature on the
development of citrus greasy spot. Proc. Florida
State Hort. Soc. 95: 66-68.
1995. Relationship between population density of
citrus rust mite (Acari: Eriophyidae) and damage to
'Hamlin' orange fruit. Environ. Entomol. 24(5):

Florida Entomologist 86(1)

Barry et al.: Potential for Lower Overflooding Ratios


'Department of Entomology, University of California, Riverside California 92521

2USDA-APHIS, P.O. Box 1040, Waimanalo, HI 96795 and Hawaiian Evolutionary Biology Program
University of Hawaii, Honolulu, Hawaii 96822.

3USDA-ARS, PBARC, 2727 Woodlawn Drive, Honolulu, Hawaii 96822.


The mating behavior of sterile, laboratory-reared, male Mediterranean fruit flies, Ceratitis
capitata (Wiedemann), was evaluated in field-cage mating competition tests with wild flies
after exposing the laboratory males to ginger root oil extract. Without exposure to ginger
root oil, sterile males obtained 12.6%, 69.0%, and 72.8% of the total matings with wild fe-
males when present in 1:1, 5:1, and 10:1 ratios of sterile males to wild males, respectively.
Sterile males, exposed to ginger root oil for 3 h, 1 d before mating trials, in a 1:1 ratio with
wild males, achieved 62.3% of the matings with wild females. These data suggest that expo-
sure to ginger oil can elevate sterile male mating competitiveness to a similar degree as el-
evated ratios of sterile to wild males. Incorporating the use of ginger root oil extract into
sterile release programs may thus increase the effectiveness of the sterile insect technique,
and/or allow a reduction in the number of sterile flies that are released.

Key Words: sterile insect technique, Ceratitis capitata, mating behavior, alpha-copaene


El comportamiento de copulacion en las moscas de la fruta, Ceratitis capitata (Wiedemann),
mediterraneas est6riles, machos criadus en el laboratorio, fue evaluado en pruebas de com-
petencia de copula enjaulas de campo fuente con moscas salvajes. Estos preubas se iniciaron
despu6s de que los insects del laboratorio fueron expuestos al extract del aceite del raiz de
jengibre. Sin la exposici6n al aceite, los machos est6riles obtuvieron 12,6%, 69,0%, y 72,8%
de las copulas totales con hembras salvajes cuando fueron presents en proporciones de 1:1,
5:1, y 10:1 de machos est6riles a machos salvajes, respectivamente. Los machos est6riles, ex-
puestos 1 d antes de copula al olor del aceite de jengibre para 3 h, en un proporcion de 1:1
con machos salvajes, alcanzaron 62,3% de las copulas con hembras salvajes. Estos datos su-
gieren que la exposici6n al aceite pueda elevar la competencia sexual del macho est6ril a un
grado igual a proporciones elevados (ca. 5:1) de machos est6riles contra machos salvajes. In-
corporarando el uso del extract del aceite en programs de accion de liberar moscas est6ri-
les, puede aumentar asi la eficiencia de la t6cnica est6ril del insect, y/o permitira una
reducci6n en el numero de moscas est6riles liberadas. Translation provided by author.

The Mediterranean fruit fly, Ceratitis captitata
(Wiedemann) (medfly), is a multivoltine, polypha-
gous (250+ plant species) insect pest that could
have a devastating economic impact if it were to
become established in California or Florida (Met-
calf 1994). In these states, the sterile insect tech-
nique (SIT) is currently used to inhibit
Mediterranean fruit fly colonization. The goal of
these programs is for sterile males to mate with
any introduced wild females, resulting in the pro-
duction of infertile eggs (Dowell et al. 2000).
There is an ongoing interest in improving the
quality of sterile males used in eradication and
control programs for the Mediterranean fruit fly.
One way to assess fly quality is through the use of

mating competition tests, where sterile males
compete with wild males for wild females. In most
cases, mating observations and tests have shown
that laboratory-reared male flies are at a disad-
vantage with wild males when competing for fe-
males (Robinson et al. 1986, Shelly et al. 1994,
Cayol et al. 1999, Lance et al. 2000, but see Taylor
et al. 2001 which found no advantage for wild
males). In one case in Hawaii, following apparent
intense selection in the field, there was almost
complete behavioral resistance by wild females to
laboratory reared sterile males (McInnis et al.
Reduced mating performance of sterile labora-
tory males is attributed to the mass-rearing pro-

Florida Entomologist 86(1)

cess, which through artificial selection, may
result in the production of flies with qualities dif-
ferent from their wild counterparts (Cayol 2000).
There are two options available to increase the
mating success of SIT flies. The first is the release
of greater numbers of sterile flies so that they
vastly outnumber wild males. Recommended ster-
ile/ wild fly release ratios vary with each strain,
but have been suggested to be 125:1 for the
Petapa strain and 100:1 for the Vienna-4/Tol-94
strain based on field cage tests using sterile/wild
fly ratios of 1:1, 5:1, 25:1, and 125:1 (Garcia et al.
1999). A second option would be to use "male-
only" strains, containing 95%+ males, which have
resulted in a significant improvement over re-
leases containing both sterile males and females
(McInnis et al. 1986, McInnis et al. 1994, Hen-
drichs et al. 1995, Rendon et al. 2000). The bene-
fits of "male-only" strains, derived from the
virtual elimination of the sterile male and sterile
female interaction, are widely accepted and as
such, these strains are now used in many rearing
and release programs. Both of these options result
in improving the success of SIT by altering the
probability of interactions between sterile and
wild flies. If, in addition, sterile fly mating com-
petitiveness could be improved, then releasing a
lower number of sterile flies would result in an
even more efficient SIT program
A survey in the 1950's identified several attrac-
tants of male medflies (Beroza & Green 1963) that
were later developed for use with medfly trapping
and monitoring programs. One of these attracta-
nts, trimedlure, was shown to increase the mating
success of males after they were exposed to it
(Shelly et al. 1996). In addition, exposure of med-
fly males to ginger root (Zingiber officinale
Roscose) oil, which contains alpha-copaene at
0.4% volume, increased their mating competitive-
ness vs. wild males (Shelly 2001, Shelly & McIn-
nis 2001). Alpha-copaene is a known male
attractant that has been identified from angelica
oil (Guiotto et al. 1972, Flath et al. 1994a, b,
Nishida et al. 2000). Shelly & McInnis (2001)
found a several-fold greater mating success of
mass-reared, sterile flies exposed to ginger root oil
compared with sterile flies not exposed. This
study adds to previous research by comparing the
mating success of sterile males at different ster-
ile:wild fly ratios, to a 1:1 ratio of sterile (exposed
to ginger root oil): wild flies.


Study Animals

Wild male and female Mediterranean fruit
flies were collected as larvae and eggs from coffee,
Coffea arabica L., from Kauai, Hawaii and from
loquats, Eriobotrya japonica Thunb., from Kula,
Maui, Hawaii in February and March, 2001.

Fruits were placed on screens above vermiculite
(22-26C) that was sifted every 5-7 d for pupae. To
obtain virgin flies for mating trials, newly
emerged flies were separated by sex <2 d after
eclosion. Adult flies were fed honey, sugar, and
protein hydrolysate until they were sexually ma-
ture (>12 d old). Twenty five adult flies were held
collectively in small plastic containers (400 ml)
with nylon mesh screening. The source of wild
flies for each mating trial was dependent on the
availability of host material (coffee and/or loquat)
and the number of flies obtained from each host.
Laboratory-reared flies of the Vienna-4 (Toli-
man) strain (male-only genetic sexing strain, car-
rying a temperature sensitive lethal (tsl)
mutation) were obtained from the Tropical Fruit
and Vegetable Research Laboratory, USDA-ARS,
in Honolulu, HI. The sexing strain was obtained
from the mass-rearing facility in Guatemala at El
Pino in 1998. The larval rearing protocol followed
that of Tanaka et al. (1969) & McInnis et al.
(1994). Before irradiation, white pupae (pre-
sumed to be females) were removed in order to
achieve a higher percentage of males. In prepara-
tion for irradiation, which occurred 2 d before
eclosion, pupae were placed in hypoxia for 1-2 h.
Sterilization was achieved using a dose of 14.5 Kr
in a Cobalt60 irradiator located at the University
of Hawaii, Manoa (McInnis et al. 1996, Rendon et
al. 1996). Adult sterile flies were fed honey, sugar,
and protein hydrolysate until they were sexually
mature (at least 4 d old). Adult flies were held in
cages (16 liters, 225-250 flies per container) with
nylon mesh screening until testing in the field.
(Different size containers were used for wild and
sterile, laboratory flies, in part because of the
numbers of each type needed for each experiment
and also to reduce the mortality of wild flies,
which is usually high in the laboratory. Carey et
al. (1995) have investigated the affects of different
fly densities on mortality.

Mating Tests

On the day before a mating trial, one male type
was marked (wild and Vienna-4 were alternated
with each replicate) by placing a small drop of
enamel paint on the thorax. This allowed later
identification of male type in mating pairs. In the
ginger root oil treatment, 1 day before the mating
trial, 25 laboratory male flies were exposed in
each of two small plastic containers (400 ml) for 3
h to 20 pl of ginger root oil (Citrus and Allied Es-
sences Ltd., Lake Success, NY) placed on a 1 cm2
piece of blotter paper. These flies were exposed in
an isolated room that was distant from all other
Circular, nylon-screened, field cages (2.5 m
high, 2.5 m diameter, and containing a 2 m tall
guava tree, Psidium guajava L.) were used for the
mating trials (McInnis et al. 1996). Four treat-

March 2003

Barry et al.: Potential for Lower Overflooding Ratios

ments were randomly assigned to four field cages
for each of five replications over time. The treat-
ments were 1:1, 5:1, and 10:1 ratios of sterile lab-
oratory-reared males to wild males, and a 1:1
ratio of sterile laboratory-reared males exposed to
ginger oil to wild males. In each cage, there were
always 25 wild males and 25 wild females, so dif-
ferent ratios of laboratory to wild males were ob-
tained by altering the number of laboratory-
reared males (i.e., the 1:1 ratio had 25 laboratory-
reared males, 5:1 had 125, and 10:1 had 250).
On the day of a mating experiment, males were
released first into each of the cages, followed by
the females after an interval of 5-10 min. Flies
that were dead, incapable of flight, or noticeably
damaged in any way at the time of release were
replaced. Two field observers, alternating be-
tween cages every 15 min, located and removed
mating pairs without replacement. Observations
were made from approximately 0900 until 1400 h.
Temperature ranged from 23-29C and relative
humidity ranged from 47-68% (HOBO@ datalog-
ger, Pro Temp/ RH, Onset Computer Corporation,
Bourne, MA).
An ANOVA was performed on data, after arc-
sine transformation, for the three treatments
with sterile males not exposed to ginger root oil to
determine if there were significant differences in
mating based on the sterile:wild fly ratio. A two-
way t-test was used to compare transformed data
from all four treatments.


Mating Tests

With flies not exposed to ginger root oil, the
sterile: wild fly ratio had a significant effect on the
percentage of mating pairs involving a sterile
male (F = 14.98; df 2, 12;P = 0.001; Fig. 1). Results
with the 1:1 ratio treatment, with sterile flies not
exposed to ginger oil, were significantly different
from all other treatments (1:1 to 5:1, P = 0.0057,
df = 4; 1:1 to 10:1, P = 0.0068, df = 4; and 1:1 to 1:1
ginger exposed, P = 0.013, df = 4). Pair-wise com-
parisons among the three remaining treatments
(1:1 with ginger exposed sterile males, 5:1, and
10:1) were not significant (P > 0.05, df= 4).


Specially treated mass-reared flies have the
potential to outperform wild flies in mating com-
petitiveness trials (Shelly & McInnis 2001). In our
study, ginger root oil was responsible for in-
creased mating success with sterile laboratory-
reared flies. At a 1:1 ratio of sterile (exposed): wild
flies, sterile males had a success (62% of the mat-
ings) similar to wild males. In comparison, a 1:1
ratio of sterile (not exposed): wild flies, resulted in
sterile males failing to mate in 3 of 5 replicates.

1:1 5:1 10:1
Ratio of sterile to wild males
Fig. 1. Mating success (mean + SE) of Vienna-4 (Toli-
man) sterile males not exposed (.), or exposed to ginger
root oil (o), for 3 h, 1 d before mating competition trials
with wild flies. The 1:1 treatment with sterile flies not
exposed to ginger oil was significantly different from all
other treatments (1:1 to 5:1, P = 0.0057, df = 4; 1:1 to
10:1, P = 0.0068, df = 4; and 1:1 to 1:1 ginger exposed,
P = 0.013, df= 4).

Sterile males (not exposed) showed mating suc-
cess similar to, or above that of, wild males only at
elevated ratios of 5:1 and 10:1 (sterile: wild
males). These results suggest that exposure to
ginger root oil elevates male mating success to a
degree comparable with elevated ratios (5:1 or
10:1) of sterile/ wild males.
During a medfly infestation, in areas where
eradication is the goal (i.e., Florida and Califor-
nia), it is likely that sterile:wild male ratios will
be more skewed than the highest ratio that was
tested in this study, 10:1, because of daily releases
of sterile flies. Assuming that these cage tests are
indicative of what may happen during mass re-
leases, there could be as much as a 1/5 reduction
in the number of flies released if they were pre-
treated with ginger root oil or if numbers of flies
remained unchanged, then the flies released
would have five times the mating success com-
pared with the present system. The use of ginger
root oil should result in a fly that is qualitatively
better, regardless of the sterile:wild fly ratio.
In this study, the number of available wild flies
was a limiting factor. Using higher numbers of
wild flies in each cage would increase the power of
the statistical tests, but would also increase the
number of sterile flies that would be needed in or-
der to maintain the ratios used. Using more than
300 flies in a field cage of the size we used (see Ma-
terials and Methods) could raise concerns about
unnaturally high densities of flies, and whether
the results could reasonably be extrapolated to
open field conditions. Fly density in this study
ranged from 15-60 flies/ m2 (75 to 300 flies in a 4.9
m2 cage). The Preventative Release Program

(PRP) in Southern California releases approxi-
mately 300 million flies each week over an area of
approximately 6,446 km2 (CDFA 2001). This cor-
responds to a density of approximately 0.048 flies/
m2 (or 1 fly/ 21 m2), assuming 0% mortality for re-
leased flies, 100% mortality for flies released the
previous week, and uniform fly distribution. Test-
ing higher ratios of sterile to wild flies, as well as
using more total flies could be better accom-
plished with the use of larger field cages, however,
the ability to find almost all mating pairs then be-
comes more difficult.
Further study is needed to determine how to
incorporate ginger root oil into mass-rearing pro-
grams. The impact of ginger root oil on mating
performance should also be evaluated against
other strains of wild flies and using other sterile
male strains. In addition, the economics and prac-
tical methodology of incorporating ginger root oil
exposure into the large Mediterranean fruit fly
SIT programs is now being investigated (TES and
DOM, unpublished data).


We thank Steve Tam and Jason Komatsu of the
USDA-ARS Genetics Laboratory (Honolulu, HI) for sup-
plying flies, assisting in field observations, and provid-
ing comments on the design of this experiment; Linda
Tran for assisting with field work; and R. D. Goeden, R.
V. Dowell, and T. S. Bellows for reviewing the manu-
script. This research was funded in part by the Califor-
nia Citrus Research Board and forms part of the Ph.D.
dissertation of JDB.


BEROZA, M., AND N. GREEN. 1963. Materials tested as
insect attractants. USDA-ARS Agriculture Hand-
book No. 239. U.S. Dept. ofAgric., Washington, D.C.
CAREY, J. R., P. LIEDO, AND J. W. VAUPEL. 1995. Mortal-
ity dynamics of density in the Mediterranean fruit
fly. Experimental Gerontology 30: 605-629.
CAYOL, J. 2000. Changes in sexual behavior and life his-
tory traits of Tephritid species caused by mass-rear-
ing processes, pp. 843-860. In M. Aluja and A. L.
Norrbom [eds.], Fruit flies (Tephritidae): phylogeny
and evolution of behavior. CRC Press, Boca Raton,
New indices and method to measure the sexual com-
patibility and mating performance of Ceratitis capi-
tata (Diptera: Tephritidae) laboratory-reared strains
under field cage conditions. J. Econ. Entomol. 92:
CDFA. 2001. Mediterranean fruit fly preventative re-
lease program: report to the legislature. California
Department of Food and Agriculture: Plant Health &
Pest Prevention Services. March, 2001.
SPAUGY. 2000. Mediterranean fruit fly preventative
release programme in southern California, pp. 369-
375. In K. H. Tan [ed.], Area-Wide Control of Fruit
Flies and Other Insect Pests. Penerbit Universiti
Sains Malaysia, Pulau Pinang, Malaysia.

March 2003

JOHN. 1994a. Male lures for Mediterranean fruitfly
(Ceratitis capitata Wied.): Structural analogs of al-
pha-copaene. Journal of Chemical Ecology 20: 2595-
JOHN. 1994b. Additional male mediterranean fruit-
fly (Ceratitis capitata Wied.) attractants from angel-
ica seed oil (Angelica archangelica L.). Journal of
Chemical Ecology 20: 1969-1984.
CASTANEDA. 1999. Evaluation of competitiveness be-
tween two strains of Mediterranean fruit fly Cerati-
tis capitata (Wied.) Villa Canales, Guatemala, pp.
39, Working group on fruit flies of the western hemi-
sphere, Guatemala City, Guatemala.
1972. Investigations of attractants for males of Cer-
atitis capitata. Farmaca. Ed. Sci. 27: 663-669.
creased effectiveness and applicability of sterile in-
sect technique through male-only releases for
control of Mediterranean fruits flies during fruiting
seasons. J. Appl. Entomol. 119: 371-377.
JACKSON. 2000. Courtship among sterile and wild
Ceratitis capitata (Diptera: Tephritidae) in field
cages in Hawaii and Guatemala. Ann. Entomol. Soc.
Am. 93: 1179-1185.
MCINNIS, D. O., T. T. Y. WONG, AND S. Y. T. TAM. 1986.
Mediterranean fruit fly [Ceratitis capitata] (Diptera:
Tephritidae): Suppression efficiencies of unisexual
and bisexual sterilized release populations in field
cages. Ann. Entomol. Soc. Am. 79: 931-937.
Behavioral resistance to the sterile insect technique
by Mediterranean fruit fly (Diptera: Tephritidae) in
Hawaii. Ann. Entomol. Soc. Am. 89: 739-744.
1994. Population suppression and sterility rates in-
duced by variable sex ratio, sterile insect releases of
Ceratitis capitata (Diptera: Tephritidae) in Hawaii.
Ann. Entomol. Soc. Am. 87: 231-240.
METCALF, R. L. 1994. Biography of the Medfly, pp. 43-
48. In J. G. Morse, R. L. Metcalf, J. R. Carey and R.
V. Dowell [eds.], The Medfly In California: Defining
Critical Research. University of California, Center
for Exotic Pest Research, Riverside, CA.
KANESHIRO. 2000. alpha-Copaene, a potential ren-
dezvous cue for the Mediterranean fruit fly, Ceratitis
capitata? Journal of Chemical Ecology 26: 87-100.
RENDON, P., G. FRANZ, AND R. J. WOOD. 1996. Assess-
ment of irradiation doses for TSL (thermal sensitive
lethal) strain Vienna 42, pp. 193-196. In B. A.
McPheron and G. J. Steck [eds.], Fruit Fly Pests: A
world assessment of their biology and management.
St. Lucie Press, Delray Beach, FL.
2000. Comparison of Medfly male-only and bisexual
releases in large scale field trials, pp. 517-525. In K.
H. Tan [ed.], Area-Wide Control of Fruit Flies and
Other Insect Pests. Penerbit Universiti Sains Malay-
sia, Pulau Pinang, Malaysia.
CAPPARELLA. 1986. Field cage studies with a genetic
sexing strain in the Mediterranean fruit fly, Ceratitis
capitata. Entomol. Exp. Applic. 41: 231-236.

Florida Entomologist 86(1)

Barry et al.: Potential for Lower Overflooding Ratios

SHELLY, T. E. 2001. Exposure to alpha-Copaene and al-
pha-Copaene-containing oils enhances mating suc-
cess of male Mediterranean fruit flies (Diptera:
Tephritidae). Annals of the Ecological Society of
America 94: 497-502.
SHELLY, T. E., AND D. O. MCINNIS. 2001. Exposure to
ginger root oil enhances mating success of irradi-
ated, mass-reared males of Mediterranean fruit fly
(Diptera: Tephritidae). J. Econ. Entomol. 94: 1413-
1994. Sterile insect release and the natural mating
system of the Mediterranean fruit fly, Ceratitis capi-

tata (Diptera: Tephritidae). Ann. Entomol. Soc. Am.
87: 470-481.
1996. Trimedlure affects mating success and mate
attraction in male Mediterranean fruit flies. Ento-
mol. Exp. Applic. 78: 181-185.
MOTO. 1969. Low-cost larval rearing medium for
mass-production or oriental and Mediterranean
fruit flies. J. Econ. Entomol. 62: 967-968.
2001. Sexual competitiveness of Vienna 4/Tol-94 'ge-
netic sexing' sterile Mediterranean fruit fly males in
Israel. Phytoparasitica 29: 7-14.

Florida Entomologist 86(1)

March 2003


University of Florida, IFAS, Citrus Research and Education Center, 700 Experiment Station Road
Lake Alfred, Florida 33850


The seasonal abundance of various life stages of Diaprepes abbreviatus (L.), was monitored
in separate years in adjacent irrigated and non-irrigated citrus plantings, as well as thickets
of Brazilian-pepper located near Poinciana, FL in Osceola County. Adult emergence, esti-
mated by weekly catches in cone-shaped ground traps, occurred throughout the year with a
peak in mid-June in both citrus and Brazilian-pepper plantings. Onset of adult emergence
coincided with an increase in soil moisture and temperature. Trap counts were highest when
soil water potential increased to 3-5 centibars at a depth of 15-30 cm and soil temperature
averaged 22-24EC. In the non-irrigated citrus planting, the adult emergence peak was of
shorter duration, but of greater magnitude, compared to the irrigated planting. Although the
alternate host, Brazilian-pepper, produced fewer weevils than did citrus, the seasonal emer-
gence pattern was virtually the same.
Adult abundance within the citrus plantings was also monitored weekly using modified Ted-
ders traps. The number of adults captured approximated the number caught in ground
traps. Adult number caught weekly changed seasonally, particularly in the fall when adult
populations were the highest. Ground traps caught a larger number of adults in the spring.
The number of egg masses collected weekly in the tree canopy and the number of neonates
caught weekly beneath the tree canopy were both correlated with the number of adults cap-
tured weekly in modified Tedders traps. These data suggest that adults caught in modified
Tedders traps provide a reliable indicator for estimating the seasonal abundance of all life
stages within a citrus planting. Larvae of different instars, pupae, and general adults were
recovered from the soil rhizosphere after periodic tree removal. No diseased or parasitized
life stages were observed in the study. Most life stages were present in the soil at each sam-
ple date, but the proportion of larvae in various instars changed seasonally. The implications
of this study for understanding the population dynamics of D. abbreviatus are discussed in
relation to current and future IPM strategies.

Key Words: Diaprepes abbreviatus, population dynamics, citrus, Brazilian-pepper


Se realize un monitoreo sobre la abundancia estacional de varias etapas de vida de Diapre-
pes abbreviatus (L.) en aios separados en siembras de citricos irrigadas y no irrigadas en
campos adjacentes, asi como en matorrales de la pimienta de Brasil (Brazilian peppertree,
Schinus terebinthifolius) cerca de Poinciana, Florida en el condado de Osceola. La emergen-
cia de los adults, estimada por el numero de adults atrapados en las trampas de forma de
un cono puestas sobre el suelo, aconteci6 a trav6z del ano con el punto mas de la poblaci6n
ocurriendo en el medio de junio en siembras de citricos y en Schinus terebinthifolius. El ini-
cio de la emergencia de los adults coincidio con un aumento en la humedad del suelo y de
la temperature. El numero de Diaprepes atrapados por trampa fue el mas elevado cuando el
potential del agua de suelo aument6 al 3-5 centibarras a una profundidad de 15-30 cm y a
un promedio de la temperature del suelo de 22-24C. En siembras de citricos no irrigados, el
punto mas alto de la emergencia de los adults dur6 menos tiempo, pero fu6 de mayor mag-
nitud, comparada con las siembras no irrigadas. Aunque el hospedero alternative, Schinus
terebinthifolius produj6 menos picudos de lo que fueron producidos en los citricos, el patron
de emergencia estacional fu6 virtualmente el mismo.
Tambi6n, se realize un monitoreo semanalmente sobre la abundancia de adults entire la
misma siembra de citricos usando trampas modificadas del tipo "Tedders". El numero de
adults capturados fu6 aproximadamente el numero recolectado en las trampas del suelo. El
numero de adults recolectados semanalmente cambi6 segun la estaci6n, particularmente
en el otono cuando la poblaci6n de adults fu6 la mas alta. Las trampas de suelo capturaron
un mayor numero de adults en la primavera. El numero de masas de huevos recolectados
semanalmente en la copa del arbol y el numero de neonatas (larvas reci6n nacidas) captura-
dos semanalmente debajo la copa del arbol fueron ambos correlacionados con el numero de

McCoy et al.: Seasonal Abundance ofD. abbreviatus

adults capturados semanalmente en las trampas modificadas de tipo "Tedders". Estos datos
sugieren que los adults capturados en las trampas modificadas de tipo "Tedders" proven
un indicador confiable para estimar la abundancia estacional de todas las etapas de vida
dentro de una siembra de citricos.
Larvas de diferentes estadios, pupas y adults tenerales fueron recobradas de la riz6fera
despu6s de la eliminaci6n de arboles hecha peri6dicamente. No fueron observadas las etapas
de vida con enfermedades o parasitadas en el studio. La mayoria de las etapas estuvieron
presents en el suelo en cada fecha de muestreo, pero la proporci6n de larvas en las varias
etapas cambi6 segun la estaci6n. Se discuten las implicaciones de este studio para entender
la dinamica de la poblaci6n de D. abbreviatus en relaci6n con las estrategias actuales y fu-
turas de Manejo Integrado Plagas (MIP).

Diaprepes abbreviatus (L.) a root weevil native
to the Caribbean region (O'Brien & Wibmer 1982)
is a major localized pest of commercial citrus, or-
namental plants, and some agronomic crops in
Florida since its introduction in 1964 (Woodruff
1964, McCoy 1999). It has recently spread to cit-
rus and ornamentals in Texas (French & Skaria
2000) and has been intercepted frequently by reg-
ulatory officials in California (Kris Godfrey, per-
sonal communication). Currently, it ranks as a
high risk pest of California agriculture. The adult,
egg, and neonate stages appear on above-ground
parts of the host plant and all larval stages, pupae
and general adults occur below ground (Wolcott
1936). Although this weevil can be univoltine on
citrus, the life cycle varies in duration with many
over-lapping generations and many different life
stages may be present simultaneously. Upon
hatching, neonates fall from tree and enter the
soil. The small neonates feed on fibrous roots,
whereas later instars feed on larger structural
roots, causing deep grooves as they consume the
outer bark and cambium layer. From year to year
root injury appears to accumulate, and feeding
sites can serve as infection courts for root rot dis-
eases such as Phytophthora spp. (Graham et al.
1996, 2002), thereby exacerbating economic
Current integrated pest management (IPM)
strategies for weevil suppression include horticul-
tural practices such as irrigation and fertilization,
Phytophthora control in the soil, and a mix of con-
trol tactics aimed at reducing larval and adult
populations (McCoy & Duncan 2000). Monitoring
of adults using visual counts or modified Tedders
traps has been deployed to better time foliar and
soil applications of adulticides and larvicides in
citrus groves (Stansly et al. 1997, Duncan et al.
2001). Although trapping has been useful in de-
termining seasonal patterns of adult emergence
from the soil, a lack of knowledge of the develop-
mental biology and ecology of D. abbreviatus has
limited our understanding of trap accuracy and
population dynamics of various life stages in the
The purpose of this 2-yr study was to assess
the relative abundance of various life stages of
D. abbreviatus in adjacent distinct citrus plant-

ings and in Brazilian-pepper thickets surround-
ing the citrus using different monitoring methods.
Brazilian-pepper is a host to both adults and lar-
vae of Diaprepes (Simpson et al. 1996) where they
feed on leaves and roots, respectively. Relative
abundance of different life stages was then re-
lated to abiotic factors such as air temperature,
soil temperature, rainfall, and soil moisture. No
irrigation was deployed in the first year. This in-
formation was deemed fundamental to the devel-
opment and application of various IPM strategies.


Experimental Site

Independent studies were conducted near
Poinciana, FL in Osceola County in two distinct
plantings of Hamlin orange grafted to Swingle cit-
rumelo rootstock set at 6.1 x 8.5 m. These adjacent
plantings, designated north (40 acres) and south
(50 acres), were planted on two-row beds, in a
poorly drained alfisol soil type classified as Flori-
dana fine sand (68.8% sand, 11.8% silt, 19.4%
clay). The surface layer was 35.6 cm loam and the
subsurface layers 76.2 cm gray fine sand followed
by clay. The soil had a low to moderate organic
matter content with a pH of 4.8. Both plantings
had been infested with D. abbreviatus for at least
10 yr according to local reports. In the north
planting, an estimate of weevil larvae in the soil
was made in early February 2000, shortly before
our study was begun. Three trees were removed
and soil sieved to recover larvae. These trees pro-
duced 83-86 late instar larvae/0.4 m3 of soil. Al-
though tree decline was severe in the north
planting at this time, high larval populations in
combination with poor horticultural care (no irri-
gation) resulted in greater tree decline by years
end. The rapidity of tree decline and death in the
north planting in 2000, necessitated moving our
study to the south planting in 2001. In so doing,
our original experimental plan to replicate by
year was lost.
In the north planting, 150 productive trees
with decline symptoms were purposely selected
for studies in 2000. These trees were non-irri-
gated and received no chemical treatments dur-

Florida Entomologist 86(1)

ing the study period. The grove was mowed
regularly to reduce weed growth and experimen-
tal trees received a herbicide application (glypho-
sate at standard rate) as needed to prevent weed
growth at the tree canopy margin and beneath the
tree. Some trees were pruned lightly to improve
accessibility beneath the tree. One fertilizer ap-
plication was applied to the healthy trees in mid-
April using a standard citrus mixture (6-6-6).
Brazilian-pepper trees, Schinus terebinthifolius
Raddi, were growing wild throughout the north
planting along with other woody weed hosts of
D. abbreviatus.
By contrast, the south planting used for study
in 2001, was highly productive and had been un-
der regular grove care that included irrigation on
a need basis, fertilization twice per year, and reg-
ular weed and pest control. Fertilization and weed
control only were continued in 2001 following the
same schedule.

Monitoring Adult Weevils

In each planting, seasonal adult emergence
from the soil and adult abundance in the grove
were monitored using cone-shaped screened
ground traps (0.9 m base dia.) and modified pyra-
midal Tedders traps (Tedders & Wood 1994, Mc-
Coy et al. 2000), respectively. A trap of each type
was placed as pairs opposite each other beneath a
tree, midway between the tree trunk and the can-
opy dripline. Within each planting, 100 randomly
selected trees received a pair of traps. Traps were
monitored weekly from 17 March through
18 December 2000 and from 6 March through 11
December 2001. In 2001, an additional 100 cone-
shaped ground traps covered with mylar plastic to
simulate no irrigation, were monitored weekly
within the grove from 6 March through 29 Au-
The whole citrus planting was bordered on
three sides by pasture that harbored various al-
ternate host plants for D. abbreviatus such as tal-
low tree, Sapium sebiferum L., coffee weed,
Sesbania exaltata (Raf.) Rydb. ex A. W. Hill, and
the fore-mentioned Brazilian-pepper. In two
thickets of Brazilian-pepper located in the adja-
cent pasture within 90 m of the south citrus plant-
ing, 100 ground traps were distributed randomly
beneath trees and monitored from 6 March
through 11 December 2001 only.

Monitoring Oviposition in the Tree Canopy

In 2001 only, egg mass abundance within the
accessible part of the tree canopy was monitored
weekly by systematic 15 min visual inspections of
each of 20 trees from 6 March through December
2001. Inspection entailed searching for detection
of attached or folded leaves glued by female wee-
vils during oviposition.

Monitoring Neonate Fall to the Soil Surface

The seasonal abundance of neonates falling
from the trees was assessed using modified pitfall
traps (funnel traps). Traps consisted of a 20.3 cm
diameter plastic funnel attached to 3.1 x 50.0 cm
length of PVC pipe support elevated about 30 cm
above the ground. A screw top conical tube (50
ml), containing 5 ml of ethylene glycol, was at-
tached to the funnel tip using duct tape to serve as
a collection unit.
Eight traps were placed under each tree in the
cardinal directions, four traps at 30 cm from the
trunk and 4 traps at 30 cm from the dripline. Lar-
vae were collected from 25 trees in 2000 and
20 trees in 2001; and collection tubes were
changed weekly. In the laboratory, neonates were
identified and counted. Diaprepes abbreviatus ne-
onates were identified by the frontal suture join-
ing the epicranial suture to form an inverted "Y"
on the head capsule (Beavers & Woodruff 1971).
Trapping was performed weekly from 1 May
through 26 December 2000 and from April
through 31 December 2001.

Monitoring Weevil Life Stages in the Soil

The seasonal changes in relative abundance of
the various life stages ofD. abbreviatus in the soil
was assessed by periodically removing four or
more trees and recovering all detectable life
stages from the soil beneath the extracted trees.
Trees were topped using a chainsaw and the re-
maining roots and surrounding soil were removed
using a back hoe. Most of the soil adhering to the
roots was removed by shaking and/or probing
with a shovel. Soil from the roots and beneath the
tree was then placed into buckets for subsequent
sieving. Approximately 0.59 m3 of soil was col-
lected per tree to a depth of 30 cm according to the
procedures of Duncan et al. (1996). All life stages
ofD. abbreviatus except larvae younger than fifth
instar were visually detectable and recovered
from the soil using a motor-driven shaker and
0.64-cm mesh sieve. The numbers of larvae, pupae
and adults recovered from each tree were re-
corded. Larval instar was determined by head
capsule width measurement (Quintela et al.
1998). All healthy larvae exhibiting normal be-
havior were recorded as'live'. Live larvae exhibit-
ing abnormal behavior and dead larvae were
placed in a disposable Petri dish (50 x 9 mm) on
moistened filter paper. Cadavers were examined
microscopically every other day for 7 d to detect
characteristic signs indicative of bacterial, fungal,
or nematode infection (Lacey & Kaya 1999).

Weather Monitoring

Soil temperature and moisture were continu-
ously monitored at 15 and 30 cm depths using

March 2003

McCoy et al.: Seasonal Abundance ofD. abbreviatus

paired thermistors and calcium block transducers
set to record hourly readings. Rainfall (amount
and duration) was measured using a tipping-
bucket rain gauge linked to a day recorder. Free
water applied via irrigation was also manually re-
corded. Ambient air temperature and relative hu-
midity were recorded using a weekly chart
hygrothermograph. Weekly data were integrated
and stored in a T21X micrologger (Campbell Sci-
entific Inc., Logan, Utah).

Statistical Analysis

SAS System for Windows, release 6.11 was
used for analysis (SAS Institute Inc. 1990). Pro-
portions of adults captured in the two trap types
were compared with contingency table analysis
and x2 tests using PROC FREQ, SAS System for
Windows 6.11 (SAS Institute Inc. 1990). Correla-
tion analysis used PROC CORR.


Adult Weevil Emergence and Abundance

Cone-shaped ground traps, designed specifi-
cally to catch emerging adults, captured 385 wee-
vils from mid-July through the end of October in
the non-irrigated north planting in 2000. The
same type traps caught 428 weevils from mid-
May through the end of November in the irrigated
south planting in 2001. Peak adult emergence oc-
curred in mid-June in the non-irrigated planting
and at approximately the same time the following
year in the irrigated planting (Figs. 1A and B).
Adult emergence in the non-irrigated planting
was closely-related to soil moisture (Fig. 1A).
When soil water potential at a depth of 15-30 cm
increased to 3-5 centibars emergence began
shortly thereafter and continued throughout the
summer. In late September, soil temperature be-
gan to decline and adult emergence dropped off,
even though soil moisture remained at or below
10 centibars (Fig. 1A). Low levels of adult emer-
gence did occur in the early spring when soil tem-
peratures were 2-4EC lower than during the mid-
summer period. Adult emergence in the irrigated
planting in 2001 was lower in magnitude, but
more continuous throughout the year (Fig. 1B).
Soil water potential of <-4 centibars combined
with soil temperatures ranging from 22-24EC in
the summer appeared to favor a longer emergence
period in the irrigated grove (Fig. 1B) compared to
the non-irrigated grove (Fig. 1A).
Tedders traps, designed to catch both emerging
adults and those attracted from the surrounding
area, captured 3113 weevils from mid-March to
mid-December in the non-irrigated planting in
2000. The same type of trap caught 7337 in the ir-
rigated planting in 2001.Although a greater num-
ber of adults were caught in Tedders traps, both

traps detected a strong late spring emergence in
both years (Figs. 1 and 2). However, only Tedders
traps detected a late season peak in adult abun-
dance in September and October that exceeded
the spring peak in magnitude (Figs. 2A and C).
Although there was a significant correlation
between the number of adults caught using the 2
trap types when all year 2000 data was pooled (r
= 0.6682, df = 39, P = 0.0001), this relationship
disappeared during the active emergence period
(22 June through 12 October) when >90% of adult
captures were recorded (r = 0.2682, df = 15, P =
0.2979). In 2001 in the irrigated planting, there
was no significant correlation between catches
from the two trap types even when all data were
pooled (r = 0.2305, df = 38, P = 0.1524). By com-
parison, total weevils captured in cone traps were
428, 1297, and 100, respectively, in irrigated, non-
irrigated citrus and Brazilian-pepper, respec-
tively, in the south planting in 2001. The season-
ality of adult emergence under the three
conditions was quite similar when percent cumu-
lative emergence was compared from mid-March
to late August (Fig. 3A). However, trap counts for
certain sample dates were significantly different
for percent cumulative emergence (P # 0.05) be-
tween irrigated citrus and non-irrigated citrus
based on contingency table analysis and the X2
test. There was no difference between non-irri-
gated citrus and Brazilian-pepper (Fig. 3A). Ex-
ceeding cumulative adult emergence increased
from 29 August to mid-November in both irri-
gated citrus and Brazilian-pepper plantings with
no significant difference between sites (Fig. 3B).
Sampling was terminated in August in the non-ir-
rigated citrus planting to prevent further water
stress to the trees.

Egg Mass Abundance

A total of 811 egg masses ofD. abbreviatus was
detected on leaves in 2001 in the irrigated plant-
ing from 13 June through 11 December with a
peak in mid-Sept (Fig. 2D). The seasonal pattern
for the number of egg masses recovered during
weekly monitoring was similar to the number of
adults caught in modified Tedders traps (Figs. 2C
and D). The correlation between egg mass number
and adult number as indicated by modified Ted-
ders trap counts was highly significant (r =
0.7885, df = 38, P = 0.0001).

Neonate Abundance

In 2000 in the non-irrigated planting, 2408 ne-
onates were recovered from modified pitfall traps
compared to 4841 in the irrigated planting in
2001. Neonate drop was detected from 12 June
through 18 December, 2000 in the non-irrigated
planting and 28 May through 31 December 2001
in the irrigated planting (Figs. 2B and E). Neo-

Florida Entomologist 86(1)


Ground Traps B
60- N = 428






Soil Temperature
30 cm -- / --- ^ -- ~~
20 -
Soil Moisture E 15cm
I1 30 cm



IJ I I t t . ,l i i I . z ,
3/6 4/3 5/1 5/29 6/26 7/24 8/21 9/18 10/16 11/13 12/11


22 C

22 C

Fig. 1. Weekly records of soil moisture, soil temperatures, and numbers of adult weevils captured in cone traps:
(A) non-irrigated citrus planting, 2000, and (B) the irrigated citrus planting, 2001.




March 2003



McCoy et al.: Seasonal Abundance ofD. abbreviatus

3 25
2 20
o 15
o 10


0 5

.- A. N = 3113 1(
0 ,

o- B


i 0

10 1(
50- JH

3/16 4/13 5/11 6/8 7/6 8/3 8131 9128 10126 11123 12/21
Year 2000

. C.
800 N=7337
1 700- C
g 600:

z 200

' D. N= 811
3 60


WS500 -
0 400 -
Z 300 -
S200 -
o ....10,gi nlanll ll
3/6 4/3 5/1 5/29 6/26 7/24 8/21 9/1810/1611/1312/11

Year 2001

Fig. 2. Weekly numbers of adult weevils captured in
Tedders traps, numbers of egg masses counted during
systematic visual canopy inspections and numbers of
neonates captured in funnel traps during 2000 (A, B)
and 2001 (C, D, E).

nate drop was significantly correlated with num-
bers of adults caught in Tedders traps in both
years (year 2000, r = 0.3971, df = 32, P = 0.0201;
year 2001, r = 0.8467, df= 32, P = 0.0001) but nei-
ther egg mass nor neonate abundance were corre-
lated with numbers of adults caught in cone traps
neonatess vs. cone traps, year 2000, r = 0.0143, df
= 32, P = 0.9360; year 2001, r = -0.0884, df= 32, P
= 0.6190; and egg masses vs. cone traps, r =
0.0462, df = 38, P = 0.7769).

)00 A
-0 Irrigated Citrus ab
-o- Non-irrigated Citrus a



3/6 4/3 5/1 5/29 6/26 7/24 8/21

60 I
8/29 9/26 10/24 11/21

Fig. 3. Comparison of the cumulative percent emer-
gence of adult weevils as estimated by cone traps in ir-
rigated citrus, non-irrigated citrus, and Brazilian-
pepper from 6 March through 29 August 2001 (A); and
for irrigated citrus and Brazilian-pepper for 29 August
through 11 December 2001 (B). An "a" denotes signifi-
cance (contingency table analysis, f2 test, P = 0.05 level)
for irrigated citrus compared to non-irrigated citrus and
Brazilian-pepper with no difference between the latter
two; whereas, a "b" denotes significance for irrigated cit-
rus compared to non-irrigated citrus but not Brazilian-
pepper, and no difference between the latter two.

Abundance of Subterranean Life Stages

In the non-irrigated citrus planting in 2000,
1253 life stages ofD. abbreviatus were recovered
from 0.59 m3 of sieved soil collected from the tree
rhizosphere of each of 32 trees. In the irrigated
planting in 2001, 785 life stages were recovered
from 28 trees using the same procedures. Many
fourth instar larvae and younger stages probably
escaped detection. No diseased or parasitized lar-
vae were found among live and dead larvae held
in the laboratory for symptom expression. All de-
tectable larval instars were present in the soil
rhizosphere throughout most of the year, but, the
proportion of the population represented by dif-
ferent life stages often changed seasonally (Figs. 4
and 5). Pupae and general adult recovery was
more erratic.
In the non-irrigated planting in 2000, the only
instar that failed to show significant change in
proportional representation during the year was
the eighth instar (n = 319) whereas, in the irri-
gated planting in 2001, the only instar failing to

Florida Entomologist 86(1)

122125 76

I i i I Ii
Year 2000
Year 2000

146 177116 140 108 85 59 80 19


I- > 0 >
< 5 .5 0 0

Year 2001

7 N = 2038
H Pupae


Fig. 4. Percentage of the weevil population in each life stage in the soil rhizosphere in non-irrigated (2000) and
irrigated (2001) citrus plantings. The total number of life stages recovered monthly is indicated above the stacked
bar for that month, and the overall total is given above the legend.

show this change was the eleventh instar (n = 17;
contingency table analysis, P > 0.05; Fig. 5). The
scarcity of tenth and eleventh instars during the
emergence period suggest an earlier onset to pu-
pation. Pupae and adults showed strong seasonal
peaks in abundance that correlated to peak adult
emergence as monitored with ground traps. More-
over, in 2001 in the irrigated planting, ninth in-
stars were extremely rare in October, whereas
fourth to eighth instars were unusually abundant
at that time.

Comparison of data from both irrigated and
non-irrigated studies suggest that the onset and
magnitude of adult emergence of D. abbreviatus
can be delayed by soil moisture deficit. This re-
sponse to moisture can be related to more frequent
rainfall (Figs. 1A and B) or free water via irrigation
supplemented with rainfall (Fig. 3). A positive ef-
fect of soil moisture on weevil emergence has been
suggested in the literature from the Caribbean
(McCoy 1999) and Florida (Beavers & Selhime
1975). In the laboratory, Lapointe & Shapiro (1999)
found that soil moisture ranging from 20-80% af-
fected larval survival and pupation. Our results
represent the first field study where soil moisture
was linked to seasonal adult emergence. The effect
of soil moisture on adult weevil emergence is well-
documented for the pecan weevil, Curculio caryae

(Horn) (Harris & Ring 1980). Generally, pecan wee-
vils emerge in large numbers after rainfall. Fur-
thermore, Harris & Ring (1980) found that drought
hardened clay soil impeded weevil emergence but
irrigation allowed normal emergence. In addition,
they found that weevil emergence can be post-
poned by withholding moisture, but cannot be ac-
celerated by the addition of moisture before the
normal emergence window.
In many citrus groves located in central and
east coastal areas of Florida, one annual peak of
adult abundance has been detected using Tedders
traps (Stansly et al. 1997, Duncan et al. 2001),
single peak population density generally occurred
typically in the May/June period. In the non-irri-
gated and irrigated citrus plantings in 2000 and
2001 respectively, adult abundance estimated us-
ing Tedders traps demonstrated as two distinct
seasonal peaks, one in late spring and another in
late summer, the latter being the greatest. These
data suggest that the fall peak cannot be ex-
plained by adult emergence. Since adult weevils
are relatively long-lived in the field, at least some
of the late-summer peak in adult abundance
might be due to an accumulation of older weevils.
However, this trend has not occurred in other lo-
cations where adult abundance was monitored
with Tedders traps (Stansly et al. 1997, Duncan et
al. 2001). A more likely explanation for the fall
peak is adult migration from alternate host
plants. As shown in Fig. 3, adults can emerge con-



11th Instar
[ 10th Instar
E39th Instar
8th Instar
S7th Instar
6th Instar
El 5th Instar
] 4thlnstar


March 2003

.I I'

McCoy et al.: Seasonal Abundance ofD. abbreviatus

C 20- c c ,

3 40 E. 9th Instar N= 427
S d b b d d d
0 cd cd d b b c d
20 b
0 ab

C) F. 8th Instar b N = 542
S40- ab cd ab ab cd d
a ab ab
C cd abb ab cd a cdab cd
20 lo- i ll_ ilab
C 2o

40 -G. 7th Instar N= 380
a c
Sa a b bc bc a bc
20. bc a be ab a

0 H. 4-6thInstar N = 191
d d
20- a cd a a '
ba a bc bIc c a b a a
bbc ab ab bc bc
ig 4 | 0 0
Year 2000 Year 2001
Fig. 5. Statistical comparison of the percentage of
the population of each weevil life stage recovered in
time from the soil rhizosphere in non-irrigated (2000)
and irrigated (2001) citrus plantings. For each life stage,
bars with common letters are not significantly different
at the P = 0.05 level based on contingency table analysis
and X2 test.

tinuously from alternate hosts such as Brazilian-
pepper throughout the summer.
As previously mentioned, the Tedders trap is
widely used in scientific research to trap adult
weevils (Duncan et al. 2001). When placed be-
neath the tree midway between the canopy mar-
gin and the tree trunk, Tedders traps are effective
for measuring changes in adult citrus root weevil
abundance within a season, and the onset of adult
emergence from the soil (Stansly et al. 1997, Dun-

can et al. 2001, Nigg et al. 2001). These data show
conclusively a close association of adult trap
counts, egg mass counts, and neonate counts sub-
stantiating the conclusions of the above authors.
Tedders traps, therefore, provide useful esti-
mates, not only of adult number, but also repro-
ductive activity in the field.
Entomopathogenic fungi and nematodes can re-
duce weevil populations in the soil under natural
conditions; however, their spatial distribution and
persistence in the soil is variable (McCoy & Duncan
2000). In independent studies conducted in both
the north and south citrus plantings to determine
the field efficacy of commercial preparations of en-
tomopathogenic nematodes against Diaprepes lar-
vae in the soil, no control was obtained. In addition,
only a few larvae recovered from the soil were par-
asitized by native nematodes and no microbe infec-
tions were diagnosed in the laboratory (McCoy et
al. 2002). These findings agree with the present
data that indicate no soil-inhabiting natural ene-
mies were found in wild larvae collected from the
rhizosphere of extracted trees. The alfisol soil type
found in these citrus plantings has fine particle size
and may be antagonistic to these natural enemies
which, in turn, might partially explain why overall
weevil densities were exceptionally high and tree
injury so severe in both plantings.
Our data on the number of neonates dropping
from trees into funnel traps compared to the num-
ber of adult weevils being captured under trees in
cone-shaped emergence traps suggests that mor-
tality during this part of the life cycle might often
exceed 98%. In 2000, in the non-irrigated grove,
the number of dropping neonates averaged 376.25
per m2; whereas the number of emerging adults
averaged 6.05 per m2, for a survival rate of 1.61%.
In 2001, in the irrigated grove, the number of
dropping neonates averaged 945.51 per m2;
whereas the number of emerging adults averaged
6.73 per m2, for an average survival rate of 0.71%.
Our data on the various developmental stages
of D. abbreviatus have important practical impli-
cations for timing the application of various con-
trol agents used as foliar or soil treatments. Since
prolonged dry weather can delay adult emergence
and, conversely, extended periods of high soil
moisture can extend emergence, seasonal adult
emergence can be highly variable from year to
year. It would appear that monitoring with ground
traps or Tedders traps is worthwhile. Since trap-
ping is not grower friendly, a weather related
model might be of benefit in timing foliar treat-
ments to control emerging adults as they enter the
tree canopy. If adult emergence is prolonged by fa-
vorable soil conditions, the benefits of foliar sprays
to control adults will be lost because of their gen-
erally short residual activity and the possible neg-
ative side effects of multiple applications. Future
control strategies might be more effective if they
could target the general adult stage in the soil.

Peak neonate drop to the soil occurred from
mid-July through October in both citrus plantings
(Fig. 2), suggesting that soil barrier treatments
might be more effective when applied at this time
of the year (McCoy & Duncan 2000). As is true for
foliar sprays for adult control, the residual effect
of a soil barrier treatment with chemicals is too
short to assure prolonged efficacy, suggesting the
need for multiple applications during the period
of highest neonate drop.
A broad range of larval instars occur in the tree
rhizosphere throughout the year. The fact that all
larval stages are present year round with overlap-
ping generations presents enormous problems
when attempting chemical control since suscepti-
bility and appropriate dosage vary with larval age.
Although entomopathogenic nematodes also ex-
hibit differential larval susceptibility to Diaprepes,
they will infect virtually all instars (Shapiro et al.
1999) at levels unachievable by chemicals.


The authors would like to thank Mr. Drew Kelly for
supplying the citrus plantings for this study and the as-
sistance of Ian Jackson, Harry Anderson, Jerry Fojtik,
Angelique Hoyte, Nadine Cuyler, Jeannette Barnes, and
Kata Blythe. The authors would like to thank Drs. J. P.
Michard and Jeffrey R. Brushwein for critical review of
the manuscript. This research was supported by grants
from the Florida Citrus Production Research Advisory
Council Grant No. 942-18E and the USDA, CSREES
Special Research Grant Project. This paper was ap-
proved for publication as Florida Agricultural Experi-
ment Station Journal Series No. R-08964.


BEAVERS, J. B., AND A. G. SELHIME. 1975. Population dy-
namics of Diaprepes abbreviatus in an isolated citrus
grove in central Florida. J. Econ. Entomol. 69: 9-10.
BEAVERS, J. B., AND R. E. WOODRUFF. 1971. A field key
for separating larvae of four species of citrus weevils
in Florida (Coleoptera: Curculionidae). Florida Dept.
Agr. and Cons. Serv., Div. Plant Ind., Entomol. Cir.
112: 1-2.
1996. Estimating sample size and persistence of en-
tomogenous nematodes in sandy soils and their effi-
cacy against the larvae of Diaprepes abbreviatus in
Florida. J. Nematol. 28: 56-57.
HAM, AND R. F. MIZELL. 2001. Estimating the rela-
tive abundance of adult citrus root weevils
(Coleoptera: Curculionidae) with modified Tedders
traps. Environ. Entomol. 30: 939-946.
FRENCH, J. V., AND M. SKARIA. 2000. Citrus root weevil
identified. Texas A & M Kingsville Newsletter.
18(6): 1-4.
Insect-plant pathogen interactions: Preliminary
studies of Diaprepes root weevils injuries and Phy-
tophthora infections. Proc. Florida State Hort. Soc.
109: 57-62.
Phytophthora Diaprepes weevil complex: Phytoph-

March 2003

thora spp. relationship with citrus rootstocks. Plant
Dis. (in press).
HARRIS, M. K., AND D. R. RING. 1980. Adult pecan weevil
emergence related to soil moisture. J. Econ. Ento-
mol. 73: 339-343.
LACEY, L., AND H. KAYA. 1999. Field manual of tech-
niques in invertebrate pathology. Kluwer Academic
Publishers. pp. 1-911.
LAPOINTE, S. L., AND J. P. SHAPIRO. 1999. Effect of soil
moisture on development of Diaprepes abbreviatus
(Coleoptera: Curculionidae). Florida Entomol. 82(2):
McCOY, C. W. 1999. Arthropod pests of citrus roots. pp.
146-156. In L. W. Timmer and L. W. Duncan, eds. Cit-
rus Health Management, St. Paul: APS Press.
McCoY, C. W., AND L. W. DUNCAN. 2000. IPM: An
emerging strategy for Diaprepes in Florida citrus.
Diaprepes Short Course. Florida Agr. Expt. Stat. pp.
McCoY, C. W., D. I. SHAPIRO, AND L. W. DUNCAN. 2000.
Application and evaluation entomopathogens for cit-
rus pest control. pp. 577-596. In L. Lacey and H. K.
Kaya, eds. Field Manual of Techniques in Insect Pa-
thology. Klumer Academic Publishers, Dordrecht,
The Netherlands.
K. NGUYEN. 2002. Field efficacy of commercial prep-
arations of entomopathogenic nematodes against
larvae of Diaprepes abbreviatus (Coleoptera: Curcu-
lionidae) in alfisol type soil. Florida Entomol. (in
J. M. HARRISON, AND N. CUYLER 2001. Distribution
and movement of adult Diaprepes abbreviatus (Co-
leoptera: Curculionidae) in a Florida citrus grove.
Florida Entomol. 84(4): 641-651.
O'BRIEN, C. W., AND G. J. WIBMER 1982. Annotated
checklist of the weevils (Curculionidae sensu lato of
North America, Central America, and the West In-
dies (Coleoptera: Curculionidae). Memoirs of the
Amer. Entomol. Inst., No. 34., 382 pp.
QUINTELA, E. D., J. FAN, AND C. W. MCCOY. 1998. De-
velopment of Diaprepes abbreviatus (Coleoptera:
Curculionidae) on artificial and citrus root sub-
strates. J. Econ. Entomol. 91: 1173-1179.
C. W. MCCOY. 1999. Effects of temperature and host
age on suppression of Diaprepes abbreviatus (Co-
leoptera: Curculionidae) by entomopathogenic nem-
atodes. J. Econ. Entomol. 92: 1086-1092.
ADAIR. 1996. Diaprepes abbreviatus (Coleoptera:
Curculionidae): host plant associations. Environ.
Entomol. 25: 333-349.
Monitoring Diaprepes abbreviatus with Tedders
traps in southwest Florida citrus. Proc. Florida State
Hort. Soc. 110: 22-26.
TEDDERS, W. L., AND B. W. WOOD. 1994. A new tech-
nique for monitoring pecan weevil emergence (Co-
leoptera: Curculionidae). J. Entomol. Sci. 29: 18-30.
WOLCOTT, G. N. 1936. The life cycle of Diaprepes abbre-
viatus at Rio Piedras. P. R. J. Agr., Univ. of Puerto
Rico 20: 883-914.
WOODRUFF, R. E. 1964. A Puerto Rican weevil new to
the United States (Coleoptera: Curculionidae). Flor-
ida Dept. Agr. Div. Plant Ind., Entomol. Circ. 30: 1-2.

Florida Entomologist 86(1)

Deyrup: List of Florida Ants


Archbold Biological Station, P.O. Box 2057, Lake Placid, FL 33862


A list of ants of Florida published in 1989 is replaced to accommodate 49 additional species
now known from Florida, and 34 name changes in species already on the 1989 list. Cur-
rently, 218 species of ants are reliably reported from Florida.

Key Words: exotic species, faunistics


Se present una nueva lista de las hormigas de Florida para reemplazar una lista publicada
en 1989. La nueva lista incluya 49 mas species y 34 cambios de nomenclatura. Al present,
218 species de hormigas son sabidas a hallarse en Florida.

In 1989 Deyrup et al. published a list of the
ants of Florida. In the ensuing thirteen years
there has been considerable myrmecological ac-
tivity, both survey work in Florida, and taxonomic
descriptions and revisions. Consequently, the
1989 list is drastically out of date: there are 49 ad-
ditional species to be added to the list, 34 name
changes that apply to species already on the 1989
list, and 4 species that have been removed from
the list because the records are probably based on
These advances do not mean that there is no
need for further work in the inventory and taxon-
omy of Florida ants. Included in the list are a
number of species that are awaiting description
by various specialists. The species epithets in the
entire genus Brachymyrmex are suspect, al-
though it is clear that at least five species of Bra-
chymyrmex occur in Florida. It is probable that
there are additional species of ants still to be
found in Florida. This can be deduced from the
fact that there are a number of species on the list
that are known from only one or two collections, or
from one or two sites; there are probably other
equally rare species that nobody has been lucky
enough to find. There is reason to suppose that ex-
otic ants will continue to become established in
Florida (Deyrup et al. 2000); even now there are
likely to be some species of localized exotics that
have not yet been reported. On the other hand,
there are four species of ants on the list that have
not been found for many years and may have been
extirpated. These are the native species Formica
subsericea and Solenopsis xyloni, and the exotic
species Myrmelachista ramulorum and Tetramor-
ium lanuginosum.
The 49 species that have been added to the list
were overlooked before for various reasons. They
are not, fortunately, primarily exotics that have
invaded Florida since 1989, although there are

nine exotic species that have been added to the
list. Most of the added Florida records are either
native species whose Florida populations have
been recently discovered, or native species that
have recently been described or are awaiting de-
The number of species listed below, 218, is the
largest ant fauna known from any state in eastern
North America, and is likely to remain so, even af-
ter other large states have received as much at-
tention as Florida. The reasons for this lie in the
convergence of various faunal elements. There is
a set of Antillean species, such as Leptothorax al-
lardycei and L. torreyi in the southern Peninsula.
There are many tropical exotics, including both
Old World species, such as Technomyrmex albipes
and S'r ....... i....., emmae, and New World species,
such as Wasmannia auropunctata and
Pseudomyrmex gracilis. There are southeastern
coastal plain species such as Camponotus socius
and Paratrechina arenivaga. There are, mostly in
north Florida, species probably derived from the
southern Appalachians, such as Pyramica ros-
trata and P pulchella. There are species repre-
senting western lineages, such as Pogonomyrmex
badius and Neivamyrmex texanus. All this diver-
sity, however, still cannot compare with that of
southwestern states such as Arizona, where an al-
most intact, pre-ice age, dryland fauna (Madroter-
tiary) is augmented at higher elevations by north
temperate species.
The following alphabetical list generally fol-
lows the nomenclature in Barry Bolton's catalog
(1995), combined with the nomenclature in his re-
vision of the Dacetini (2000). These two works are
currently the foundations of North American ant
nomenclature. Numbers in parentheses following
a name in the list usually refer to variances from
the nomenclature or from the lists of species in
Bolton 1995 and 2000. These variances are ex-

Florida Entomologist 86(1)

plained in the section following the list. In cases
in which a name appearing in the 1989 list has
been changed and the changes are referenced in
one of the two works by Bolton mentioned above,
the 1989 usage is noted but the original source of
the change in nomenclature may be omitted.
An asterisk (*) denotes a species that was ab-
sent from the 1989 list.
Florida specimens of all but seven species are
deposited in the collection of Archbold Biological


Acanthomyops claviger (Roger).* Okaloosa Co.
Acanthomyops interjectus (Mayr).* Liberty Co.
Amblyopone pallipes (Haldeman). Widespread.
Anochetus mayri Emery. Dade and Palm Beach
Cos. Introduced.
Aphaenogaster ashmeadi (Emery). Widespread
in north FL, south into Highlands Co.
Aphaenogaster carolinensis Wheeler. North
FL; distribution unclear: confounded with miami-
ana. (1)
Aphaenogaster flemingi M. R. Smith. Wide-
Aphaenogaster floridana M. R. Smith. Wide-
spread in north FL, south into Highlands Co.
Aphaenogaster fulva Roger. Widespread in
north FL, south into Orange and Volusia Cos.
Aphaenogaster lamellidens Mayr. Widespread
in north FL, south into Highlands and St. Lucie
Aphaenogaster mariae Forel. Liberty Co. Rare.
Aphaenogaster miamiana Wheeler.* South FL;
distribution unclear: confounded with carolinen-
Aphaenogaster tennesseensis (Mayr). North-
ernmost counties of peninsular Florida.
Aphaenogaster treatae Forel. Widespread.
Aphaenogaster umphreyi Deyrup & Davis.*
Scattered sites in north FL, south into Highlands
Co. (2).
Brachymyrmex brevicornis Emery?* (Genus in
disarray). Columbia Co. Rare. Introduced.
Brachymyrmex depilis Emery? (Genus in dis-
array). Widespread.
Brachymyrmex minutus Forel?* (Genus in dis-
array). Dade and Monroe Cos. Introduced.
Brachymyrmex musculus Forel?* (Genus in
disarray). Widespread. Introduced.
Brachymyrmex obscurior Forel? (Genus in dis-
array). Widespread.
Camponotus caryae (Fitch). Liberty Co. Rare.
Camponotus castaneus (Latreille). Wide-
Camponotus decipiens Emery. Widespread.
Camponotus discolor (Buckley).* Scattered
sites in peninsular Florida, Alachua into High-

lands Cos. Records in 1989 list under sayi Emery
apparently refer to this species.
Camponotus floridanus (Buckley). Wide-
Camponotus impressus (Roger). Widespread.
Listed under Colobopsis in 1989 list.
Camponotus nearcticus Emery. Widespread.
Camponotus pennsylvanicus (DeGeer). North-
ernmost FL, including Panhandle.
Camponotus planatus Roger. South FL, north
into Hillsborough and Orange Cos. Introduced.
Camponotus pylartes Wheeler.* North-central
Camponotus sexguttatus (Fabricius).* Dade
Co. Rare. Introduced.
Camponotus snellingi Bolton. Widespread.
Records under C. pavidus Wheeler in 1989 list re-
fer to this species.
Camponotus socius Roger. Widespread, south
into Broward Co.
Camponotus tortuganus Emery. South FL,
north into Volusia Co.
Cardiocondyla emeryi Forel. Widespread. In-
Cardiocondyla nuda (Mayr). Widespread. In-
Cardiocondyla venustula Wheeler. Wide-
spread. Introduced.
Cardiocondyla wroughtonii (Forel). Wide-
spread in peninsular FL. Introduced.
Cardiocondyla sp.* Widespread in peninsular
FL. Introduced. (3)
Cephalotes varians (F.Smith). Dade and Mon-
roe Cos. (4)
Crematogaster agnita Wheeler.* Monroe Co.
Rare. Introduced.
Crematogaster ashmeadi Mayr. Widespread.
Crematogaster atkinsoni Wheeler. Widespread.
Crematogaster cerasi (Fitch). Widespread in
north FL, south into Highlands Co.
Crematogaster lineolata (Say). Widespread in
north FL, south into Hernando and Sumter Cos.
Crematogaster minutissima Mayr. Wide-
Crematogaster missuriensis Emery.* Panhan-
dle. (5)
Crematogaster pilosa Emery. Widespread.
Crematogaster vermiculata Emery. Scattered
sites in north FL, south into Hillsborough Co.
Crematogaster sp. A (pine species).* Wide-
spread in north and central FL, south into Lee
and Palm Beach Cos.
Crematogaster sp. B (large species in man-
groves).* Monroe Co. Rare.
Cryptopone gilva (Roger). Widespread in north
FL, south into Highlands Co.
Cyphomyrmex minutus Mayr. Widespread in
south FL, north into Hillsborough and Volusia
Cyphomyrmex rimosus (Spinola). Widespread.

March 2003

Deyrup: List of Florida Ants

Discothyrea testacea Roger. Widespread.
Dolichoderus mariae Forel.* Leon Co. Rare.
Dolichoderus pustulatus Mayr.* Scattered
sites throughout FL. Records in 1989 list under D.
plagiatus Mayr refer to this species.
Dorymyrmex bossutus (Trager).* Widespread
in peninsular FL, west into Leon Co. (6)
Dorymyrmex bureni (Trager).* Widespread.
Records in 1989 list under Conomyrma insana
(Buckley) refer to this species. (6)
Dorymyrmex elegans (Trager).* Central penin-
sular ridges. (6)
Dorymyrmex flavopectus M. R. Smith. Central
peninsular FL, south Highlands Co. into Marion
Co. Listed under Conomyrma in1989 list. (6)
Dorymyrmex grandulus (Forel).* North FL,
south into Citrus Co. (6)
Dorymyrmex medeis (Trager).* Scattered sites
in north FL, south into Highlands Co. (6)
Dorymyrmex reginicula (Trager).* Central
peninsular FL, southern Highlands Co. north into
Volusia Co. (6)
Eurhopalothrix floridana Brown & Kempf.
Widespread in peninsular FL.
Forelius pruinosus (Roger). Widespread.
Forelius sp.* Scattered sites in north FL, south
into Citrus Co.
Formica archboldi M. R. Smith. Widespread in
peninsular FL, west into Liberty Co.
Formica pallidefulva Latreille. Widespread in
north FL, south into Highlands Co.
Formica schaufussi dolosa Wheeler. North FL,
south into Lake Co.
Formica subsericea Say. Liberty Co. Rare, not
seen in recent decades.
Gnamptogenys triangularis (Mayr). Isolated
records from Dade, Broward and Escambia Cos.
Rare. Introduced. Record in 1989 list under G. ac-
uleaticoxae (Santschi) refers to this species.
Hypoponera inexorata (Wheeler). Widespread.
Hypoponera opaciceps (Mayr). Widespread.
Hypoponera opacior (Forel). Widespread.
Hypoponera punctatissima (Roger). South FL,
north into Alachua and Bradford Cos. Introduced.
Lasius alienus (Foerster). North FL, south into
Marion Co.
Lasius neoniger Emery. Scattered sites in
north FL.
Lasius umbratus (Nylander). A few sites in
Leptogenys manni Wheeler. Highlands Co.,
north and west into Leon Co. Records in 1989 list
under L. elongata (Buckley) refer to this species.
Leptothorax allardycei (Mann). Monroe and
Lee Cos. Listed under Macromischa in 1989 list.
Leptothorax bradleyi Wheeler. North FL, south
to Highlands Co.
Leptothorax curvispinosus Mayr. Northern-
most FL.
Leptothorax pergandei Emery. Widespread.

Leptothorax schaumii Roger. North FL, south
into Highlands Co.
Leptothorax smith Baroni Urbani. Scattered
sites in north FL, south into Highlands Co.
Records under L. wheeleri in 1989 list refer to this
Leptothorax texanus Wheeler. Widespread. (7)
Leptothorax torrei (Aguayo). Dade, Monroe,
Martin Cos. Listed under Macromischa in 1989
Leptothorax sp.* Undescribed. Liberty and
Leon Cos. (7)
Linepithema humile (Mayr). Scattered sites
throughout FL. Introduced. Listed under Iri-
domyrmex in 1989 list.
Monomorium destructor (Jerdon). Scattered
sites in south FL, chiefly in Key West. Introduced.
Monomorium ebeninum Forel. Monroe Co. In-
Monomorium floricola (Jerdon). South FL,
north into Pinellas Co. Introduced.
Monomorium pharaonis (Linnaeus). Wide-
spread. Introduced.
Monomorium trageri DuBois.* Scattered sites
throughout FL. Records listed under M. mini-
mum (Buckley) in 1989 list refer to this species.
Monomorium viride Brown. Widespread.
Myrmecina americana Emery. Widespread.
Myrmecina sp.* Undescribed. Bradford Co.
Myrmelachista ramulorum Wheeler. Polk Co.
Rare or extirpated. Introduced.
Myrmica punctiventris Roger.* Walton, Sta.
Rosa, Escambia Cos. Rare.
Neivamyrmex carolinensis (Emery). Scattered
sites in north FL, south into Highlands Co.
Neivamyrmex opacithorax (Emery). Scattered
sites throughout FL.
Neivamyrmex texanus Watkins. Scattered sites
in north FL, south into Highlands Co.
Ochetellus glaber (Mayr). Orange and Volusia
Cos. Introduced. Listed under Iridomyrmex in
1989 list.
Odontomachus brunneus (Patton). Widespread
in Peninsula, in Panhandle west into Leon Co.
Odontomachus ruginodis M. R. Smith. South
FL, north into Orange and Co. Introduced.
Odontomachus sp. Undescribed. Highlands
into Orange, Citrus Cos. Records listed under O.
clarus in 1989 list refer to this species.
Pachycondyla stigma (Fabricius). South FL,
north into Orange and Volusia Cos. Introduced.
Paratrechina arenivaga (Wheeler). Wide-
Paratrechina bourbonica (Forel). South FL,
north into Orange and Hillsborough Cos. Intro-
Paratrechina concinna Trager. Widespread.
Paratrechina faisonensis (Forel). Widespread
in north FL, south into Highlands Co.

Florida Entomologist 86(1)

Paratrechina guatemalensis (Forel). South FL,
north to Sarasota and Indian River Cos. Intro-
Paratrechina longicornis (Latreille). South FL,
north to St. Johns and Columbia Cos. Introduced.
Paratrechina paruula (Mayr). North FL, south
to Volusia Co.
Paratrechina phantasma Trager. Peninsular
FL, Alachua Co. south into Palm Beach Co.
Paratrechina pubens (Forel). Dade, Sarasota
and Palm Beach Cos. Introduced.
Paratrechina vividula (Nylander). North FL,
south into Highlands Co.
Paratrechina wojciki Trager. Widespread.
Paratrechina sp. A.* Undescribed. Workerless
parasite of wojciki. Highlands Co.
Paratrechina sp. B.* Undescribed. Workerless
parasite of faisonensis. Hamilton Co.
Pheidole adrianoi Naves. Widespread.
Pheidole bicarinata vinelandica Forel.* North
FL, south into Citrus Co.
Pheidole carrolli Naves. Scattered sites in
north FL, south into Citrus Co. Rare.
Pheidole crassicornis Emery. Scattered sites in
north FL, south into Alachua Co.
Pheidole dentata Mayr. Widespread.
Pheidole dentigula M. R. Smith. Widespread.
Pheidole diversipilosa Wheeler.* A few sites in
Panhandle, east into Columbia Co. Rare.
Pheidole flauens Roger.* South FL, north into
Martin Co. Introduced.
Pheidole floridana Emery. Widespread.
Pheidole lamia Wheeler. Leon and Jackson
Cos. Rare.
Pheidole littoralis Cole. Peninsular FL, west
into Franklin Co., south into Highlands Co.
Pheidole megacephala (Fabricius). Scattered
sites in south FL, north into Hillsborough Co. In-
Pheidole metallescens Emery. Widespread in
north FL, south into Highlands Co.
Pheidole moerens Wheeler. Widespread. Intro-
Pheidole morrisi Forel. Widespread.
Pheidole obscurithorax Naves*. Panhandle,
east into Leon Co. Introduced. (8)
Pheidole tysoni Forel. Alachua and Madison
Platythyrea punctata (F. Smith). South FL,
north into Highlands Co.
Pogonomyrmex badius (Latreille). Widespread.
Polyergus lucidus Mayr. Alachua, Columbia,
Leon, Sta. Rosa Cos. Rare.
Ponera exotica M. R. Smith. Scattered sites in
north FL, south into Highlands Co.
Ponera pennsylvanica Buckley. North FL,
south into Marion Co.
Prenolepis imparis (Say). North FL, south into
Orange Co.
Prionopelta antillana Forel. Marion Co. Intro-

Proceratium crassicorne Emery.* Sta. Rosa
and Liberty Cos. Rare. (9)
Proceratium croceum (Roger). Scattered sites
in north FL, south into Levy Co.
Proceratium pergandei (Emery). North FL,
south into Pinellas and Highlands Cos.
Proceratium silaceum Roger. North FL, south
into Highlands Co.
Proceratium sp. A.* Undescribed. Leon and
Lafayette Cos. Rare.
Proceratium sp. B.* Undescribed. Liberty Co.
Pseudomyrmex cubaensis (Forel). South FL,
north into Polk Co.
Pseudomyrmex ejectus (F. Smith). Widespread.
Pseudomyrmex elongatus (Mayr). South FL,
north to Highlands Co.
Pseudomyrmex gracilis (Fabricius). South FL,
north into Alachua Co. Introduced. Records listed
under P mexicanus in 1989 list refer to this spe-
Pseudomyrmex leptosus Ward. A few sites in
Peninsula, north into Alachua Co.
Pseudomyrmex pallidus (F. Smith). Wide-
Pseudomyrmex seminole Ward. Widespread in
south FL, Orange and Bay Cos. in north FL.
Pseudomyrmex simplex (F. Smith). South FL,
north into Orange Co.
Pyramica abdita (Wesson & Wesson).* Alachua
and Leon Cos. Rare.
Pyramica angulata (M. R. Smith).* A few sites
in north FL. Rare.
Pyramica apalachicolensis Deyrup & Luber-
tazzi.* Leon Co. Rare. (10)
Pyramica archboldi (Deyrup & Cover).* North
FL, south into Volusia Co., west into Jefferson Co.
Pyramica bunki (Brown). Scattered sites in
north FL, south into Highlands Co. Under Smith-
istruma in 1989 list.
Pyramica carolinensis (Brown). A few sites in
north FL. Rare. Under Smithistruma in 1989 list.
Pyramica cloydi (Pfitzer).* Lake Co. Rare.
Pyramica clypeata (Roger). North FL, south
into Highlands Co. Under Smithistruma in 1989
Pyramica creightoni (M. R. Smith). North pen-
insular FL, south into Highlands Co. Under
Smithistruma in 1989 list.
Pyramica deyrupi Bolton.* Marion Co.
Pyramica dietrichi (M. R. Smith). Widespread.
Under Smithistruma in 1989 list.
Pyramica eggersi (Emery). South FL, north
into Alachua Co. Introduced. Under Strumigenys
in 1989 list.
Pyramica gundlachi Roger. Dade, Monroe, Col-
lier Cos. Introduced. Under Sr ... i.-..., in 1989
Pyramica hexamera (Brown). Marion and Her-
nando Cos. Rare. Introduced. Under Epitritus in
1989 list.

March 2003

Deyrup: List of Florida Ants

Pyramica inopina (Deyrup & Cover).* Ala-
chua, Marion, Putnam Cos. Rare.
Pyramica laevinasis (M. R. Smith). Walton Co.
Rare. Under Smithistruma in 1989 list.
Pyramica margaritae (Forel). Scattered sites
in north FL, south into Marion Co. Introduced.
Under Smithistruma in 1989 list.
Pyramica membranifera (Emery). Widespread.
Introduced. Under Trichoscapa in 1989 list.
Pyramica missouriensis (M. R. Smith).* Ala-
chua and Highlands Cos. Rare.
Pyramica ohioensis (Kennedy & Schramm).
Scattered sites in north FL, south into Alachua
Co. Under Smithistruma in 1989 list.
Pyramica ornata (Mayr). North FL, south into
Highlands and Sarasota Cos. Under Smithis-
truma in 1989 list.
Pyramica pilinasis (Forel).* Scattered sites in
north FL, south into Highlands Co.
Pyramica pulchella (Emery). North FL, west
into Bay Co., south into Highlands Co. Under
Smithistruma in 1989 list.
Pyramica reflexa (Wesson & Wesson). North
FL, south into Polk Co. Under Smithistruma in
1989 list.
Pyramica rostrata (Emery).* Northern tier of
FL counties. Under Smithistruma in 1989 list.
Pyramica talpa (Weber). North FL, south into
Highlands Co. Under Smithistruma in 1989 list.
Pyramica wrayi (Brown).* Leon and St. Johns
Cos. Rare.
Solenopsis abdita Thompson.* Widespread.
Solenopsis carolinesis Forel. Widespread.
Solenopsis corticalis Forel. Monroe Co.
Solenopsis geminata (Fabricius). Widespread.
Solenopsis globularia littoralis Creighton.
South FL, north into Alachua and Franklin Cos.
Solenopsis invicta Buren. Widespread. (11)
Solenopsis nickersoni Thompson. Collier Co.
north into Leon Co.
Solenopsis pergandei Forel. Widespread.
Solenopsis picta Emery. Widespread.
Solenopsis tennesseensis M. R. Smith. Wide-
Solenopsis tonsa Thompson.* North FL, Leon
Co. east to Alachua Co., south to Orange Co.
Solenopsis truncorum Forel.* Alachua Co.
Solenopsis xyloni McCook. Western Panhan-
dle. Not seen in recent decades, possibly extir-
Solenopsis sp.* Undescribed parasite of
Pheidole dentata. Gilchrist Co. Rare.
Stenamma foveolocephalum M. R. Smith.*
Walton Co. Rare.
i 'r, .... -.'-.. ., emmae (Emery). South FL, north
into Hillsborough and Volusia Cos. Under
Quadristruma in 1989 list.
a 'r ..f.... -:.-,,. lanuginosa Wheeler. Dade, Lee
and Monroe Cos. Rare. Introduced.
S .... .i louisianae Roger. Widespread.

s'r .., .... ;.... rogeri Emery. South FL, north
into Orange Co. Introduced.
'r .... '... ...., silvestrii Emery. A few sites, from
Monroe into Gadsden Cos. Rare. Introduced.
Tapinoma litorale Wheeler. South FL, north
into Pinellas Co.
Tapinoma melanocephalum (Fabricius). South
FL, north into Brevard Co. Introduced.
Tapinoma sessile (Say). Widespread.
Technomyrmex albipes (F. Smith).* Wide-
spread. Introduced.
Tetramorium bicarinatum (Nylander). Wide-
spread. Introduced.
Tetramorium caldarium (Roger). South FL,
north into Lake Co. Introduced.
Tetramorium lanuginosum Mayr. Hernando,
Holmes, Jackson Cos. Rare, extirpated? Intro-
duced. Under Triglyphothrix in 1989 list.
Tetramorium simillimum (F. Smith). South
FL, north into St. Johns Co. Introduced.
Trachymyrmex jamaicensis (Andre). Martin
and Monroe Cos. Rare.
Trachymyrmex septentrionalis (McCook).
Wasmannia auropunctata (Roger). South FL,
north into Orange Co. Introduced.
Xenomyrmex floridanus Emery. South FL,
north into Orange Co.

(1). Aphaenogaster carolinensis is listed as a
subspecies ofA. texana Wheeler in Bolton 1995. It
was raised to species level by Gary Umphrey
(1996). Aphaenogaster carolinensis is difficult to
separate from A. miamiana.
(2). Aphaenogaster umphreyi is a recently de-
scribed species (Deyrup and Davis 1998).
(3). There are at least five species of Cardio-
condyla in Florida, differentiated by structural
character states, coloration and habitat prefer-
ences. There is also some variation in color and
sculpture within apparent species. This could de-
note additional species, but I have kept in mind
that members of this genus are fully capable of
founding inbred populations from the introduc-
tion of a single female, so variation could result
from the introduction of a species from more than
one source. If this occurred, there might be vari-
ant forms, even sympatric variant forms, that re-
flected geographic variation within the natural
range of the species, or genetic drift. The persis-
tence of these forms would reflect the degree of in-
breeding in the population. I have resorted to
using Creighton (1950), which leaves one species
without a name. There are several names that
might be applied to this species, or to other Flor-
ida species that I have identified using the
Creighton key. These names include obscurior
Wheeler, ectopia Snelling, mauritanica Forel, and
minutior Forel. Inconvenient though it might
seem, the most sensible approach to understand-

ing the taxonomy of exotic Cardiocondyla would
be to begin by looking at specimens from the na-
tive ranges of the species.
(4). Cephalotes varians was, until recently,
placed in the genus Zacryptocerus, now synony-
mized with Cephalotes (Andrade and Baroni Ur-
bani 1999).
(5). Crematogaster missuriensis (the original
description lacks the "o" in Missouri), was consid-
ered a subspecies ofC. minutissima by Creighton
(1950). Although the two species are similar, they
differ in nest site, in the sculpture of the me-
sopleuron (Creighton 1950), and are sympatric in
a large part of their ranges. A paper on the subge-
nus Orthocrema in eastern North America is in
(6). Species of Dorymyrmex were listed under
the genus Conomyrma in the 1989 list. In the
1989 list, several species of Dorymyrmex reported
from Florida by James Trager (1988) were ex-
cluded from the list. This was a mistake, and an
injustice to Trager's excellent work. Of the seven
species known to occur in Florida, four made their
way into Bolton's catalog, three did not. Extensive
field work in Florida since 1989 has confirmed
that there are at least seven species in Florida.
Whether any of these are identical with species
known from the southwestern U.S. is an open
question until the southwestern species have
been studied as intensively as those in the South-
east. In 2001, I did a little collecting of Dory-
myrmex during a short stay in Arizona. The
results of even so brief an exposure have con-
vinced me that the southwestern Dorymyrmex are
as complex and challenging as they are fascinat-
(7). The subspecies Leptothorax texanus davisi
Wheeler, listed in the 1989 list, was raised to spe-
cies rank by Mackay (2000). In Florida and else-
where davisi does not seem to be recognizable
either as a species or subspecies; a paper dealing
with this and the undescribed Florida Leptotho-
rax listed above is in preparation.
(8). Pheidole obscurithorax is listed in Bolton
(1995) as a subspecies of P fallax Mayr. Deyrup et
al. (2000) treated obscurithorax as a species.
(9). Proceratium crassicorne was synonymized
with P silaceum by Creighton (1950), but Maria
de Andrade will be reviving this species in a forth-
coming revision of the genus; she has sent me a

March 2003

series of identified Florida specimens of both spe-
cies that seem to justify this treatment.
(10). Pyramica apalachicolensis is a recently
described species (Deyrup and Lubertazzi 2001).
(11). Solenopsis invicta is listed as S. wagneri
Santschi in the 1995 catalog; the former name has
been conserved.


Many of the species added to the 1989 list were
found for the first time in Florida by various entomolo-
gists, who recognized them as new records and kindly
sent or brought specimens to me. These collectors I
gratefully list below; the parenthesis after each name
denotes the number of species that each person discov-
ered: Lloyd Davis (7), Stefan Cover (3), Clifford Johnson
(3), David Lubertazzi (3), Zachary Prusak (1), Vincent
Golia (1), Paul Skelley (1). I thank Lloyd Davis and Hi-
lary Swain for detailed and thoughtful comments on the
first draft of this paper.


versity and adaptation in the ant genus Cephaplotes,
past and present. Stuttgarter Beit. Naturkunde, Ser.
B 271: 1-889.
BOLTON, B.1995. A new general catalogue of the ants.
Harvard University Press, Cambridge. 504 pp.
BOLTON, B. 2000. The ant tribe Dacetini. Mem. Ameri-
can Entomol. Inst. 65: 1-1028.
CREIGHTON, W. S. 1950. The ants of North America.
Bull. Mus. Comparative Zool. Harvard 104: 1-585, +
57 plates.
DEYRUP, M., AND L. DAVIS. 1998. A new species ofAph-
aenogaster from upland habitats in Florida. Ento-
mol. News 109: 88-94.
DEYRUP, M., L. DAVIS, AND S. COVER 2000. Exotic ants
in Florida. Trans. American Entomol. Soc. 126: 293-
WHEELER 1989. A preliminary list of the ants of
Florida. Florida Entomol. 72: 91-101.
DEYRUP, M., AND D. LUBERTAZZI. 2001. A new species of
ant (Hymenoptera: Formicidae) from North Florida.
Entomol. News 112: 15-21.
TRAGER, J. C. 1988. A revision of Conomyrma (Hy-
menoptera: Formicidae) from the southeastern
United States, especially Florida, with keys to the
species. Florida Entomol. 71: 11-29.
UMPHREY, G. J. 1996. Morphometric discrimination
among sibling species in the fulva-rudis-texana com-
plex in the ant genus Aphaenogaster (Hymenoptera:
Formicidae). Canadian J. Zool. 74: 528- 559.

Florida Entomologist 86(1)

Cherry: Predators in Florida sugarcane


Everglades Research and Education Center, 3200 E. Palm Beach Road, Belle Glade, FL 33430

Arthropod ground predators were sampled with pitfall traps in Florida sugarcane fields.
More red imported fire ants, Solenopsis invicta Buren, were caught in pitfall traps than all
other predators combined. Sugarcane harvesting did not affect pitfall trap catches of arthro-
pod ground predators. However, replanting reduced arthropod catches for five to six months.
These data show that for most of its three to five year crop cycle, Florida sugarcane is a sta-
ble ecosystem at ground level for arthropod ground predators.

Key Words: Sugarcane, predators, Florida, pitfall traps, ants


Se muestrearon los depredadores artr6podos del suelo con trampas de suelo ("pitfall", tram-
pas donde la presa cae en un hoyo en el suelo) en campos de cana de azucar en Florida. Se
capturaron mas hormigas de fuego importadas, Solenopsis invicta Buren, en las trampas
que todos los otros depredadores juntos. La cantidad de depredadores artr6podos capturados
en las trampas no fue afectada al cosechar la cana de azucar. No obstante, resembrando re-
duj6 la cantidad de artr6podos capturados durante cinco a seis meses. Estos datos muestran
que por la mayor parte de su ciclo de cultivo, de tres a cinco anos, la cana de azucar en Florida
es un ecosistema stable al nivel de suelo para los depredadores artr6podos del suelo.

Sugarcane (Saccharum spp.) is a major field
crop in Florida and is primarily grown in the Ev-
erglades area of southern Florida. Numerous
studies have been published about various biolog-
ical control agents in Florida sugarcane. A list of
many of these studies is provided by Hall (1988).
In a later report, Hall & Bennett (1994) discuss in
greater detail the overall biological control and
IPM of sugarcane pests in Florida sugarcane.
However, there are no published reports on the
population dynamics of arthropod ground preda-
tors in Florida sugarcane. Florida sugarcane is a
long-term crop and few tillage practices are re-
quired over the entire course of a 3 to 5 year plant-
ing (Hall & Bennett 1994). Hence, what the effect
of yearly harvesting and eventual replanting of
sugarcane is on arthropod ground predators is an
interesting question. The objective of this study
was to determine the effects of harvesting and re-
planting on arthropod ground predators in Flor-
ida sugarcane.


Four sugarcane fields in southern Florida were
sampled starting in June, 2000. Two of the fields
were eighteen months old at the start of sampling.
These fields were left in production after harvest
(ratooned) and were used to measure the effect of
harvest on activity of arthropod ground predators.
In this paper, I consider arthropods to be preda-
ceous if they belong to a taxonomic group in which

most members are predaceous. Two of the fields
were three and one half years old at the start of
sampling. These older fields were replanted to
sugarcane (successive planting) after harvest and
were used to measure the effect of replanting on
activity of arthropod ground predators. The two
ratooned fields were harvested during February,
2001. Harvesting consisted of burning the sugar-
cane to remove litter and removal of sugarcane
stalks by mechanical harvesting. The two succes-
sively planted fields were harvested and re-
planted during November, 2000. Harvesting was
as described for ratooned fields. Replanting con-
sisted of fields being disced, sugarcane seedpieces
placed in furrows, Thimet 20G (AI = phorate)
placed in furrows on cane at 4.55 kg AI/hectare,
and then seedpieces covered with soil.
Pitfall trap sampling in all four fields started
June, 2000 and continued until June, 2001. Each
pitfall trap consisted of a nine cm diameter plastic
cup containing 100 ml of ethylene glycol.A five cm
deep plastic collar was also cut from the 9 cm plas-
tic cups. The top of this collar was taped in the
middle of a 26 cm diameter paper plate with it's
center removed. This collar was then inserted into
the pitfall trap and the plate loosely covered with
soil. This arrangement prevented soil subsidence
around the trap rim thus allowing arthropods
easy access to the trap. A small metal roof was
also placed above each trap to prevent rainfall
from filling traps. Five traps were used in each
field. The first trap was located mid-field in a sug-

Florida Entomologist 86(1)

arcane row 50 m into the field to avoid possible
edge effects. The next four traps were placed 5 m
apart in the row into the field. Traps were used for
two weeks each month. After each two week pe-
riod, traps were taken to a laboratory and sam-
ples drained into paper towels and frozen.
Thereafter, ants (Formicidae), earwigs (Der-
maptera), ground beetles (Carabidae), rove bee-
tles (Staphylinidae), spiders (Araneida), and
centipedes (Chilopoda) were counted under a mi-
croscope. Taxonomic determinations of ants and
spiders were made since these were the most
abundant predators found in traps. The relative
abundance of predators in all traps was deter-
mined. For statistical analysis, data from the two
ratooned fields were pooled as were data from the
two replanted fields. The mean monthly catch of
ants, spiders, and total predators in pitfall traps
in ratooned fields and replanted fields was com-
pared using Least Significant Difference (LSD)
tests (SAS 1996).


A total of 4,255 arthropod ground predators
were caught in pitfall traps during the one year
study (Table 1). Of these, the vast majority were
ants being 67.6% of the total catch. Among ants,
the imported fire ant, Solenopsis invicta Buren
was clearly the dominant ant species being 79.2%
of all ants found in traps. These data are consis-
tent with the report of Cherry & Nuessly (1992)

that showed that S. invicta had become the domi-
nant ant species in Florida sugarcane since first
being found there in 1970. In fact, more S. invicta
(2,279) were caught in pitfall traps in this study
than all other predators combined. There is a
wealth of literature on S. invicta as a predator in
sugarcane and other ecosystems and this is re-
viewed by Reagan (1986).
Hall & Bennett (1994) have noted that insect
pests of sugarcane are good candidates for classi-
cal biological control because some pest damage
may be generally tolerated, sugarcane is a long
term crop, and few tillage practices are required
over the entire course of the three to five year
planting. They also note that pre-harvest burning
is the most disruptive practice that may interfere
with biological control. However, the effects of
burning on arthropod populations are complex
and not always predictable. For example, ants
were the most frequently caught predators in this
study and MacKay et al. (1991) noted that fire
may reduce species richness of ants, increase ant
activity, or have no effect on ant populations. Data
in Table 2 show that there were significant differ-
ences in catches of ants, spiders, and total preda-
tor numbers among different months in ratooned
fields. However, there were no significant differ-
ences in catches of these groups in the month im-
mediately preceding harvest and following
harvest in ratooned fields. Also, catches of these
groups during the three month post-harvest pe-
riod were not significantly different than the


Predator Number % of total catch

Ants 2877 67.6
Brachymyrmex obscurior Forel 50 1.2
Monomorium pharaonis (Linn.) 52 1.2
Odontomachus ruginodis Wheeler 96 2.3
Pheidole moerens Wheeler 126 3.0
Solenopsis invicta Buren 2279 53.4
Strumigenys louisianae Roger 46 1.1
Tetramorium simillimum Smith 65 1.5
Wasmannia auropunctata (Roger) 60 1.4
Unknown 109 2.6
Earwigs 252 5.9
Ground Beetles 76 1.8
Rove Beetles 89 2.1
Spiders 913 21.5
Corinnidae 116 2.7
Gnaphosidae 49 1.2
Linyphiidae 69 1.6
Lycosidae 633 14.9
Unknown 46 1.1
Centipedes 48 1.1
Total 4255 100.0

March 2003

Cherry: Predators in Florida sugarcane



Month Ants Spiders Total2

June 2000 30.5 + 34.8 A 3.1 +1.5 BC 34.5 + 35.9 A
July 19.1 + 19.4 AB 13.2 + 17.8 A 35.0 + 29.5 A
August 7.3 + 8.5 BC 5.4 + 4.8 BC 13.0 + 10.1 B
September 7.3 + 11.2 BC 8.1 + 4.7 AB 16.8 + 14.7 B
October 2.3 + 2.8 C 3.6 + 2.9 BC 5.9 + 4.8 B
November 1.1 + 1.4 C 3.3 + 6.3 BC 4.4 + 7.2 B
December 2.7 + 1.9 C 1.8 + 1.8 C 5.0 + 2.9 B
January 2001 1.7 + 2.7 C 1.8 + 1.9 C 3.5 + 4.3 B
February Harvest Harvest Harvest
March 9.2 + 10.9 BC 1.9 + 1.6 C 13.7 + 11.6 B
April 1.8 + 1.5 C 4.1 + 2.9 BC 6.4 + 4.1 B
May 4.0 + 4.7 C 2.2 + 1.9 C 10.7 + 6.8 B

'Mean + SD. Means in a column followed by the same letter are not significantly different (alpha = 0.05) using the LSD test (SAS 1996).
2 Total predators = all predators noted in Table 1.

three month pre-harvest period. These data show planting is more disruptive to the soil habitat
that the sugarcane harvesting, including the than harvesting because replanting involves not
burning of the fields, did not reduce overall activ- only burning of the field and mechanical harvest-
ity of ants, spiders, or total predator number in ra- ing, but also discing, and the use of a soil insecti-
tooned fields. cide.
Predator catches in pitfall traps in successively To summarize, my data show that sugarcane
planted fields of Florida sugarcane are shown in harvesting had no significant effect on total num-
Table 3. Pitfall trap catches of ants, spiders, and bers of arthropod ground predators caught in pit-
total predators all decreased in the month follow- fall traps. In contrast, replanting significantly
ing replanting versus the month immediately be- reduced total numbers of ground predators in pit-
fore replanting. Also, total predator catches fall traps, but these numbers resurged after 5 to 6
remained low for the first four months after re- months to preharvest levels. These data show
planting compared to pre-planting catches and that through most of its 3 to 5 year crop cycle,
then increased dramatically at five to six months Florida sugarcane is a stable ecosystem at ground
after planting. These data make sense since re- level for most arthropod ground predators.



Month Ants Spiders Total2

June 2000 61.6 + 120.9 A 5.7 + 3.2 CDE 69.0 + 120.3 A
July 16.5 + 19.8 B 11.7 + 6.9 BC 34.7 + 28.0 ABC
August 9.5 + 11.9 B 8.2 + 6.6 BCD 9.4 + 17.0 BC
September 10.6 + 14.2 B 13.2 + 12.1 B 27.1 + 18.3 BC
October 21.3 + 42.0 B 21.6 + 18.6 A 44.3 + 48.2 AB
November Replant Replant Replant
December 2.4 + 2.3 B 2.2 + 1.7 DE 5.4 + 3.5 C
January 2001 1.6 + 1.8 B 1.2 + 1.2 E 3.0 + 2.5 C
February 1.3 + 1.6 B 0.7 + 0.8 E 2.7 + 2.1 C
March 2.1 + 2.6 B 0.4 + 0.7 E 4.5 + 2.6 C
April 14.9 + 32.3 B 2.0 + 2.0 DE 18.3 + 32.5 BC
May 32.3 + 43.9 AB 0.9 + 1.0 E 48.2 + 39.7 AB

'Mean + SD. Means in a column followed by the same letter are not significantly different (alpha = 0.05) using the LSD test (SAS 1996).
2 Total predators = all predators noted in Table 1.

52 Florida Ento


This research was supported by the Florida Agricul-
tural Experiment Station, and approved for publication
as Journal Series No. R-08912. I thank Dr. G.B. Ed-
wards of the Division of Plant Industry, Gainesville, FL
for help in identifying spiders.


CHERRY, R., AND G. NUESSLY. 1992. Distribution and
abundance of imported fire ants (Hymenoptera: For-
micidae) in Florida sugarcane fields. Environ. Ento-
mol. 21: 767-770.
HALL, D. 1988. Insects and mites associated with sugar-
cane in Florida. Florida Entomol. 71: 138-150.


ologist 86(1) March 2003

HALL, D., AND F. BENNETT. 1994. Biological control and
IPM of sugarcane pests in Florida, pp. 297-325. In D.
Rosen, F. Bennett, and J. Capinera (eds.). Pest man-
agement in the subtropics: biological control A Flor-
ida perspective. Intercept. Paris.
D. GONZALES, AND S. VINSON. 1991. Impact of the
slashing and burning of a tropical rain forest on the
native ant fauna (Hymenoptera: Formicidae). Socio-
biology. 18: 257-268.
REAGAN, T. 1986. Beneficial aspects of the imported fire
ant: a field ecology approach, pp. 58-71. In C. Lofgren
and R. Vander Meer (eds.). Fire ants and Leaf-cut-
ting ants biology and management. Westview
press. London.
SAS INSTITUTE. 1996. SAS Systems for Windows. Ver-
sion 6.12. SAS Institute, Cary, NC.

Lopez et al.: Medfly Parasitoid


1Programa MOSCAMED, Laboratorio la Aurora, Avenida Hincapie y 18 Calle Zona 13
Cuidad de Guatemala 01013, Guatemala

2USDA-ARS, CMAVE, P.O. Box 14565, Gainesville, FL 32604

3USDA-APHIS-PPQ-CPHST, 4a Avenida Zona 10, Cuidad de Guatemala 01013, Guatemala

4USDA-APHIS-PPQ-CPHST, 1913 SW 34th St., Gainesville, FL 32608

5USDA-APHIS-PPQ-CPHST-NBCI, Center for Biological Control, Florida A&M University, Tallahassee, FL 32307

6International Centre of Insect Physiology and Ecology, P.O. Box 30772, Nairobi, Kenya

7Department of Entomology, Texas A&M University, College Station, TX 77843

'Instituto de Ecologia A. C., Apdo. Postal 63, 91000 Xalapa, Veracruz, Mexico


Fopius ceratitivorus Wharton is a recently discovered braconid parasitoid of the Mediterra-
nean fruit fly (= medfly), Ceratitis capitata (Wied.). Unlike other parasitoids previously used
in medfly biological control, F. ceratitivorus was originally collected from medfly in its pur-
ported region of origin, east Africa. Shipments of Ceratitis spp. pupae from Kenya to a newly
constructed quarantine facility in Guatemala yielded both F. ceratitivorus and its congener
F. caudatus (Szepligeti). Only the former species was successfully colonized through the use
of medfly infested coffee berries. In the process of colonization it was determined that F. cer-
atitivorus oviposited into the eggs and recently hatched larvae of medflies and completed de-
velopment in the hosts' puparia. This is a relatively rare behavior among fruit fly parasitoids
and, because tephritid eggs near the surface of fruits are particularly vulnerable to attack,
one that might contribute to its success as a biological control agent.

Key Words: biological control, mass-rearing, medfly


F ceratitivorus Wharton es un parasitoide Braconido de la mosca del Mediterraneo (= mos-
camed), Ceratitis capitata (Wied.), recientemente descubierto. A diferencia de otros parasi-
toides previamente usados en el control biol6gico de la moscamed, F. ceratitivorus fue
colectado originalmente de moscamed en su supuesta region de origen, al este de Africa. En-
vios de pupa de tephritidos desde Kenia hacia la recientemente construida instalaci6n de
Cuarentena en Guatemala, produjeron especimenes deF. ceratitivorus y su congener F. cau-
datus (Szepligeti). Solo la primera especie fue colonizada exitosamente mediante el uso de
frutos de caf6 infestados por moscamed. En el process de colonizaci6n se determine que F. ce-
ratitivorus oviposita sobre los huevos y larvas recientemente eclosionadas de moscamed, y
que complete su desarrollo en la pupa hu6sped. Este es un comportamiento relativamente
raro dentro de los parasitoides de moscas de la fruta, y debido a que los huevos de los tephri-
tidos cercanos a la superficie del fruto son particularmente vulnerable, ello podria contri-
buir a su 6xito como agent de control biol6gico.

By the end of the 19th century the Mediter- (Wied.), had spread from its African homeland
ranean fruit fly (= medfly), Ceratitis capitata to tropical and subtropical countries around the

Florida Entomologist 86(1)

world. After finding medfly in the Honolulu area
in 1910, the progressive Hawaiian agricultural
community of the time financed an African col-
lection oftephritid natural enemies by Silvestri
(1914) in order to bring the fly under biological
control. By 1918 there were several parasitoids
from various parts of the world established in
Hawaii, including three species of opiine Bra-
conidae (Pemberton & Willard 1918). Over time,
subsequent expeditions resulted in additional
Hawaiian establishments (Gilstrap & Hart
1987), the most effective of which for suppres-
sion of both medfly and oriental fruit fly (Bac-
trocera dorsalis [Hendel]) proved to be the
braconid Fopius arisanus (Sonan) (Bess et al.
While F arisanus is a common parasitoid of
medfly in Hawaii, it is an Asian species that was
originally obtained from the pupae of oriental
fruit fly (Wharton & Gilstrap 1983). In fact, to
our knowledge, none of the braconids success-
fully disseminated for the control of medfly orig-
inated from collections of medfly (Wharton &
Gilstrap 1983; Ovruski et al. 2000). This short-
age of "true-medfly" parasitoids is not due to a
lack of candidates since a recent Kenyan survey
of Ceratitis spp. yielded 10 species of hy-
menopterous parasitoids (Wharton et al. 2000;
see also Steck et al. 1986), but probably reflects
the historical difficulty of transporting live in-
sects from Africa to afflicted agricultural areas
such as Hawaii or Central America (e.g., van
Zwaluwenburg 1937).
We here describe the shipment to Guatemala
and subsequent colonization of Fopius cerati-
tivorus Wharton, a recently discovered parasitoid
of the medfly that is both a true, African natural
enemy of medfly and, like its congener F arisa-
nus, an egg-pupal parasitoid. This combination of
characteristics suggests that this species may be a
particularly attractive candidate for biological


Origin of Insects

Fopius ceratitivorus has been obtained only
from coffee, Coffea arabica L., in central Kenya
and in particular from plantations at Ruiru
(15.72'S, 36054.22'E at 1609 m) and Rurima
(0038.39'S,37029.69'E at 1228 m) (Wharton 1999;
Wharton et al. 2000). Mean annual rainfalls in
these areas are 1.06 m and 0.9 m, respectively,
and the mean temperature ranges are 12.8-25
and 15.5-280C, respectively (Wharton et al. 2000).
Collections were made throughout the November-
July coffee harvest season. The tephritids in the
shipments, in order of abundance, were: C. capi-
tata, C. rosa Karsch, and Trirhithrum coffeae
Bezzi (Wharton 1999).

Insect Arrival

Field collection procedures and handling pro-
cedures were described by Wharton et al. (2000).
Pupae were shipped by air in lots of 4,000-23,000
insects to the Guatemala International Airport,
cleared through customs, and then brought by car
to the USDA-APHIS/MOSCAMED quarantine fa-
cility at San Miguel Petapa, Guatemala (Table 1).

Quarantine Facility

Packages were brought to the USDA-APHIS /
MOSCAMED quarantine facility at San Miguel
Petapa outside of Guatemala City, Guatemala.
Initially, packages holding pupae were removed
from containers in a large (0.8 x 0.8 x 0.8 m)
sleeved cage separated from the remainder of the
quarantine facility by a locked door. Adult parasi-
toids were captured individually and transferred
to smaller (21 x 21 x 21 cm) cages containing
honey and a water-wick while adult flies were
placed in 70% ethanol and preserved for latter ex-



Date Collection No. Pupae % Parasitism F. ceratitivorus F. caudatus D. fullaway

9/19/2000 Koru 4,310 4.11 0 177 0
10/31/2000 Koru 2.14 0 215 0
10/31/2000 Ruiru 10,052 0.73 73 0 0
12/22/2000 Koru 0.46 0 95 8
12/22/2000 Ruiru 22,507 1.57 290 0 63
12/22/200 Rurima 0.35 74 0 5
6/26/2001 Koru 15,645 12.77 0 1,998 0
12/21/2001 Koru 0.20 0 16 0
12/21/2001 Ruiru 8,043 3.68 221 0 75

March 2003

Lopez et al.: Medfly Parasitoid

amination. Caged parasitoids were then moved
into a larger room within the quarantine facility
that had both windows providing natural light
and artificial lighting (12L: 12D). Temperature in
this room was 26C and relative humidity 65-
75%. All packaging materials and biological
wastes were sterilized in an autoclave before re-
moval from the quarantine building. Specimens of
all insect species received were preserved in col-
lections at the quarantine facility.

Presentation of Hosts

Several means of host presentation were de-
veloped, including artificially placing eggs into
slits cut through the skin and pulp of coffee ber-
ries. However, the most practical and effective
means of presenting hosts to F. ceratitivorus con-
sisted of first allowing female medflies to oviposit
into firm coffee berries that were mature to the
point of color-break. In addition to conserving any
host-location cues the ovipositing medfly might
leave, this technique minimized fermentation
during the presentation. High densities of med-
flies (~4,000 males and 4,000 females) were kept
in 31 x 31 x 31 cm screen cages and allowed to lay
eggs in varying numbers of berries for a period of
24 hours. The berries had been previously strung
on thread to form "necklaces" of ~180 fruits and
then suspended from the ceiling of the cage.
Strings of berries were then transferred to screen
cages that contained ~600 female parasitoids and
a similar number of males. There was an attempt
to present infested fruit at a ratio of 3 fruit / fe-
male (~120 host egg clutches / female parasitoid).
Females at the time of first presentation were ~ 8
days of age and had been kept in the presence of
similar aged males since eclosion. During the first
2 days of this mating period cages were placed
near windows to approximate a natural light en-
vironment. Over the next 6 days the cages were
kept under full spectrum lights.
Typically, infested fruits were exposed to para-
sitoids for 48 hours; however, if berries were
small and prone to drying exposures were cur-
tailed after 24 hours. Depending on coffee avail-
ability there were 2 to 4 exposures / cage / week.
Female survived in the exposure cages for up to
45 days, with 50% percent of females were typi-
cally alive after 30 days. Male lifespans were
lower in quarantine, with 50% dead after only 15
In addition to coffee, medflies were allowed to
oviposit in several other fruits that were subse-
quently exposed to F. ceratitvorus : mangos
(Mangifera indica L. var. Tommy Atkins), Spon-
dias sp., papaya (Carica papaya L.), apple (Malus
pumila Mill.), peach (Prunus persica Batsch), and
pear (Pyrus communis L.). The treatment of these
fruits following exposure was as described above
for coffee berries.

Holding of Larvae

Following either 24 or 48 hours of exposure to
parasitoids, 180-360 berries were placed in 30 x
15 cm trays on dampened paper over a 0.5 kg of
moist medfly diet obtained from the USDA-
APHIS / MOSCAMED rearing facility at El Pino,
Guatemala (Schwarz et al. 1985). After spraying
the berries lightly with water, the trays were
placed in a 0.6 x 0.4 x 1.0 m rack covered with a
black plastic sheet, which allowed temperature
and relative humidity to increase to 26-27C and
98-99%, respectively. After 48 hours, the berries
were mixed into an additional 0.5kg of diet. Trays
were moved into a cooler room at 23C and 65-
75% RH, and held over sawdust for 13 days (day
19 of the process). At the end of this time, mature
larvae had left the diet and completed pupation in
the sawdust. Puparia were placed in the 31 cm x
31 cm x 31 cm cages and emerging medflies re-
moved through aspiration. Fopius ceratitivorus
males began to eclose on day 24 and females on
day 26.

Host Stage Parasitized

Standard rearing methods enabled oviposition
into both eggs and newly hatched larvae. How-
ever, they did not distinguish between these
stages. To better determine what stages) of med-
fly was being parasitized we varied parasitoid-ex-
posure schedules to present either eggs alone or
larvae alone (Fig. 1). In one case (Fig 1; A), fruit
with eggs were presented 21 hours after the ini-
tial exposure to medfly females and then removed
after 24 hours. This eliminated the possibility
that early-instar larvae where present. In the sec-
ond case (Fig. 1; B), 69 hours elapsed between the
initiation of oviposition and exposure of the fruit
to parasitoids, and only larvae were available as
hosts. In the third case (Fig.l; C), the standard ex-
posure sequence was modified so that parasitism
was limited to a 24 hour period rather than the
usual 48 hours. This again resulted in only eggs
being present during attacks. D represents the
standard sequence where both eggs and early in-
star larvae are potentially present. There were 3
replications of each of the exposure regimes.


Host Stage Parasitized

In the course of the standard rearing proce-
dure, infested berries were removed from expo-
sure to parasitoids prior to or just following egg
hatch and larvae were rarely observed when ber-
ries were first placed on dampened paper over
diet. To better determine the stage of medfly being
attacked the exposure procedure was modified to
expose only eggs or only recently hatched first in-

Florida Entomologist 86(1)


I --

Fig. 1. The timing of exposure of coffee berries con-
taining Mediterranean fruit fly eggs and larvae to the
parasitoid Fopius ceratitivorus. The various schedules
resulted in either eggs (A&C), first instar larvae (B) or
both (D) being open to attack. Dark bars refer to time
spent in the various activities of preparation of fruit to
be exposed to medflies ("prepertion"), exposure of fruit
to medflies ("infestation"), the period of egg availability
("egg"), the period during which eggs hatch ("larvae"),
and the period of exposure to parasitoids parasitismsm").
The time line in hours is at the top of the chart and the
light and shaded areas represent alternating periods of
light and darkness.





--.-.-.--.-, -.-

star larvae to parasitoids. Both stages were vul-
nerable to attack (Table 2).
In addition to F ceratitivorus, the Kenyan
shipments contained other Braconidae including
Fopius caudatus (Szepligeti). Attempts to rear F
caudatus were unsuccessful, although colonies
were sometimes maintained for up to 6 genera-
tions before increasingly male-biased sex ratios
resulted in collapse. Fopius caudatus was also an
egg-pupal parasitoid and when medfly eggs were
presented in slits cut in coffee berries it was rela-
tively easy to observe the penetration of the host
egg by the parasitoid's ovipositor. Its capacity to
attack early instar larvae is unknown.

Percent Parasitism and Colony Growth

The F. ceratitivorus colony increased over time
until at present (April 2002) weekly production
was 10,000-18,000 adults/week or roughly 2-3
adult parasitoids per berry (Fig. 2). Overall, per-
cent parasitism was 3.5-4%, but was occasionally
as high as 21%. Typical sex ratios approximated 1
male:l female, but were sometimes strongly fe-
male or male biased (Fig. 2). For example, in the
experiments to identify stage of host attacked
only 37% of the adult parasitoids were male. In
part, fluctuations in numbers reflected the sea-
sonal changes in the abundance of coffee berries
used in the rearing process. Fopius ceratitivorus
was capable of parasitizing medfly in a variety of
fruit species other than coffee (Table 3), and these
may be integrated into the rearing program in the


The establishment of F arisanus in Hawaii is
arguably the most successful instance of fruit fly
biological control in the world (e.g., Clausen
1978), and it would be useful to employ other par-
asitoids that possess the characteristics that have
made F arisanus so prominent among Hawaiian
fruit fly natural enemies. Certainly one the most
unusual attributes of F arisanus is that it is an
endoparasitic koinobiont that oviposits into the
egg, rather than the larva, of its tephritid host
(Wharton 1997)). The larval parasitoid persists in
the first instar until the host's puparium is
formed after which it completes its development
(Clausen 1978).The ability to parasitize eggs, as
do F arisanus, ceratitivorus and caudatus, is oth-
erwise rare among fruit fly parasitoids. We are
aware of only one other species known to do so,
Utetes canaliculatus (Gahan), a North American
parasitoid of Rhagoletis and another opiine bra-
conid (Prokopy & Webster 1978).
Several reasons have been proposed why egg-
parasitism might account for the success of F
arisanus, including its early presence inside the
host compared to other braconids that attack var-

March 2003

Lopez et al.: Medfly Parasitoid


Egg Larvae Standard Egg Standard Egg & Larvae
Exposed intervals (h) Replicates 21-45 69-93 24-48 24-72

Adults reared 1 30 7 24 248
2 143 51 21 92
3 72 7 123 333
Pupae recovered (cc) 1 120 280 103 153
2 400 250 68 62
3 65 80 98 110
% Parasitism 1 0.4 0.0 0.4 2.6
2 0.8 0.3 0.5 2.5
3 1.9 0.1 2.1 5.1

ious larval instars (Bess et al. 1961). Because it is
the first parasitoid present, a F arisanus larva
would be in a position to eliminate or suppress the
growth of its competitors. In addition, tephritid
eggs located near the surface of a fruit or vegeta-
ble are particularly vulnerable to parasitism.
Fruit fly larvae that feed in the pulp or seeds of
fruit can be difficult for parasitoids to reach with
their ovipositors, and there is a well established
negative relationship between fruit size and par-
asitism by larva-attacking braconids (e.g., Sivin-
ski et al. 1997, Lopez et al. 1999).
Thus the capacity to parasitize vulnerable
eggs and early instar larvae is potentially a valu-
able trait for a biological control agent (Bess et al.
1961), and medfly control in Central America
may particularly benefit from the availability of
an effective natural enemy. At present, there is
little parasitism of medfly in the New World by
native tephritid parasitoids, and only local and
sporadic parasitism by introduced species such
as Diachasmimorpha longicaudata (Ashmead)
(Eskafi 1990; Sivinski et al. unpublished data).
Unlike its exemplary performance in Hawaii, F
arisanus has either failed to become established
in the Americas (Ovruski et al. 2000) or failed to
flourish after establishment (Wharton et al.
In addition to being a potential candidate for
establishment, F ceratitivorus might prove to be
important in regional eradication programs. Med-
fly is now widely distributed across the Latin
American tropics and subtropics. The northward
spread of medfly into Mexico, and ultimately into
the United States, has been prevented by a Sterile
Insect Technique (= SIT) / insecticide-bait spray
barrier maintained along the Mexican / Guatema-

lan border by the international organization
MOSCAMED (United States, Mexico, and Guate-
mala). Recently, this barrier has been expanded
and the possibility of regional eradication of the
medfly is under consideration (P. Rendon, per-
sonal communication.).
In a region-wide eradication program the
medfly must be attacked in a variety of environ-
ments, some of which may not be amenable to re-
peated applications of insecticide-bait sprays,
such as organic growing areas, urban / suburban
locations, water-sheds, and national parks. In
these areas it will be important to maximize the
impact of the biological components of the vari-
ous control options. There is accumulating evi-
dence that augmentative parasitoid releases may
be an efficacious means of suppressing fruit fly
populations, perhaps to a level where SIT can
then be used to complete eradication (e.g., Wong
et al. 1991, Sivinski et al. 1996, Montoya et al.
The potential of F. ceratitivorus for mass-
rearing and augmentative release is unknown.
The mean parasitism rate in the present colony
of ~4% is an order of magnitude or more lower
than the laboratory parasitism rates of better
established medfly parasitoids (e.g., Baeza et al.
2002). However, experience suggests that
greater familiarity with the species' require-
ments will improve production. In the mean-
time a stable colony in Guatemala will allow the
experiments to be accomplished that will clarify
its usefulness in medfly biological control.
These include determination of host range and
capacity to persist through seasonal declines in
Guatemalan medfly populations when eggs are

Florida Entomologist 86(1)








12 I




4 O


=1 2001
-- F. ceratitivorus
SF. caudatus

0 0- '" Ile

5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65

i 1991
Il 1992
-- F. ceratitivorus

5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65 68


Fig 2. The production, % parasitism, and sex-ratio of Fopius ceratitivorus and F caudatus in the Guatemalan
quarantine facility over time. Sharp declines are typically due to temporary shortages of coffee berries.

5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65 68

March 2003


Lopez et al.: Medfly Parasitoid


Emerged Parasitoids
Number Pupae % J: 9
Host Fruit of Fruit per Fruit (ml) S 2 Parasitism Ratio

Mango 60 20 2,398 1,359 5.3 1.8
Pear 46 40 4,254 2,593 6.4 1.6
Coffee 1,260 1 1,511 1,343 3.3 1.1
Spondias sp. 32 2 235 99 7.2 2.4
Papaya 12 9 226 84 4.7 2.7
Peach 1 3 7 0 4.0
Apple 1 8 0 0

Jarvi Esquit6 was directly in charge of the numerous
attempts to develop a successful rearing technique. The
ideas and support of the personnel of La Aurora Parasi-
toid Rearing Facility were also critical for the coloniza-
tion. We would like to thank those who were
instrumental in the explorations that resulted in the
discovery and subsequent collections of F ceratitivorus
in Kenya: Slawomir Lux and Willam Overholt of the In-
ternational Centre of Insect Physiology and Ecology
(Nairobi, Kenya), Robert Wharton (Texas A&M Univer-
sity; USDA/NRI Grant no. 9703184), Russell Messing
(University of Hawaii; USDA-CREES Special Grant no.
96-34135) and Richard Baranowski, University of Flor-
ida; Caribbean Basin Administrative Group Grant no.
96-34135-3016). In Guatemala, Gustavo Baeza (MOS-
CAMED) oversaw the construction of the quarantine fa-
cility and Gordon Tween (USDA-APHIS-IS) provided
funds when they were most needed. Without Rony Ro-
das' (MOSCAMED, Coatepeque, Guatemala) collections
of coffee berries of the correct quantity and quality noth-
ing could have been accomplished. Gina Posey prepared
the illustrations and Valerie Malcolm the manuscript.
The use of trade, firm, or corporation names in this
publication is for the information and convenience of
the reader. Such use does not constitute an official en-
dorsement or approval by the United States Depart-
ment of Agriculture or the Agricultural Research
Service of any product or service to the exclusion of oth-
ers that may be suitable.

The effects of chilling on the fecundity and life span
of mass-reared parasitoids (Hymenoptera: Bra-
conidae) of the Mediterranean fruit fly, Ceratitis cap-
itata (Wiedemann) (Diptera: Tephritidae). Biocont.
Sci and Tech. 12: 205-215.
Fruit fly parasites and their activities in Hawaii. Ha-
waiian Entomol. Soc. 17: 367-378.
CLAUSEN, C. 1978. Tephritidae, pp. 320-335 In B. Bar-
tlett et al. (eds.) Introduced Parasites and Predators
of Arthropod Pests and Weeds: a World Review.
USDA Agr. Handbook No. 480.
ESKAFI, F. 1990. Parasitism of fruit flies Ceratitis capi-
tata and Anastrepha spp. (Diptera: Tephritidae) in
Guatemala. Entomophaga 35: 355-362.

GILSTRAP, F., AND W. HART. 1987. Biological control of
the Mediterranean fruit fly in the United States and
Central America. USDA ARS-56.
menopterous larval-pupal and pupal parasitoids of
Anastrepha flies (Diptera: Tephritidae) in Mexico.
Biol. Cont. 15: 119-129.
RERA, J. SIVINSKI, AND M. ALUJA. 2000. Biological
control ofAnastrepha spp. (Diptera: Tephritidae) in
mango orchards through augmentative releases of
Diachasmimorpha longicaudata (Ashmead) (Hy-
menoptera: Braconidae). Biol. Cont. 18: 216-224.
2000. Hymenopteran parasitoids on fruit- infesting
Tephritidae (Diptera) in Latin America and the
southern United States: diversity, distribution, taxo-
nomic status and their use in biological control. In-
teg. Pest Manag. Rev. 5: 81-107.
PEMBERTON, C., AND H. WILLARD. 1918. A contribution
to the biology of fruit-fly parasites in Hawaii. J. Agr.
Res. 8: 419-465.
PROKOPY, R., AND R. WEBSTER 1978. Oviposition-deter-
ring pheromone of Rhagoletis pomonella, a
kairomone for its parasitoid Opius lectus. J. Chem.
Ecol. 4: 481-494
1985. Mass production of the Mediterranean fruit fly
at Metapa, Mexico. Florida Entomol. 68: 467-477.
SILVESTRI, F. 1914. Report of an expedition to Africa in
search of the natural enemies of fruit flies (Trypa-
neidae). Terr. Hawaii Brd. Agr. & Forest. Bull. No. 3.
SIVINSKI, J., M. ALUJA, AND M. LOPEZ. 1997. Spatial and
temporal distributions of parasitoids on Mexican
Anastrepha spp. (Diptera: Tephritidae) within the
canopies of fruit trees. Ann. Entomol. Soc. America
90: 604-618.
G. DODSON. 1996. Suppression of a Caribbean fruit
fly (Anastrepha suspense [Loew] Diptera: Tephriti-
dae) population through augmented releases of the
parasitoid Diachasmimorpha longicaudata (Ash-
mead) (Hymenoptera: Braconidae). Biol. Cont. 6:
1986. Braconid parasitoids of Tephritidae (Diptera)
infesting coffee and other fruits in west-central Af-
rica. Entomophaga 31: 59-67.

60 Florida Entc

VAN ZWALUWENBURG, R. 1937. West African notes. Ha-
waiian Planters' Rec. 41: 57-83. 16
WHARTON, R. 1997. Generic relationships of Opiine Bra-
conidae (Hymenoptera) parasitic on fruit- infesting
Tephritidae (Diptera). Cont. American Entomol.
Inst. Vol. 30, No. 3.
WHARTON, R. 1999. A review of the Old World genus
Fopius Wharton (Hymenoptera: Braconidae: Opii-
nae), with description of two new species reared from
fruit-infesting Tephritidae (Diptera). J. Hym. Res. 8:
WHARTON, R., AND F. GILSTRAP. 1983. Key to and status
of opiine braconid (Hymenoptera) parasitoids used in
biological control of Ceratitis and Dacus s. 1.(Diptera:
Tephritidae). Ann. Entomol. Soc. America 76: 721-


ologist 86(1) March 2003

W. HART. 1981. Hymenopterous egg-pupal and lar-
val-pupal parasitoids of Ceratitis capitata andAnas-
trepha spp. (Diptera: Tephritidae) in Costa Rica.
Entomophaga 26: 285-290.
HAMED, AND J. SIVINSKI. 2000. Parasitoids of medfly,
Ceratitis capitata, and related tephritids in Kenyan
coffee: a predominantly koinobiont assemblage.
Bull. Entomol. Res. 90: 517-526.
J. NISHITO, AND J. HERR 1991. Augmentative re-
leases of Diachasmimorpha tryoni (Hymenoptera:
Braconidae) to suppress a Mediterranean fruit fly
(Diptera: Tephritidae) population in Kula, Maui, Ha-
waii. Biol. Cont. 1: 2-7.

Stuart et al: Predation on Diaprepes Neonates


University of Florida, IFAS, Citrus Research and Education Center
700 Experiment Station Road, Lake Alfred, FL 33850


The root weevil, Diaprepes abbreviatus (L.), is a major pest of Florida citrus. When neonate
larvae hatch from egg masses in the citrus canopy and drop to the soil surface before bur-
rowing down to the roots for feeding, they are vulnerable to ant predation. However, neo-
nates are reported to produce a chemical repellent that lasts up to four days and reduces ant
predation by about 40%. We assessed the daily pattern of neonate drop from egg masses un-
der laboratory conditions (24 C, 70% RH, L:D = 12:12), examined the role of ants as preda-
tors of neonates (<48 h post hatch) on the soil surface in three citrus groves in central
Florida, and tested for chemical repellency in the field by comparing predation rates on 5-
day versus 1-2 h old neonates. Neonate drop was not well synchronized within or among egg
masses, occurred during all hours of the light and dark phases, and extended over 5 to 23 h
(mean = 11.97, SE = 0.866) for individual egg masses (n = 29). However, the drop rate was
highest during the second half of the light phase (52.4%) and lowest during the second half
of the dark phase (8.0%). Predation occurred in 104 of 199 replicates (52.3%) in the three
groves with a total of 475 of the 3980 larvae (11.9%) removed by predators within 20 mins.
Predation pressure varied within and among groves, and involved eight ant species (Hy-
menoptera: Formicidae) and a single predation event by a nymph of the big-eyed bug, Geo-
coris floridanus Blatchley (Hemiptera: Lygaeidae). For data pooled among groves, the red
imported fire ant, Solenopsis invicta Buren, was responsible for 29.5% of the predation,
Pheidole moerens Wheeler 27.8%, Dorymyrmex reginicula (Trager) 9.7%, Brachymyrmex ob-
scurior Forel 8.8%, Dorymyrmex bureni (Trager) 8.6%, Cardiocondyla emeryi Forel 8.0%,
Paratrechina bourbonica Forel 4.8%, and Pheidole morrisi Forel 2.5%. In our test for age-de-
pendent chemical repellency, a total of 2620 of the 3840 neonates (68.2%) were preyed upon
within 30 mins but the predation rate on old versus young neonates did not differ at 68.8%
and 67.7%, respectively. In this experiment, 368 of the predation events (14.0%) were ob-
served directly with P. moerens responsible for 62.5%, S. invicta 25.3%, C. emeryi 10.1%,
B. obscurior 1.4%, Cardiocondyla wroughtonii (Forel) 0.5%, and D. bureni 0.3%. We conclude
that ants are important predators of Diaprepes neonates in central Florida citrus groves and
have potential for a conservation biological control program.

Key Words: predation, ants, biological control, integrated pest management, red imported
fire ant, Curculionidae, Formicidae


El picudo de la raiz, Diaprepes abbreviatus (L.), es una plaga important de citricos en la Flo-
rida. Cuando las larvas neonatas (reci6n nacidas) esclosionan de las masas de huevos en la
copa de los arboles de citricos y caen a la superficie del suelo escavando hacia abajo en las
raizes para alimentarse, ellas estan vulnerable a la depredaci6n por las hormigas. No obs-
tante, se reportan que las neonatas produce un quimico repelente que dura hasta cuatro
dias y reduce la depredaci6n por las hormigas por aproximadamente 40%. Nosotros evalua-
mos el patron diario de la caida de las neonatas de las masas de huevos bajo condiciones en
el laboratorio (24 C, 70% RH, L:D = 12:12) [RH = Humedad Relativa; L:D = Luz:Oscuridad],
examinamos el papel de las hormigas como depredadores de las neonatas (<48 h despu6s de
esclosi6nar) sobre la superficie del suelo en tres huertos de citricos en central Florida, y pro-
bamos la abilidad del quimico para repelar en el campo comparando las tasas de depredaci6n
sobre las neonatas de 5-dias versus 1-2 horas de edad. La caida de las neonatas no fu6 bien
sincronizada dentro de la misma o en diferentes masas de huevos, mas ocurri6 durante todas
las horas de las fases de luz y oscuridad, y se extendi6 de 5 a 23 horas (promedio = 11.97,
SE = 0.866) para las masas de huevos individuals (n = 29). No obstante, la tasa de la caida
fu6 la mas alta durante la segunda mitad de la fase de luz (52.4%) y la mas baja durante la
segunda mitad de la fase (8.0%). La depredaci6n ocurri6 en 104 de las 199 replicas (52.3%)
en tres huertos con un total de 475 de los 3980 larvas (11.9%) eliminadas por los depredado-
res dentro de 20 minutes. La presi6n de los depredadores vari6 dentro del mismo y en dife-

Florida Entomologist 86(1)

rentes huertos, y abarc6 ocho species de hormigas (Hymenoptera: Formicidae) y un solo
event de depredaci6n por parte de una ninfa del chinche de ojos grandes, Geocoris flori-
danus Blatchley (Hemiptera: Lygaeidae). Por los datos colectados entire todos los huertos, la
hormiga de fuego importada, Solenopsis invicta Buren, fu6 responsible por el 29.5% de la de-
predaci6n, Pheidole moerens Wheeler 27.8%, Dorymyrmex reginicula (Trager) 9.7%, Bra-
chymyrmex obscurior Forel 8.8%, Dorymyrmex bureni (Trager) 8.6%, Cardiocondyla emeryi
Forel 8.0%, Paratrechina bourbonica Forel 4.8%, y Pheidole morrisi Forel 2.5%. En nuestra
prueba de la repelencia quimico dependiendo de la edad, un total de 2620 de los 3840 neo-
natas (68.2%) fueron atacadas dentro de 30 minutes pero la tasa de depredaci6n sobre neo-
natas viejas versus neonatas jovenes no fu6 diferente siendo 68.8% y 67.7%,
respectivamente. En este experiment, 368 de los events de depredaci6n (14.0%) fueron ob-
servados directamente con P. moerens responsible por 62.5%, S. invicta 25.3%, C. emeryi
10.1%, B. obscurior 1.4%, Cardiocondyla wroughtonii (Forel) 0.5%, y D. bureni 0.3%. Noso-
tros concluimos que las hormigas son depredadores importantes de neonatas de Diaprepes
en los huertos de citricos en central Florida y tienen potential en un program de conserva-
ci6n de control biol6gico.

The root weevil, Diaprepes abbreviatus (L.)
(Coleoptera: Curculionidae), is a major pest of cit-
rus, ornamentals, and other crops, and has spread
widely in Florida since it was first detected in
1964 (Graham et al. 1996, McCoy 1999, McCoy et
al. 2001).Adults are long lived and feed on foliage,
especially new growth. Mating occurs in the can-
opy, and eggs are laid in masses between leaves
that are glued together by an adhesive secreted by
the female during oviposition. The larvae hatch,
escape from the sealed leaf envelope, drop to the
soil, and burrow down to the roots where they be-
gin feeding. As they grow, the larvae move to
larger roots, and pupate in the soil after 9-11 in-
stars (Woodruff 1985, Quintela et al. 1998, McCoy
1999). In citrus, larval feeding reduces yield, gir-
dles trees, and facilitates infections by plant
pathogens such as Phytophthora spp. The combi-
nation of Diaprepes and Phytophthora can cause
severe tree decline and destroy groves within a
few years of an initial infestation (Graham et al.
1996). In developing an effective integrated pest
management (IPM) program for D. abbreviatus, it
is important to maximize the effectiveness of nat-
ural enemies. Preliminary research indicates that
some of the major mortality agents of Diaprepes
eggs, larvae, and adults are predators; and that
the primary predators are ants (Whitcomb et al.
1982; Richman et al. 1983, Stuart et al. 2002, in
press, Stuart & McCoy, in press).
Ants are recognized as important predators of
pest insects in various agroecosystems and are
subject to conservation in some IPM programs
(Way & Khoo 1992, Perfecto & Castiieiras 1998,
Eubanks 2001). Indeed, the use of ants to control
citrus pests in Asia dates back to at least 304 AD,
is the earliest known example of biological con-
trol, and is still practiced today (Way & Khoo
1992). Florida has a rich and diverse ant fauna
numbering over 200 species (Deyrup et al. 2000),
and ants can be extremely abundant in Florida
citrus groves (Whitcomb et al. 1982, Richman et
al. 1983; Tryon 1986, Stuart & McCoy, in press,
Stuart et al. in press). Under the proper condi-

tions and with appropriate management, ants
could constitute a major weapon in our fight
against Diaprepes, and a conservation biological
control program focusing on appropriate ant spe-
cies might well be the key to controlling this in-
sect (Whitcomb et al. 1982, Jaffe et al. 1990,
Stuart & McCoy, in press; Stuart et al. in press).
However, at present, it is unclear which ant spe-
cies are the most effective predators of Diaprepes
on the soil surface, in the canopy, and below
ground, and what strategies might be most effec-
tive in promoting and conserving beneficial ant
species (Whitcomb et al. 1982, Richman et al.
1983, Stuart & McCoy, in press, Stuart et al. in
press). Natural variability in the abundance and
distribution of ants, combined with various possi-
ble influences of citrus management practices
could contribute to considerable variability in pre-
dation pressure by ants on Diaprepes within and
among groves across the state (e.g., see McCoy et
al. 2001). Our present research begins to address
these issues by assessing the role of ants as pred-
ators of Diaprepes neonate larvae on the soil sur-
face in citrus groves in central Florida.
The timing of Diaprepes egg hatch, neonate es-
cape from sealed leaf envelopes, and neonate drop
to the soil surface could influence the relative ex-
posure of neonates to predation (Whitcomb et al.
1982, Richman et al. 1983, Stuart et al. 2002).
Jones & Schroeder (1983) found that a consider-
able period often elapsed between egg hatch and
neonate escape from leaf envelopes, estimated av-
erage larval age at the time of neonate drop to be
about 48 h, and found that neonates dropped be-
tween 1100 and 2400 h. Ant foraging on the soil
surface during this period is reported to be low
compared to early morning hours, and the timing
of neonate drop could be an adaptation to avoid
peak foraging periods (Whitcomb et al. 1982;
Richman et al. 1983). Unfortunately, Jones &
Schroeder (1983) examined neonate drop for only
five egg masses, and did not report any informa-
tion on the light cycles used in their laboratory ex-
periments. Additional research on the activity

March 2003

Stuart et al: Predation on Diaprepes Neonates

patterns of predators, the factors that stimulate
egg hatch and neonate drop, and the conditions
that promote neonate survival in the canopy, on
the soil surface, and below ground is necessary for
a more thorough understanding of how these fac-
tors might shape D. abbreviatus life history and
survival strategies (Stuart et al. 2002).
Jaffe et al. (1990) observed that first instar Di-
aprepes larvae were preyed upon by various ant
species but that the larvae appeared to be some-
what repellent, and Jaffe et al. suggested that the
larvae might have chemical defenses. Pavis et al.
(1992) investigated this chemical repellency with
respect to the fire ant, Solenopsis geminata (F.),
on the island of Guadeloupe in the Caribbean and
identified two bicyclic sesquiterpene aldehydes
that appeared to be responsible for the effect. The
concentration of the repellent was highest in
newly-hatched neonates, decreased with larval
age, and was absent after about four days. Pavis
et al. (1992) suggested that ant predation on neo-
nate larvae during the first few hours after hatch-
ing would be reduced by about 40% because of
chemical repellency. Various coccinellid species
readily consume Diaprepes neonates under labo-
ratory conditions with no indication of repellency
(Stuart et al. 2002), and no experimental demon-
strations of neonate repellency against other ant
species have yet been reported (Stuart & McCoy,
in press).
Our objectives in the present study were to (i)
assess the daily temporal pattern of neonate drop
from Diaprepes egg masses under controlled labo-
ratory conditions, (ii) evaluate predation pressure
on neonates on the soil surface in a series of citrus
groves in central Florida, and (iii) test for neonate
age-dependent chemical repellency by comparing
predation rates on neonates of different ages un-
der field conditions in central Florida.


Neonate Drop

Egg masses laid between wax paper strips
were obtained from a Diaprepes colony that was
maintained on citrus foliage in a greenhouse at
the Citrus Research and Education Center, Lake
Alfred, FL. Egg masses were transferred to envi-
ronmental chambers (Gaffney Engineering,
Gainesville, FL) at least 4-5 days prior to neonate
drop and were held in glass funnels (6 cm dia.),
each of which was placed over a moving clock face
coated with Tanglefoot (The Tanglefoot Com-
pany, Grand Rapids, MI, 40504). The clock face
was fashioned from the lid of a plastic Petri dish
(150 x 15 mm) on which 24 segments were marked
to correspond to a 24-h clock. All one-hour seg-
ments will be referred to by the time at the begin-
ning of the hour (e.g., the 0800 h one-hour time
segment refers to the period from 0800 h to

0859 h). The clock face was secured to the top of
the mechanical clock apparatus of a hygrothermo-
graph (model H-302, Weather Measure Corpora-
tion, P.O. Box 41257, Sacamento, CA 95841). The
top of the funnel was covered with the base of a
small plastic Petri dish (60 x 15 mm) that was se-
cured to the funnel with Parafilm (American Na-
tional Can, Menasha, WI 54952). Attached to the
inside of the Petri dish was a vial of water sealed
with a cotton plug to prevent the egg mass from
drying out. The environmental chamber was set
to 24C, 70% RH, L:D = 12:12. Onset of the light
phase was at 0800 h and of the dark phase was at
2000 h. We conducted 42 replicates during the pe-
riod from 14 May through 28 October 2001, and
each replicate involved a single egg mass. This in-
cluded nine replicates that were conducted as con-
trols in which the clock face was immobile to
determine whether neonates would move be-
tween clock segments under these experimental
A two-way ANOVA(PROC GLM, SAS Institute
Inc. 1990) with seven levels for the factor "date"
and 24 levels for the factor "time" was conducted
on the percent neonate drop from each egg mass
after arcsin transformation. Means comparisons
used the LSMEANS procedure (SAS Institute Inc.
1990). A similar ANOVA was conducted for the
data pooled into four time intervals representing
the first and second halves of the light and dark


We conducted direct observations of predation
on lab-reared Diaprepes neonates (<48 h post
hatch) placed in citrus groves, and used proce-
dures similar to those of Whitcomb et al. (1982)
and Richman et al. (1983). Egg masses were ob-
tained from a Diaprepes colony that was main-
tained on citrus foliage at the Citrus Research
and Education Center, Lake Alfred, FL. Egg
masses were held at room temperature in plastic
containers that were arranged so that neonates
hatching in the upper chamber could drop
through a screen into the lower chamber. Neo-
nates were removed from the lower chambers on a
daily basis at 0800 h for use in experiments con-
ducted that day. Each replicate consisted of 20 ne-
onates placed in an open plastic dish (4 mm high
x 48 mm dia, Millipore Petrislide containers, Mil-
lipore Corporation, Bedford, MA) on the soil sur-
face under the canopy and observed for 20 min. A
thin layer of fine sand that passed through a No.
40 sieve was placed in the bottom of the dish and
effectively discouraged neonates from crawling
out (see Richman et al. 1983), whereas roughen-
ing the outside and inside vertical surfaces of the
dish with sand paper facilitated the entrance and
exit of predators. A predation event was scored
whenever a neonate was removed from the dish.

Florida Entomologist 86(1)

In most cases, predators were identified visually
by field observers based on previously identified
samples. In some cases, additional samples were
taken to confirm identifications.
We conducted a total of 199 replicates in three
groves located near the following municipalities
in central Florida: Lake Alfred (63 replicates) and
Alturas (58 replicates) in Polk County; and South-
port (78 replicates) in Osceola County. In the Lake
Alfred grove, which was located at the Citrus Re-
search and Education Center of the University of
Florida, we worked in two blocks, one of which
was a mature planting of Marsh seedless grape-
fruit on Swingle citrumelo rootstock whereas the
other was a close-set 5-year old planting of Navel
orange on Swingle citrumelo rootstock. In the
Southport grove, we worked in two mature blocks
of Hamlin orange on Swingle citrumelo rootstock.
The south block was being managed for commer-
cial production whereas the north block had been
abandoned because of severe tree decline, was no
longer irrigated, and was subject to relatively lit-
tle weed control. The Alturas grove was a uniform
planting of mature Flame grapefruit on Swingle
citrumelo rootstock and was under management
for commercial production. Observations were
conducted between 1 August and 11 September
2000, between 1200 and 1700 h. Minimum and
maximum daily temperatures during this period
at Lake Alfred, FL, ranged from 20.1 to 35.8C.
Statistical comparisons of the proportion of
replicates in which predation was observed and
the proportion of neonates preyed upon in differ-
ent groves and blocks were conducted using con-
tingency table analyses and the X2 test (PROC
FREQ, SAS Institute Inc. 1990).

Chemical Repellency

Twelve assay stations were established under
the canopy of mature grapefruit trees (Marsh
seedless grapefruit on Swingle citrumelo root-
stock) with one station per tree at the Citrus Re-
search and Education Center of the University of
Florida, Lake Alfred, FL. Diaprepes egg masses
laid between sheets of wax paper were obtained
from the USDA rearing facility, Fort Pierce, FL.
Egg masses were held as in the previous experi-
ment; and neonates were harvested on a daily ba-
sis at 0800 h, and hourly on the day of the
experiment. At each station, for each replicate, 20
neonates aged five days post hatching and 20 aged
1-2 h post hatching were placed in separate
paired open plastic dishes (as described in the
previous experiment), ca. 2-4 cm apart, on the soil
surface near the trunk, for 30 min. The number of
neonates remaining in the dishes was counted un-
der a microscope at the termination of each repli-
cate. Eight replicates were conducted, with the
position of the dishes (i.e., left versus right) being
randomized for the first replicate at each station,

and then alternated in successive replicates. Data
were collected from six stations simultaneously,
with stations 1-6 and 7-12 forming two distinct co-
horts within the experiment, which were gener-
ally run sequentially. Direct observations of
predation events, defined as a neonate being re-
moved from an assay dish by a predator, were con-
ducted during the experimental period for all
replicates by systematically observing the dishes
for 1-2 minute intervals as the test progressed.
Observations were conducted during the period
21-26 July 2001, between 0850 and 1550 h.
Weather data were recorded during the first 15
mins of each replicate by FAWN (Florida Auto-
mated Weather Network, University of Florida,
Gainesville, FL) in an open field at the Citrus Re-
search and Education Center, Lake Alfred, FL.
These data include air temperature (measured
61.0 cm above the soil surface), soil temperature
(measured 7.6 cm below the soil surface), relative
humidity, and solar radiation.
A three-way ANOVA (PROC GLM, SAS Insti-
tute Inc. 1990) with two levels for the factor "age"
(old and young), two levels for the factor "position"
(left and right), and twelve levels for the factor
"station" (1-12) was conducted on the percent pre-
dation in each dish after arcsin transformation.
Means comparisons used the LSMEANS proce-
dure (SAS Institute Inc. 1990). Percent predation
on old versus young larvae by various ant species
was compared using contingency table analysis
and the x2 test (PROC FREQ, SAS Institute Inc.


Neonate Drop

The 42 egg masses contained from 21 to 127
eggs (total = 2687, mean = 63.98, SE = 3.983). A
total of 491 eggs failed to hatch (18.3%), and an
additional 60 eggs hatched but the neonates
failed to exit the egg masses (2.2%). In two cases,
the entire egg mass failed to hatch (4.8%). A total
of 259 neonates dropped in the nine control repli-
cates. Of these, 11 neonates moved across one
clock-segment boundary (4.2%), and none moved
across more than one boundary. In the 33 experi-
mental replicates, less than 20 neonates dropped
in four replicates, and only the remaining 29 rep-
licates were considered in assessing the temporal
pattern of neonate drop. In these 29 replicates, 23
to 125 neonates dropped per egg mass (total =
1865, mean = 64.31, SE = 5.410).
Neonates dropped during every hour of the
light and dark phases, but there were significant
differences in the percentage of neonates drop-
ping among one-hour and six-hour time intervals
(Fig. 1A, B). For the analysis involving 24 one-
hour time intervals (Fig. 1A), the dates on which
replicates were conducted had no influence on the

March 2003

Stuart et al: Predation on Diaprepes Neonates

. 6

o 0
a. 50


a C b

e f


P= 0.001

d d

ba bbbbb
a a a a a a
212223 241 2 3 4 5 6 7
P = 0.001

20 c
C | c
10 -- --- ------- ..-J--I

0800 1300 1400 1900 2000 0100 0200-0700
Time (24-h clock)
Fig. 1. Comparison of the percentage of neonates
(mean + SE) that dropped from egg masses at various
times of the day when the onset of the light phase was
at 0800 h and the dark phase was at 2000 h with the
data partitioned either hourly (A) or into six-hour time
intervals (B). Statistical analysis was by way of ANOVA
and LSMEANS. Common letters above bars indicate no
significant difference at the P = 0.05 level.

results (ANOVA: F = 1.01, df= 6, 528, P = 0.4148)
but time was a significant factor (ANOVA: F=
12.50, df= 23, 528, P = 0.0001) and the interaction
between date and time was not significant
(ANOVA: F = 1.22, df= 138, 528, P = 0.0676). Sim-
ilarly, for the analysis involving four six-hour time
intervals (Fig. 1B), the dates on which replicates
were conducted had no influence on the results
(ANOVA: F = 0.13, df= 6, 88, P = 0.9928) but time
was a significant factor (ANOVA: F= 22.32, df= 3,
88, P = 0.0001) and the interaction between date
and time was not significant (ANOVA: F = 1.47, df
= 18, 88, P = 0.1192). On average, 17.9% of neo-
nates dropped from 0800-1300 h, 52.4% from
1400-1900 h, 21.7% from 2000-0100 h, and 8.0%
from 0200-0700 h.
The pattern of neonate drop for individual egg
masses over the four six-hour time intervals de-
fined above shows that the timing of neonate drop
was not highly synchronized within or among egg
masses (Fig. 2). For individual egg masses, neo-
nates dropped over a period of 5 to 23 h (mean =
11.97, SE = 0.866). Of the 29 egg masses referred
to above, 28 had neonates drop during both the
light and dark phases (96.6%), 12 during all four
of the six-hour time intervals (41.4%), 24 during
at least three intervals (82.8%), 28 during at least

two intervals (96.6%), and one during only one in-
terval (3.4%). All 29 egg masses had some neo-
nates drop during the peak 6-h time interval from
1400-1900 h, 23 had more than 25% drop during
this period (79.3%), 17 had more than 50% drop
during this period (58.6%), and 8 had more than
75% drop during this period (27.6%). Two egg
masses had more than 50% of their neonates drop
from 0800-1300h (6.9%), three had more than
50% drop from 2000-0100 h (10.3%), and one had
more than 50% drop from 0200-0700 h (3.4%).

We observed predation in 104 of 199 replicates
(52.3%) with a total of 475 of the 3980 neonates
being removed from assay dishes by predators
(11.9%). All but one of the observed predation
events were by ants (Hymenoptera: Formicidae).
The single exception was by a nymph of the big-
eyed bug, Geocoris floridanus Blatchley (Hemi-
ptera: Lygaeidae), that attacked a neonate in the
Lake Alfred grove. There was no significant differ-
ence in the proportion of replicates in which pre-
dation was observed at the three sites (Fig. 3A),
but a greater proportion of neonates was preyed
upon at the Alturas grove than at the other sites
(Fig. 3B). Significant differences in the level of
predation occurred within different areas of the
groves. At the Lake Alfred grove, we observed no
significant difference in the proportion of repli-
cates that resulted in predation events in the
grapefruit block versus the orange block (Fig. 3C),
but a greater proportion of neonates was preyed
upon in the orange block (Fig. 3D). At the South-
port grove, we observed a highly significant differ-
ence in the proportion of replicates that resulted
in predation events in the abandoned north block
versus the cultivated south block (Fig. 3E), and a
greater proportion of neonates was also preyed
upon in the north block (Fig. 3F).
Eight ant species preyed on neonates in this
experiment. For the data pooled from all sites, the
red imported fire ant Solenopsis invicta Buren
and Pheidole moerens Wheeler were the most ac-
tive predators, and accounted for 29.5% and
27.8% of the predation events, respectively. Other
predatory ant species in decreasing order of per-
cent predation for the pooled data included Dory-
myrmex reginicula (Trager) (9.7%),
Brachymyrmex obscurior Forel (8.8%), Dory-
myrmex bureni (Trager) (8.6%), Cardiocondyla
emeryi Forel (8 0 i Paratrechina bourbonica
Forel (4.8%), and Pheidole morrisi Forel (2.5%).
Predation pressure by different ant species was
variable within and among groves and probably
reflects differences in the abundance and distri-
bution of those species (Fig 4A, B). It is notewor-
thy that the Alturas grove was the site with the
highest predation pressure on neonates (Fig. 3B)
and was also the site in which S. invicta was the
most active predator (Fig. 4A, B).

Florida Entomologist 86(1)

60 68 65 80 90 52 98 23 50 30 68 70 33 12444 40 64 91 67 43 31 30125114 53 44 52 45 111

a. -
2 80


M 60-

S 50 -.

40 -

40 26 28 32 3019 27 8 20 34 7 3714 38 2 13 33 22 1 31 5 36 392316 35 4 4125
Replicate Number
S 0800 1300 h 1400 1900 h ZE 2000 0100h M 0200 0700 h
Fig. 2. Comparison of the percentage of neonates that dropped from egg masses at various times of the day for
data partitioned into six-hour time intervals and ranked from left to right in descending order of the percentage of
neonates that dropped during the peak time interval, 1400-1900 h (see Fig. 1). The replicate number is indicated
below each bar, and the number of neonates dropping from the egg mass in that replicate is indicated above the bar.

Casual observations of the behavior of the ants
in this study indicate that workers of different
species often respond quite differently to Di-
aprepes neonates. Workers of both Dorymyrmex
species are relatively large and fast moving, and
often failed to respond to neonates as they quickly
passed through assay dishes. S. invicta workers
are also relatively large but move more slowly,
and responded to neonates more frequently. We
also noted that S. invicta often appeared to sting
neonates before carrying them out of the dish,
and that sometimes S. invicta workers appeared
to have difficulty removing their sting from the
body of the neonate and would wander about the
dish with the neonate attached to the tip of their
abdomen by the sting. Workers of the smaller spe-
cies (i.e., P. moerens, B. obscurior, and C. emeryi)
appeared to detect and respond to neonates more
readily than the larger species. However, P. moer-
ens workers typically seized and carried off neo-
nates almost immediately upon contact whereas
B. obscurior was more hesitant, often picked up
and dropped neonates repeatedly, and sometimes
abandoned the assay dish without a neonate. This
handling difficulty is reminiscent of some of the
behavior observed by previous researchers and

might be indicative of repellency (Jaffe et al.
1990, Pavis et al. 1992). C. emeryi was intermedi-
ate to the other two small species in terms of its
handling efficiency. We did not observe mass re-
cruitment of nestmates to assay dishes by any ant
species in this experiment. Rather, predation ap-
peared to be conducted by individually foraging
workers that discovered and carried off a neonate,
and often returned repeatedly to an assay dish to
prey on additional neonates.

Chemical Repellency

A total of 2620 of the 3840 larvae (68.2%) were
preyed upon in this experiment but the predation
rate on 5-day old versus 1-2 h old Diaprepes neo-
nates showed no significant difference at 68.8%
and 67.7%, respectively (ANOVA: F = 0.44, df= 1,
144, P = 0.5067). Both station and position were
significant factors (ANOVA: station, F = 19.27, df
= 11, 144, P = 0.0001; position, F = 5.56, df = 1,
144, P = 0.0197), and there was a highly signifi-
cant interaction between station and position
(ANOVA: F = 6.11, df = 11, 144, P = 0.0001). The
position of the assay dishes, whether left or right,
had a significant impact on the percent predation

March 2003

Stuart et al: Predation on Diaprepes Neonates

P = 0.001

C P= 0.215 D P=0.001
0] 36
0 b
0 a 540
0 I I I I
Grapefruit Orange Grapefruit Orange
E a P= 0.001 F P=0.001


44 680 b
i 1 880

S North South North South

Fig. 3. Comparison of the percentage of replicates in
which predation was observed (A, C, E) and the percent-
age of neonates preyed upon (B, D, F) in the three groves
(A, B), and for different areas within two groves, Lake
Alfred (C, D) and Southport (E, F). The numbers above
the bars indicate the total number of replicates (A, C, E)
or neonates (B, D, F). P-values indicate the results ofX2
tests on 2 x 2 or 2 x 3 contingency tables. Common let-
ters above bars indicate no significant difference at the
P = 0.05 level. No letters appear above groups of bars
where the overall contingency table analysis was not
significant at the P = 0.05 level.

at five of the 12 stations, and predation was more
intense at some stations than at others (Fig. 5A).
At one station (#11), predation was 100% in both
dishes in every replicate.
A total of 368 of the 2620 predation events
(14.0%) were observed directly. All observed pre-
dation was by six ant species, with P. moerens and
S. invicta preying on 62.5% and 25.3% of the neo-
nates, respectively. Other ant species in decreas-
ing order of percent predation included C. emeryi
(10.1%), B. obscurior (1.4%), Cardiocondyla
wroughtonii (Forel) (0.5%), and D. bureni (0.3%).
C. wroughtonii was not observed as a predator in
the previous experiment, and is the ninth preda-
tory ant species reported here. The two predation
events by C. wroughtonii occurred at the same as-
say station on successive days between 1520 and
1600 h. A comparison of the number of predation
events by P. moerens, S. invicta, or the other ant
species pooled on old versus young larvae re-
vealed no significant differences (2 x 3 contin-

100 Replicates
SA P = 0.336
60 63 5
-- 78

Lake Southport Alturas

a a 1160
1260 1560
I- I-I I
Lake Southport Alturas


. 8
-" 6

o 5
0 cJ A I di dId d d O ff
'a Lake Alfred Southport Alturas
)4 50 c, N=12
S5 B M Lake Alfred, N= 120
40 b M Southport. N = 142
S40I Alturas. N =213
30 b
25 a
25 a b
20 ; a a
15 a- b b a
10 a c
b b c aa jbb

0 0

< 4* 0' R

Fig4.4. Comparison of the percent predation by differ-
ent ant species within and among the three groves. Sta-
tistical analysis was by way contingency tables and the
X2 test. Common letters above bars indicate no signifi-
cant difference at the P = 0.05 level either among species
within sites (A) or within species among sites (B).

agency table, x2 = 4.170, df = 2, P = 0.124; Fig. 5B).
Thus, we found no evidence for age-dependent re-
pellency of neonates, either overall or for particu-
lar ant species.
There was no correlation between time of day
and percent predation for data pooled among sta-
tions and assay dish positions for stations 1-6
(Fig. 6A) but there was a significant increase in
predation rate as the day progressed for stations
7-12 (Fig. 6B). There were no significant correla-
tions between predation rate and air tempera-
ture, relative humidity, or solar radiation for
either experimental cohort (Table 1). Soil temper-
ature was correlated with predation rate for sta-
tions 7-12 but not for stations 1-6 (Table 1).


This study indicates that ants are important
predators of Diaprepes neonates on the soil sur-
face in the citrus groves of central Florida. Our
laboratory data indicate that neonate larvae drop
from the canopy during all hours of the day and
night but that peak drop occurs during the after-
noon. Our field tests demonstrate that at least
nine ant species prey on neonates on the soil sur-

Florida Entomologist 86(1)




Left I Right
100 A ns ns ** ns ns
T- ns ...
80 ns

60 T


20 ns

1 2 3 4 5 6 7 8 9 10 11 12
An -

7 B

I I Young

P. moerens S. invicta

P= 0.124


Fig. 5. Comparison of percent predation (mean + SE)
on neonate larvae by ants for different assay-dish posi-
tions (left or right) at the 12 stations (A), and on old ver-
sus young neonate larvae by different ant species (B).
Statistical analysis was by ANOVA and LSMEANS (A),
or by contingency tables and the X2 test (B): ns, P > 0.05;
*, P < 0.05; ** P < 0.01; ***, P < 0.001.

face, and are active during the peak drop period,
but that predation pressure overall and by partic-
ular species can be highly variable within and
among groves. Ants preying on Diaprepes neo-
nates in this study included Brachymyrmex ob-
scurior, Cardiocondyla emeryi, C. wroughtonii,
Dorymyrmex bureni, D. reginicula, Paratrechina
bourbonica, Pheidole moerens, Ph. morrisi, and
Solenopsis invicta. A single predation event on a
neonate by a nymph of the big-eyed bug, Geocoris
floridanus, was also observed. We found no field
evidence for differential predation on neonates of
different ages by ants in this community and,
hence, no evidence for age-dependent chemical re-
pellency of neonates (see Jaffe et al. 1990, Pavis et
al. 1992). However, behavioral observations indi-
cate that ant species differ widely in their abili-
ties to detect and handle this prey item and,
consequently, the relative importance of particu-
lar species as predators of neonates is likely to de-
part markedly from their relative abundance in
this agroecosystem. Our results reinforce the
view that ants are among the primary mortality
agents of Diaprepes, and have potential as the ba-
sis for a conservation biological control program
(Whitcomb et al. 1982, Jaffe et al. 1990, Stuart &
McCoy, in press, Stuart et al. in press).



0800 1000 1200 1400 1600
Time (24-h clock)

Fig. 6. Relationship between percent predation on
neonate larvae and time of day for stations 1-6 (A) and
7-12 (B) of the chemical repellency experiment. The
equation for the regression line, the Pearson correlation
coefficient (r), and the P-value are given.

Previous research on ant predation on Di-
aprepes neonates in a central Florida citrus grove
in Forest City, Seminole County, identified
Pheidole dentata Mayr, P. floridana Emery, and
Tetramorium simillimum Roger as the primary
species involved (Whitcomb et al. 1982, Richman
et al. 1983, Tryon 1986) but none of these species
were detected in the present study. Moreover,
whereas S. invicta and P. moerens were relatively
minor predators in the previous studies, they
were major predators in the present study. It is
unclear whether this difference represents mere
variability among sites or whether the ant fauna
in central Florida citrus groves has undergone a
dramatic change during this period. Florida has
over 200 ant species, more than 50 of which are
thought to be introduced, exotic species (Deyrup
et al. 2000). Both S. invicta and P. moerens are ex-
otic, and were classified by Deyrup et al. (2000) as
"possible ecological villains" since they occur in
both disturbed and undisturbed habitats, appear
to dominate their trophic roles, and might dis-
place native competitors. Indeed, both S. invicta
and P. moerens appear to have spread "explo-
sively" in Florida since they were first reported in
1950 and 1975, respectively (Deyrup et al. 2000).
However, whereas S. invicta is considered a major
invasive pest in a broad range of habitats and is
often subject to intensive control efforts (see Vin-
son 1997), P. moerens has remained obscure (Dey-
rup et al. 2000). Notably, of the predatory ants

A y = 0.0028x + 65.276
P= 0.8072


y = 0.0224x + 47.262
r = 0.8381
P= 0.0094

March 2003

Stuart et al: Predation on Diaprepes Neonates


with % Predation

Station No. Variable N Mean SE Range R P

1-6 Predation (%) 8 62.15 2.326 52.5-71.3 -
Air Temperature (C) 8 28.31 0.873 26.0-33.2 0.2660 0.5244
Soil Temperature (C) 8 28.61 0.462 26.9-31.1 -0.4654 0.2452
Relative Humidity (%) 8 72.78 3.785 51.5-84.9 -0.3654 0.3734
Solar Radiation (W/m2) 8 416.14 87.233 61.0-722.0 0.2637 0.5281
7-12 Predation (%) 8 74.33 2.557 61.7-82.9 -
Air Temperature (C) 8 28.67 1.138 24.3-33.7 0.2842 0.4951
Soil Temperature (C) 8 28.76 0.370 27.2-30.6 0.7543 0.0306
Relative Humidity (%) 8 72.04 4.905 52.0-90.7 -0.4979 0.2093
Solar Radiation (W/m2) 8 411.63 89.858 46.0-853.0 0.0287 0.9463

recorded in the present study, only the two Dory-
myrmex species, P morrisi, and possibly B. obscu-
rior, are considered native (Deyrup et al. 2000);
and very little is known concerning the ecological
dynamics of the highly synthetic ant community
we now find in Florida citrus.
The results of the present study show how
variable ant predation pressure on Diaprepes ne-
onates can be in Florida citrus groves. Our exper-
iments detected variation in predation rates
among groves, among blocks within groves,
among assay stations under a series of grapefruit
trees, and between paired assay dishes only a few
centimeters apart. This spatial variability proba-
bly reflects heterogeneity in the distribution and
abundance of ant nests of various species. Unfor-
tunately, there is little information on factors that
might influence the distribution, growth, and ac-
tivity levels of ant colonies of most species in this
particular agroecosystem, but such information is
necessary for the selective manipulation of spe-
cies within a conservation biological control pro-
gram. One exception is the red imported fire ant,
Solenopsis invicta, since many aspects of its biol-
ogy are well known (Vinson 1997). However, S. in-
victa is sometimes considered a citrus pest since
it feeds on foliage and bark, can girdle young
trees, tends aphids and scales, and disrupts har-
vesting by stinging grove workers (Banks et al.
1991, McCoy 1999, Michaud et al. 2002). Chemi-
cal suppression of S. invicta tends to increase ant
diversity (McCoy et al. 2001) but, since S. invicta
is such a voracious predator, it is unclear whether
manipulating the ant community in this manner
has a positive or a negative impact on the biolog-
ical control of insect pests in general or of Di-
aprepes in particular. Whatever the case, S.
invicta has become the dominant ant in many
Florida citrus groves (Banks et al. 1991) and fur-
ther studies of the positive and negative impacts
of this and other ant species in this agroecosys-

tem seem warranted. In particular, we know vir-
tually nothing about the biology of P. moerens
(Deyrup et al. 2000) and its appearance as a ma-
jor predator of Diaprepes neonates in the present
study justifies further research.
Our results on the temporal pattern of neonate
drop complement and extend those obtained pre-
viously. On the basis of five egg masses, Jones &
Schroeder (1983) concluded that neonate drop oc-
curred from 1100 to 2400 h but provided no fur-
ther information on the temporal pattern
observed, variability within or among egg masses,
or on the light cycle under which these data were
obtained. With a larger data set of 29 egg masses
and controlled laboratory conditions (24C, 70%
RH, L:D = 12:12), we found that a peak in neonate
drop occurred during the second half of the light
phase but that neonate drop could occur during
all hours of the light and dark phases, and that
there was considerable variability in the timing of
neonate drop within and among egg masses. In
central Florida, neonate drop is closely associated
with adult abundance and has been recorded
from mid June to mid December (McCoy et al. in
press, Nigg et al. in press). Further research is
necessary to explore patterns of neonate drop un-
der the range of light cycles and environmental
conditions that occur during this period. At
present, it is unclear what proximate or ultimate
factors influence neonate drop, and to what ex-
tent temporal patterns are under genetic, devel-
opmental, or environmental control. Such
patterns might constitute adaptive responses to
the activity patterns of predators (Whitcomb et
al. 1982, Richman et al. 1983, Stuart et al. 2002),
or be influenced by environmental conditions
such as rainfall, which could have an important
impact on the ability of neonates to penetrate soil
(Jones & Schroeder 1983).
Daily cycles in the foraging activity of ants in
Florida citrus groves could be an important factor

Florida Entomologist 86(1)

regulating predation pressure on Diaprepes neo-
nates. When Whitcomb et al. (1982) examined ant
predation on neonates in a central Florida citrus
grove using similar procedures to ours, they
found no predation at 1200 h and 1500 h but 62%,
44% and 49% predation at 0700 h, 1800 h, and
2400 h, respectively. Richman et al. (1983) found
predation levels of only 9.6% from 1200 to 1530 h
in the same grove as the previous researchers
but, since they used 50 neonates per dish rather
than 20, their result might better be adjusted to
24.0% for a reasonable comparison. In the present
experiments, we found a predation rate of 11.9%
from 1200 to 1700 h when using a 20 min expo-
sure period similar to those cited above, but a rate
of 68.2% from 0850 to 1550 h when using a 30 min
exposure period. Moreover, in the latter case,
there was no evidence of a decline in predation
rate during the day as suggested by the previous
researchers. Indeed, in one cohort, there was a
significant increase over the course of the day.
Current data suggest that ant foraging activ-
ity could be highly variable temporally and spa-
tially in Florida citrus groves, and might depend
more on the species involved, their relative abun-
dance, and ambient environmental conditions
than on time of dayper se. According to H11dobler
& Wilson (1990), every ant species can be ex-
pected to have a distinctive foraging schedule. In
some species, circadian rhythms have been dem-
onstrated but can apparently be over-ridden or
phase shifted by colony hunger, patterns of food
availability, or other factors. To some extent, dif-
ferent foraging schedules among similar sympat-
ric species might ultimately reflect coevolution
and the temporal partitioning of resources, but
could be based proximately on different humidity
and temperature preferences or tolerances. How-
ever, in a detailed study of the foraging patterns
of Solenopsis invicta, Porter & Tschinkel (1987)
found that soil temperature, season, and rainfall
explained most of the variation in foraging activ-
ity as indicated by food discoveries and recruit-
ment to baits. Factors like time of day, even the
difference between day and night, were not re-
lated to foraging activity. S. invicta foraged when
soil temperatures at a depth of 2 cm ranged from
15 to 43C, and exhibited maximum foraging
rates between 22 and 36C. Foraging was unusu-
ally low in late fall and was reduced during peri-
ods of rainfall. In the present research, we found
that predation rates were related to time of day
and soil temperature at some assay stations but
not at others. Unfortunately, our weather records
were based on measurements made in a nearby
open field rather than under the citrus canopy
where the experiments took place, and more de-
tailed information regarding environmental con-
ditions in this particular microhabitat might have
been revealing. Further research is necessary to
elucidate the foraging patterns of other ant spe-

cies in this community, the factors that regulate
them, and how this might affect Diaprepes sur-
vival strategies.
This study indicates that there is no differen-
tial predation on 5-day versus 1-2 h old Diaprepes
neonates by ants in a central Florida citrus grove.
Thus, there is no evidence that a quantitative de-
crease in the chemical ant repellents produced by
neonates over this time period as reported by
Pavis et al. (1992) has any influence on the inten-
sity of predation by this ant community. Our re-
sults are not necessarily in conflict with those of
Pavis et al. (1992) since the repellents might be
equally effective when present in small or large
amounts, or might be totally ineffective against
the ant species in this study, which did not include
the species in their study, Solenopsis geminata.
Furthermore, since Pavis et al. (1992) conducted
their research on the island of Guadeloupe in the
Caribbean, and since the Diaprepes population in
Florida was introduced, perhaps as a few small
founder populations (Bas et al. 2000), it is possi-
ble that the weevils in the two studies differ ge-
netically and that the dynamics of the purported
chemical ant-repellent system is different as well.
Coccinellids are reported to be deterred from at-
tacking alfalfa weevil larvae by their defensive
wriggling (Kalaskar & Evans 2001), and it is un-
clear whether wriggling contributes to defense for
Diaprepes larvae. Further research is necessary
to explore these possibilities.
The predation assay used in the present study
might underestimate predation pressure on Di-
aprepes neonates. This assay would miss any "sit
and wait" predators (e.g., ant lions, various spi-
ders) present under trees, and presents a novel
substrate to mobile predators that might deter
prey searching. It often appeared in our experi-
ments that ants more readily walked around assay
dishes than through them. However, these assay
dishes might also facilitate predation by depriving
neonates of cracks, crevices, and other complex el-
ements of the leaf litter environment that might
shelter them from predators or facilitate soil pene-
tration. Jones & Schroeder (1983) found that the
presence of grass stems and leaves placed verti-
cally in soil did not promote soil penetration, and
that neonates failed to penetrate dry soil. They
also found that half of the groups of 20 neonates in
their experiments that were aged 9 and 72 h re-
quired 80 and 105 mins, respectively, to penetrate
below the surface of moist soil, and that all of the
larvae in these groups disappeared below the sur-
face within 180 mins. These results indicate that
neonates might often remain on the soil surface for
relatively long periods and that exposure times of
20 or 30 mins as used in our experiments are not
excessive. Our experiments presented neonates to
predators at an initial density of 20 neonates per
assay dish, a density that would appear to consti-
tute a relatively diffuse food resource and that

March 2003

Stuart et al: Predation on Diaprepes Neonates

seems justified by the temporal pattern of neonate
drop from individual egg masses observed in the
present study. At this density, we found that ants
discovering a neonate and carrying it off to their
nest did not engage in the mass recruitment of
nestmates but would often return to the assay dish
repeatedly to prey on additional neonates, a forag-
ing response common to many ant species and
known as Ortstreue (H1lldobler & Wilson 1990).
Neonates experimentally presented to predators
at higher densities are reported to induce mass re-
cruitment by some ant species (Whitcomb et al.
1982), and predation rates observed in such exper-
iments might be considered artifacts of an unreal-
istically high neonate density. Consequently, we
suggest low densities of the kind used here as a
more realistic mode of neonate presentation for fu-
ture experiments.
In general, the extent to which ants prey on
various life stages ofDiaprepes and are capable of
controlling this insect will likely depend on the
abundance and diversity of the ant species
present. In turn, the structure of the ant commu-
nity in particular citrus groves probably depends
on an array of factors including local environmen-
tal conditions, grove management practices, and
the dynamics associated with interactions among
the various native and exotic species that become
established within groves. Given the possible
complexity of these factors and their interactions,
it is premature to speculate upon what factors
might have contributed to the variability in pre-
dation rates among groves and blocks within
groves that we observed in the present study.
Nonetheless, since ants are often extremely abun-
dant in Florida citrus groves, and can be such im-
portant predators of a broad range of insect pests,
considerable benefits could be derived from fur-
ther studies of this ant community, the positive
and negative impacts of different species, and how
they might best be managed for pest control
within the framework of a comprehensive IPM
program. It might seem unusual to consider incor-
porating invasive exotic ant species into such a
program but many of these species are now well
established components of the ecological land-
scape in Florida and exploiting their beneficial as-
pects, especially in agricultural habitats, might be
an extremely practical and cost-effective way of
using these resources.


We thank Angel Hoyte for rearing Diaprepes neonates,
Karin Crosby for providing Diaprepes egg masses, Mark
Deyrup for assistance with ant identification, Julieta
Brambila for identifying the Hemipteran, and J. P.
Michaud and H. N. Nigg for comments on the manuscript.
This research was supported by the Florida Agricultural
Experiment Station and a grant from the Florida Citrus
Production Research Advisory Council, and approved for
publication as Journal Series No. R-09033.


Damage to young citrus trees by the red imported
fire ant (Hymenoptera: Formicidae). J. Econ. Ento-
mol. 84: 241-246.
Genetic relationships among Florida Diaprepes ab-
breuiatus (Coleoptera: Curculionidae) populations.
Ann. Entomol. Soc. Am. 93: 459-467.
DEYRUP, M., L. DAVIS, AND S. COVER 2000. Exotic ants
in Florida. Trans. Am. Entomol. Assoc. 126: 293-326.
EUBANKS, M. D. 2001. Estimates of the direct and indi-
rect effects of red imported fire ants on biological
control in field crops. Biol. Cont. 21: 35-43.
Insect-plant pathogen interactions: preliminary
studies of Diaprepes root weevil injury and Phytoph-
thora infections. Proc. Florida State Hort. Soc. 109:
HOLLDOBLER, B., AND E. O. WILSON. 1990. The Ants.
Cambridge, MA: Belknap/Harvard.
Qualitative evaluation of ants as biological control
agents with special reference to predators on Di-
aprepes spp. (Coleoptera: Curculionidae) on citrus
groves in Martinique and Guadeloupe. Rencontres
Caraibes en Lutte Biologique, Guadeloupe. pp. 405-
JONES, I. F., AND W. J. SCHROEDER 1983. Study of first-
instar Diaprepes abbreviatus (Coleoptera: Curcu-
lionidae) activity for control purposes. J. Econ. Ento-
mol 76: 567-569.
KALASKAR, A., AND E. W. EVANS. 2001. Larval responses
ofAphidophagous lady beetles (Coleoptera: Coccinel-
lidae) to weevil larvae versus aphids as prey. Ann.
Entomol. Soc. Am. 94: 76-81.
McCoY, C. W. 1999. Arthropod pests of citrus roots, In:
L. W. Timmer and L. W. Duncan (eds.), Citrus Health
Management, APS Press, St. Paul, MN, pp. 149-156.
A. HOYTE. 2001. Soil surface application of chemicals
for the control of neonate Diaprepes abbreviatus (Co-
leoptera: Curculionidae) and their effect on ant pred-
ators. Florida Entomol. 84: 327-335.
J. FOJTIK, AND A. HOYTE. in press. Seasonal ecologi-
cal studies on Diaprepes in Florida citrus. Florida
CHILDERS, AND C. W. MCCOY. 2002. Other insect
pests, In L. W. Timmer (ed.), 2002 Florida Citrus
Pest Management Guide, Cooperative Extension
Service, University of Florida, Institute of Food and
Agricultural Sciences, SP-43, Gainesville, FL, pp.
COY. in press. Larval abundance in a Diaprepes ab-
breviatus (L.) infested Florida citrus grove. J. Econ.
K. JAFFE, AND C. DESCOINS. 1992. Defensive secre-
tion of first-instar larvae of rootstalk borer weevil,
Diaprepes abbreviatus L. (Coleoptera: Curculion-
idae), to the fire ant Solenopsis geminata (F.) (Hy-
menoptera: Formicidae). J. Chem. Ecol. 18: 2055-

PERFECT, I., AND A. CASTINNEIRS. 1998. Deployment of
the predaceous ants and their conservation in agroeco-
systems, In P. Barbosa (ed.) Conservation Biological
Control. Academic Press, San Diego, CA, pp. 269-289.
PORTER, S. D., AND W. R. TSCHINKEL. 1987. Foraging in
Solenopsis invicta (Hymenoptera: Formicidae): ef-
fects of weather and season. Environ. Entomol. 16:
QUINTELA, E. D., J. FAN, AND C. W. McCOY. 1998. De-
velopment of Diaprepes abbreviatus (Coleoptera:
Curculionidae) on artificial and citrus root sub-
strates. J. Econ. Entomol. 91: 1173-1179.
1983. Predation on neonate larvae of Diaprepes ab-
breviatus (Coleoptera: Curculionidae) in Florida and
Puerto Rico citrus groves. Florida Entomol. 66(2):
SAS INSTITUTE INC. 1990. SAS/STAT Guide for Personal
Computers. Cary, North Carolina: SAS Institute Inc.
STUART, R. J., AND C. W. McCOY. in press. Ants as bio-
logical control agents of the citrus root weevil, Di-
aprepes abbreviatus (Coleoptera: Curculionidae):
does the age of neonate weevil larvae influence pre-
dation? Proceedings of the First International Sym-
posium on the Biological Control of Arthropods.

March 2003

press. Ants as predators of the root weevil Diaprepes
abbreviatus (Coleoptera: Curculionidae) in Florida
citrus. Proceedings of the International Society of Ci-
COY. 2002. Ladybeetles as potential predators of the
root weevil, Diaprepes abbreviatus, in Florida citrus.
Florida Entomol. 85(3): 409-416.
TRYON, E. H. JR 1986. The striped earwig, and ant pred-
ators of sugarcane rootstock borer, in Florida citrus.
Florida Entomol. 69: 336-343.
VINSON, S. B. 1997. Invasion of the red imported fire ant
(Hymenoptera: Formicidae): spread, biology, and im-
pact. Am. Entomol. 43: 23-39.
Way, M. J., and K. C. Khoo. 1992. Role of ants in pest
management. Ann. Rev. Entomol. 37: 479-503.
Predators of Diaprepes abbreviatus (Coleoptera:
Curculionidae) larvae. Florida Entomol. 65(1): 150-
WOODRUFF, R. E. 1985. Citrus weevils in Florida and
the West Indies: preliminary report on systematics,
biology, and distribution (Coleoptera: Curculion-
idae). Florida Entomol. 68: 370-379.

Florida Entomologist 86(1)

Krecek & Scheffrahn: New Neotermes from Cuba


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

Neotermes phragmosus n. sp. is described from the imago and soldier castes. The imago
head capsule ofN. phragmosus has a distinctly phragmotic and concave frons. Plesiomorphic
characters ofN. phragmosus unique among the Kalotermitidae include partial separation of
the otherwise fused first and second marginal teeth of the left imago/worker mandible, long
subcosta and radius, and increased number of antennal articles in both images and soldiers.
This species is confined to the xeric coastal habitats of southeastern Cuba.

Key Words: new species, taxonomy, West Indies, Greater Antilles, Caribbean
El Neotermes phragmosus n. sp. es descrito de la casta imago y la casta soldado. La capsula
de la cabeza del imago N. phragmosus tiene el frente distintivamente fragmotico y c6ncavo.
Las caracteristicas plesiomorficas del N. phragmosus son unicas entire los Kalotermitidae in-
cluyen la separaci6n parcial de los primeros y segundos dientes marginales de la mandibula
izquierda del imago/trabajador, que en otros casos se encuentra fundidos; un subcosta y un
radio largados; y un mayor numero de articulos en las antenas en los images y los soldados.
Esta especie esta restringida a la zona arida costera del sureste de Cuba.

A species of Neotermes, collected in extreme
southeastern Cuba, was originally listed as Neoter-
mes sp. nr. mona (Banks) (Scheffrahn et al. 1994).
A subsequent redescription of N. mona (Krecek et
al. 2000) revealed that the Cuban Neotermes was a
new species that is described herein from the
winged imago and small and large soldier castes.
Neotermes phragmosus n. sp. is the fourth Neoter-
mes species recorded from Cuba and the sixth from
the Greater Antilles. Neotermes phragmosus and
N. cubanus (Snyder) are endemic solely to Cuba
(Snyder 1956, data herein). Of the two additional
Cuban species, N. castaneus (Burmeister) is also
recorded from the Bahamas, Cayman Islands,
Florida, Hispaniola, Jamaica, and Turks and Ca-
icos Islands (Scheffrahn et al. unpublished), while
N. jouteli (Banks) ranges into the Bahamas, Cuba,
Florida, and Mexico (Scheffrahn et al. 2000). The
remaining Greater Antillean species include N.
platyfrons Krecek and Scheffrahn (2001) from His-
paniola, and N. mona from the Bahamas, Hispani-
ola, Puerto Rico, Turks and Caicos, and Virgin
Islands (Krecek et al. 2000).


The description ofN. phragmosus is based on 87
colony samples from the authors' collection taken
from 23 localities in Guantanamo Province, Cuba,
as part of a survey of termites of Cuba and the
West Indies (Fig. 4). Collection localities were
mapped using ArcView GIS version 3.0a software
and relevant map data from Digital Map of the
World version 1.0 (Environmental Systems Re-

search Institute, Inc. Redlands, CA). Morphomet-
ric data from specimens preserved in 85% ethanol
were obtained using a stereomicroscope fitted with
an ocular micrometer. Scanning electron micro-
graphs were scanned at 300 dpi, the specimen out-
line captured with photograph-enhancing software
(Adobe Photoshop Elements, Adobe Systems Inc.,
San Jose, CA), the background converted to black,
and the scale bar digitally redrawn. The imago
head capsule photomicrograph was obtained using
a digitized three-dimensional imaging system
(Auto-Montage, Syncroscopy Inc. Frederick, MD)
and further enhanced as mentioned above.
The holotype alate and paratype large and
small soldier will be deposited at the American Mu-
seum of Natural History, New York. The additional
alate and soldier paratypes will be submitted to the
National Museum of Natural History (Smithsonian
Institution), Washington, D.C., and to the Florida
State Collection ofArthropods, Florida Department
of Agriculture and Consumer Services, Division of
Plant Industry, Gainesville, Florida. The remaining
paratypes will be held in the authors' collection at
the University of Florida Research and Education
Center, Fort Lauderdale, Florida.

Neotermes sp. nr. mona (Banks); Scheffrahn et al.
1994:217 (Cuba).

Imago (Figs. 2 and 3, Table 1).
In dorsal view, head capsule ferruginous or-
ange, except for slightly darker ferruginous ante-

Florida Entomologist 86(1)

Fig. 1. Scanning electron micrograph of anterior of the large soldier head (dorsal view) ofNeotermes phragmosus
n. sp. from Tortuguilla, Guantanamo Province, Cuba. Scale bar equals 1 mm.

rior frons and postclypeus. Compound eyes almost
black. Mandibles chestnut brown. Antennal arti-
cles 1-3 ferruginous; remaining articles ferrugi-
nous orange. Anteclypeus yellowish. Ferruginous
orange chevron patterns formed by wing scales on
pterothorax faint and wide; remaining dorsum of
body pale orange-yellow. Sclerotized wing vena-
tion ferruginous orange, remainder of wings and
abdominal sternites yellowish.
In dorsal view, head capsule suboval with
sides along and anterior to eyes slightly concave;
posterior of head capsule broadly rounded. Head
converging to anterior in ventral aspect. In ob-
lique view, frons phragmotic, broadly excavated;
depression sharply delimited by moderately
raised ridge; surface of frons covered by dense
wrinkling of variable orientation (Fig. 2). A pair of
tiny tubercles behind ocelli; lateral branches of
epicranial suture near tubercles. In lateral view,
plane of frons margin slopes weakly toward a
slightly convex vertex. Compound eyes large and
protruding, subcircular; eye margins narrowly
subrectate along ocelli and along posteroventral
area, and broadly subrectate or slightly concave
along antennal sockets. Ocelli slightly protruding,
large, elliptical; contacting or very narrowly sepa-
rated from eyes; distinctly converging anteriorly.
Mandibular bases and anterolateral corners of
head capsule with distinct striations. Left mandi-
ble with slight hump at basal two-fifths; basal
hump with several ~0.03 mm long setae; first and

second marginal teeth partially separated; each
with separate pointed apex (Fig. 3); third mar-
ginal tooth with sinuous anterior and posterior
margins. Right mandible with molar plate longer
than posterior margin of second marginal tooth
and composed of ca. 20 ridges (Fig. 3).
Several dozen setae of medium length
(~0.05mm) dispersed on head, pronotum, wing
scales, abdominal tergites, and sternites. Anten-
nae with 18-24 articles, 75% (n = 64) with 22-24
articles, 10% with 24; relative length formula
2>3>4 = 5 or 2 = 3>4 = 5. Pronotum robust, about
twice as wide as its median length. Pronotum
with anterior margin evenly concave, lateral mar-
gins faintly convex, posterolateral margins sub-
truncate or faintly concave, and posterior margin
slightly concave medially; anterior and lateral
margins with raised and rounded rim. Fore wing
with very long subcosta and radius; subcosta ter-
minating at costal margin usually beyond 1/2 of
wing length from suture and near intersection of
radius and costal margin at 2/3 of wing length.
Radial sector with 4-6 branches that fork in apical
third of wing just beyond intersection of radius
into costal margin. Median vein sclerotized and
with about four sclerotized and short posterior
branches; branches dissolve gradually into mem-
brane except for usually the two most distal
branches, that terminate at wing margin. Wing
membrane faintly and irregularly nodulate with
some nodules fused. Arolia distinct.

March 2003

Krecek & Scheffrahn: New Neotermes from Cuba

Fig. 2. Photomicrograph of the oblique view of imago head of Neotermes phragmosus n. sp. from the U.S. Naval
Base, Guantanamo, Cuba, showing deeply excavated and phragmotic frons. Scale bar equals 1 mm.


The N. phragmosus imago is unique among
congeners in that its frons is characteristically
truncated, depressed, encircled by a ridge, and
rugose. Imagos ofN. phragmosus and the allopat-
ric N. mona are the largest among the West In-
dian Kalotermitidae. The N. phragmosus imago
has less dense pilosity than N. mona on the head,
pronotum, and wing scales. Few short setae on
basal hump of mandibles present in N. phragmo-
sus imago are absent both in N. mona and N.
Compared to the sympatric N. jouteli, N.
phragmosus alates differ primarily in size, the
first species being distinctly smaller than the sec-
ond one, usually without any overlapping. Those
most distinctive characters are: 1.77-2.16 mm for
head length with labrum ofN. jouteli, versus 2.24-
2.74 mm for N. phragmosus; labrum width maxi-
mum 0.60-0.70 mm versus 0.74-0.83 mm; prono-
tum maximum length is 1.06-1.32 mm of N.
jouteli, but 1.44-1.81 in N. phragmosus; and

pronotum width with 1.75-2.05 mm, while 2.10-
2.59, respectively. Total body length is also useful;
13.92-16.05 mm in N. jouteli, versus 15.80-19.04
mm in N. phragmosus.

Soldier. (Fig. 1, Tables 2 and 3).

The soldier caste consists of two distinct mor-
phs, large and small, both usually present in ma-
ture colonies. Other than size, there are few
distinguishing characters that separate small and
large soldiers of N. phragmosus compared with
some congeners and species in several other kalo-
termitid genera.
Head capsule and labrum ferruginous orange
in dorsal view. Antennae ferruginous orange;
three proximal articles ferruginous. Anteclypeus
pale yellowish. Mandibles glossy, almost black;
basal areas dark chestnut. Epicranial sutures
faint or absent. Eyes dark gray. Thorax, including
femora and abdominal dorsum and sternum pale
yellowish. Tibiae and genae pale orange-yellow.
Postmentum pale ferruginous.

Florida Entomologist 86(1)

Fig. 3. Scanning electron micrograph of imago mandibles ofNeotermes phragmosus n. sp., dorsal view, from Tor-
tuguilla, Guantanamo Province, Cuba. Labels: first marginal tooth (A), second marginal tooth (B) of left mandible,
and molar plate (C) of right mandible. Scale bar equals 0.5 mm. Labrum removed for clarity.

In dorsal view, head capsule subsquare, with
sides subparallel, faintly convex in large soldiers,
slightly convex in small morph; posterior corners
rounded and posterior margin widely rectate in
both morphs. Head capsule, thorax, and abdomen
covered with dense mat of long setae (~0.1 mm);
occiput glabrous. Frons depressed, faintly sub-
merged, and broadly continuous with postclypeus;
depressed area faintly striate. Frontal carinae ta-
pered into distinctly protruding tubercle near an-
tennal carinae. Labrum broadly linguiform; apex
slightly convex. Mandibles elongate and rela-
tively robust, with remarkably pilose basal
humps; dentition distinct. Small soldier antennae
with 17-21 articles, usually 18; large morph with
16-20 articles, usually 18 or 20; third antennal ar-
ticle subclavate, terminal articles usually slightly
elongate; antennal formula 2<3>4 = 5. Antennal
carinae protruding and faintly rugose. Pronotum
papilionaceous, noticeably wider than head, and
more than twice as wide as long in middle. Ante-
rior margin of pronotum deeply and evenly con-
cave; anterolateral corners abruptly rounded,
sides of pronotum subparallel, faintly convex; pos-
terior margin weakly emarginate. Pterothorax
with posterolateral sides subtruncate, more so in
small soldiers than in large soldiers. All soldiers
with short wing buds.

In lateral view, head capsule dorsoventrally
flattened; principal plane of frons occupying
about half of head capsule length in small sol-
diers; about one third in large morph. Frons
slopes =15 from plane of vertex; mandibles no-
ticeably curved upward; eyes large and vertically
oriented; without peripheral satellite facets. Pi-
losity of frons and anterior vertex dense. Hind
femora moderately broadened in small soldiers
and noticeably inflated in large morphs.


No single measurement in either soldier
morph is diagnostic for separating N. phragmosus
from its nearest congener, N. mona. Nevertheless,
the small morph ofN. phragmosus is larger in the
majority of measurements than that of N. mona.
The mandibular hump pilosity of N. phragmosus
is considerably more conspicuous than that of
both N. mona and N. jouteli. The N. phragmosus
soldiers possess a distinctly protruding tubercle
on each frontal carina, which, both in N. mona
and N. jouteli, are rudimentary. Striations of
frons in N. phragmosus are considerable, while
absent or very faint in N. mona. The rugosity of
antennal carinae is faint in N. phragmosus, while
being well developed in N. mona. The eyes of N.

March 2003

Krecek & Scheffrahn: New Neotermes from Cuba


Measurement in mm (n = 9 males, 10 females from 6 colonies) Range Mean + S.D. Holotype

Head length with labrum 2.24-2.74 2.58 + 0.12 2.64
Head length to postclypeus 1.56-2.00 1.86 + 0.12 1.93
Head width, maximum at eyes 1.83-2.27 2.11+ 0.093 2.12
Head height without postmentum 1.00-1.19 1.14 + 0.050 1.18
Labrum width, maximum 0.74-0.83 0.80 + 0.026 0.80
Eye diameter with sclerite, maximum 0.56-0.68 0.64 + 0.034 0.67
Eye to head base, minimum from sclerite 0.27-0.38 0.33 + 0.029 0.34
Ocellus diameter, maximum 0.18-0.26 0.23 + 0.020 0.22
Ocellus diameter, minimum 0.16-0.20 0.19 + 0.0098 0.19
Eye sclerite to ocellus, minimum 0-0.016 0.0090 + 0.0066 0.0082
Pronotum, maximum length 1.44-1.81 1.67 + 0.10 1.77
Pronotum, maximum width 2.10-2.59 2.44 + 0.13 2.59
Total length with wings 15.80-19.04 17.88 + 0.75 17.69
Total length without wings 8.91-13.10 11.35 1.09 12.29
Fore wing length from suture 11.07-13.77 12.93 0.61 13.23
Fore wing, maximum width 3.08-4.07 3.79 + 0.28 4.07
Hind tibia length 1.60-1.90 1.78 + 0.089 1.83

phragmosus do not display peripheral facets, 3.32-3.96 mm). Pronotal length of N. phragmosus
which are typical of N. mona. Finally, the anten- large soldiers ranges between 2.15-2.52 mm while
nae ofN. phragmosus soldiers have more articles in N jouteli the length is 1.71-1.85 mm. Both sol-
compared to those ofN. mona, in which the range dier morphs ofN. phragmosus are more pilose than
is 13-19, 12-18 in N. jouteli, while in N. phragmo- N. jouteli around the anterior portion of the head
sus it is 16-21. including mandible bases. The maximum head
Compared with the sympatric N. jouteli, N. width (2.93-3.46 mm) and left mandible length
phragmosus soldiers of both forms differ in having (2.64-2.90 mm) ofN. phragmosus large soldiers do
a much wider and much more deeply concave ante- not overlap with those respective measurements
rior margin of the pronotum. The character is par- (2.34-2.70 and 2.17-2.42 mm) in N. jouteli. Al-
ticularly distinctive in large soldiers (pronotum though some small soldier measurements overlap
width in N. jouteli ranges between 2.61-3.03 mm, for both species, the N. phragmosus small soldier is
while the same measurement in N. phragmosus is larger overall than that of N. jouteli.

80,00 W ,
* t '- A

-el "

,,, *.- -A -Q...

S 21.00 N
0 Neotermes phragmosus

A Termite Collection Sites
A0 A

100 km

Fig. 4. N rm rgm n. lclii nd rmi cllcin i n Cb nd nigbring ilnd.
Fig. 4. Neotermes phragmosus n. sp. localities and termite collection sites on Cuba and neighboring islands.

Florida Entomologist 86(1)

March 2003


Measurement in mm (n = 12 from 7 colonies) Range Mean + S.D.

Head length to tip of mandibles 3.91-5.30 4.69 + 0.40
Head length to postclypeus 2.43-3.47 3.04 + 0.32
Head width, maximum 2.28-3.10 2.77+ 0.24
Antennal carinae, outside span 2.04-2.60 2.35+ 0.16
Head height, excluding postmentum 1.34-1.83 1.53+ 0.15
Labrum, maximum width 0.64-0.82 0.73 + 0.053
Postclypeus width, maximum 0.87-1.10 0.98+ 0.066
Left mandible length, tip to most distant visible point of ventral condyle 2.17-2.69 2.42 + 0.15
Postmentum, length in middle 1.88-2.47 2.17 + 0.20
Postmentum, maximum width 0.80-1.11 0.93+ 0.087
Postmentum, minimum width 0.49-0.60 0.54 0.045
Pronotum, maximum width 2.69-3.36 3.07 0.19
Pronotum, maximum length 1.63-2.20 1.93 0.16
Hind tibia length 1.38-1.95 1.71 0.15
Total length 9.72-14.85 12.48 + 1.67

Etymology. alates were collected in late August and early No-
The species name reflects the unique and
striking phragmosis of the imago frons; possibly Type material.
the most developed for this character among the
Isoptera. Holotype colony series. Cuba. Guantanamo
Province; Tortuguilla; 19.98N, 74.93W; 20-VIII-
Remarks. 1974; coll. J. Krecek; 1 female alate holotype, 13
alate paratypes, 6 paratype small soldiers and 6
The holotype colony was collected in a very xe- paratype large soldiers (CU-968).
ric coastal habitat from the dead wood of living Paratype colonies series. All material origi-
Calotropis procera Aiton (Asclepiadaceae), an ex- nates from Guantanamo Prov.: Imias; 20.07N,
otic shrub. The colony penetrated into xylem ele- 74.64W; VIII-1975; coll. L. de Armas; 1 paratype
ments within the living cambium. Other colonies small and large soldier (CU-1038). The following
were collected from dead branches and trunks of samples were collected at the U.S. Naval Base
mangroves, buttonwood, and other littoral woods. Guantanamo Bay by J. Chase, J. Mangold, and
The dispersal flight season of N. phragmosus is R.H. Scheffrahn 2-XI-2001 to 6-XI-2001: Kittery
unknown, but we suspect nocturnal autumn Beach; 19.906N, 75.089W; 1 paratype imago
flights similar to those of others congeners as (CU-1076); N. Kittery Beach; 19.905N, 75.0880W;

Table 3. Measurements ofNeotermes phragmosus large soldier.

Measurement in mm (n = 11 from 6 colonies) Range Mean + S.D.

Head length to tip of mandibles 5.30-6.09 5.69 + 0.22
Head length to postclypeus 3.61-4.16 3.87 + 0.17
Head width, maximum 2.93-3.46 3.25 + 0.17
Antennal carinae, outside span 2.54-2.97 2.78 + 0.13
Head height, excluding postmentum 1.83-2.30 2.15 + 0.14
Labrum, maximum width 0.70-0.83 0.78 + 0.049
Postclypeus width, maximum 1.06-1.21 1.13 + 0.046
Left mandible length, tip to most distant visible point of ventral condyle 2.64-2.90 2.77 + 0.080
Postmentum, length in middle 2.57-3.03 2.79 + 0.14
Postmentum, maximum width 0.93-1.14 1.06 + 0.078
Postmentum, minimum width 0.47-0.65 0.58 + 0.063
Pronotum, maximum width 3.32-3.96 3.66 + 0.16
Pronotum, maximum length 2.15-2.52 2.33 + 0.11
Hind tibia length 1.75-2.10 2.01 + 0.10
Total length 12.83-16.07 13.96 + 1.13

Krecek & Scheffrahn: New Neotermes from Cuba

1 paratype small and large soldier (CU-1343); Old
Chief's Club; 19.925N, 75.131W; 1 paratype
imago (CU-1374); Boat landing, leeward man-
groves; 19.941N, 75.152W; 1 paratype large sol-
dier (CU-1401); 1 paratype small soldier (CU-
1408); Naval Station Brig; 19.936N, 75.124W; 1
paratype small soldier (CU-1430); 1 paratype
imago (CU-1433); Evan's Point; 19.921N,
75.141W; 1 paratype imago (CU-1447); 1
paratype small and large soldier (CU-1448); Lee-
ward mangroves, Coccothrinax habitat; 19.958N,
75.165W; 1 paratype imago and small soldier
(CU-1521), 1 paratype large soldier (CU-1523).


The characters of Neotermes phragmosus re-
quire that morphological definitions for the Kalo-
termitidae be broadened for both the imago and
soldier. Plesiomorphic traits of N. phragmosis
outside of Krishna's (1961) imago diagnosis in-
clude: 1) a maximum of 24 antennal articles (in-
crease of 3), 2) separation of the second and third
marginal teeth of the left mandible, 3) the molar
plate of the right mandible longer than the poste-
rior margin of the second marginal tooth, and 4)
fore wing subcosta extending to at least mid wing
with radius intersecting costal margin well be-
yond mid wing. In the soldier, the number of an-
tennal articles is increased from 19 to 21. It is
noteworthy that for the soldier, Kambhampati &
Eggleton (2000) use the threshold gap of 20-22
antennal articles to separate the Termopsidae
from the Kalotermitidae.
Although a weak frontal concavity and rudi-
mentary phragmosis occur in several Neotropical
Neotermes images, i.e.N. jouteli (Scheffrahn et al.
2000), N. mona (Krecek et al. 2000), and N. platy-
frons (Krecek & Scheffrahn 2001), the degree of
its development in N. phragmosus is remarkable
and suggests apomorphism for defense of incipi-
ent colonies against predatory ants or competi-
tion by termites vying for nuptial microhabitats.
The evolutionary significance of pilosity of the
mandibular humps in the soldier is unclear. Man-
dibular basal pilosity is not uncommon in Neoter-
mes; it appears also in Glyptotermes,
Paraneotermes, and Incisitermes, but this trait
reaches its maximum expression in N. phragmo-
Together with the Antillitermes subtilis
(Scheffrahn & Krecek 1993), Constrictotermes
guantanamensis Krecek et al. (1996), Cryptoter-
mes spathifrons, and C. cymatofrons (Scheffrahn
& Krecek 1999), N. phragmosus is the fifth spe-
cies recently described from southeastern Cuba.

All species but C. cymatofrons are confined to xe-
ric habitats.


The authors thank James A. Chase and John R.
Mangold, Terminix International, and Luis F. de Armas,
Cuban Academy of Sciences, for specimen collection;
Tom Drake, Wildlife Technician; Paul Schoenfeld, Natu-
ral Resources Manager; Patricia Loop, Environmental
Director; USNB Guantanamo Bay, Cuba, for logistical
support; Diann Achor, University of Florida, Lake Al-
fred Citrus Research and Education Center, for assist-
ing with scanning electron microscopy; Lyle Buss and
Brian J. Cabrera for assisting with light photomicros-
copy; and William Kern Jr. and B. Cabrera for their crit-
ical reviewing of this manuscript. Florida Agricultural
Experiment Station Journal Series No. R-08789.


and phylogeny of termites, pp. 1-23. In Abe, T., D.E.
Bignell, and M. Higashi (eds.): Termites: Evolution,
Sociality, Symbioses, Ecology. Kluwer Acad. Publ.,
Dordrecht, Netherlands.
KRECEK, J., AND R.H. SCHEFFRAHN. 2001. Neotermes
platyfrons, a new dampwood termite (Isoptera, Kal-
otermitidae) from the Dominican Republic. Florida
Entomol. 84: 70-76.
Greater Antillean Nasutitermitinae (Isoptera, Ter-
mitidae): Constrictotermes guantanamensis, a new
subterranean termite from eastern Cuba. Florida
Entomol. 79: 180-187.
KRECEK, J., N.-Y. Su, AND R. H. SCHEFFRAHN. 2000. Re-
description of Neotermes mona, a dampwood termite
(Isoptera, Kalotermitidae) from the central West In-
dies. Florida Entomol. 83: 268-275.
KRISHNA, K. 1961. A generic revision and phylogenetic
study of the family Kalotermitidae (Isoptera). Bull.
American Mus. Nat. Hist. 122: 303-408.
SCHEFFRAHN, R. H., AND J. KRECEK. 1993. Parvitermes
subtilis, a new subterranean termite (Isoptera: Ter-
mitidae) from Cuba and the Dominican Republic.
Florida Entomol. 76: 603-607.
SCHEFFRAHN, R. H., AND J. KRECEK. 1999. Termites of
the genus Cryptotermes (Isoptera: Kalotermidae)
from the West Indies. Insecta Mundi 13: 111-171.
LINS, J. KRECEK, AND N.-Y. Su. 1994. Termites
(Isoptera: Kalotermitidae, Rhinotermitidae, Ter-
mitidae) of the West Indies. Sociobiology 24: 213-
SCHEFFRAHN, R. H., J. KRECEK, AND N.-Y. Su. 2000. Re-
descriptions of the dampwood termites Neotermes
jouteli and N. luykxi (Isoptera: Kalotermitidae) from
Florida, Bahamas, and Turks and Caicos Islands.
Ann. Entomol. Soc. America 93: 785-794.
SNYDER, T. E. 1956. Termites of the West Indies, the Ba-
hamas, and Bermuda. J. Agric. Univ. Puerto Rico 40:

Florida Entomologist 86(1)

March 2003


USDA-ARS, U.S. Horticultural Research Laboratory, 2001 South Rock Road, Ft. Pierce, FL 34945


Three legume species with potential as cover crops in citrus groves were studied for their ef-
fect on the developmental biology of the Diaprepes root weevil, Diaprepes abbreviatus (L.) in
greenhouse studies. All 3 cover crops were hosts for the Diaprepes root weevil. Cajanus ca-
jan (pigeon pea) was a superior host for development ofD. abbreviatus compared with citrus
rootstocks. C. cajan appeared to be allelopathic; the root mass of uninfested citrus was
greatly reduced when grown in association with C. cajan compared with citrus grown alone.
Association of citrus with C. cajan orArachis pintoi (perennial peanut) reduced chlorophyll
fluorescence, a measure of photosynthesis, compared with citrus associated with Crotalaria
pallida rattleboxx) or with another citrus seedling. When grown in close association withA.
pintoi, citrus produced the same amount of root mass as citrus seedlings grown alone. Infes-
tation with larval D. abbreviatus reduced chlorophyll fluorescence of citrus by 26%. None of
the 3 legume species tested reduced the feeding damage caused by D. abbreviatus to citrus.
Larvae reared in pots with A. pintoi, associated with citrus or alone, gained weight at the
same rate as larvae reared on the citrus rootstocks alone. Larvae recovered from pots con-
taining C. pallida associated with citrus weighed significantly more than larvae reared on
citrus alone. C. cajan appears to be particularly inappropriate as a cover crop because of its
positive effect on larval growth and reduction of citrus root mass. None of the 3 legume spe-
cies tested had a negative effect on D. abbreviatus or on feeding damage.

Key Words: Citrus, Diaprepes abbreviatus, cover crops, perennial peanut, Arachis pintoi, pi-
geon pea, Cajanus cajan, Crotalaria pallida


Tres species de leguminosas forrajeras con potential como cobertura en citricos fueron eva-
luadas por su efecto sobre el desarrollo del cucarron Diaprepes abbreviatus (L.) en un inver-
nadero. Las tres species sirvieron como hospedantes para el insecto. Cajanus cajan
(gandul) fue un hospedante superior para el desarrollo de D. abbreviatus comparado con pa-
trones de citricos. C. cajan fue alelopatico; la masa de raices de citricos no-infestadas fue me-
nor en asociaci6n con C. cajan comparado con citricos solos. La asociaci6n de citricos con C.
cajan o Arachis pintoi (mani forrajero) result en una reducci6n de fluroscencia de clor6filo,
una media de fotosintesis, comparado con citricos asociados con Crotalaria pallida o con
otra plantula de citricos. Citricos asociados con A. pintoi produjeron la misma cantidad de
raices que los citricos solos. La infestaci6n con larvas de D. abbreviatus result en una reduc-
ci6n de 26% en fluroscencia de clor6filo de plantulas de citricos. Niguna de las tres species
de leguminosas redujo el daio causado por alimentaci6n de D. abbreviatus en citricos. Las
larvas en potes conA. pintoi (asociado con citricos o solo) aumentaron de peso de la misma
manera que larvas sobre raices de citricos solo. Larvas recuperadas de potes que contenian
C. pallida asociado con citricos pesaron mas que larvas sobre citricos solo. Parece que C. ca-
jan en particular no es recomendable como cobertura debida a su efecto positive sobre el cre-
cimiento de larvas y su efecto negative sobre la masa radicular de citricos. Ninguna de las
tres leguminosas tuvo un efecto negative sobre D. abbreviatus o redujo su daio. Translation
provided by author.

Leguminous cover crops can contribute to in- selected, at a minimum, that do not harbor key
creased and sustainable crop productivity pests and may be selected to divert or deter pests
through erosion and weed control, biological ni- and contribute to the diversity and abundance of
trogen fixation, and by providing refuge for natu- natural enemies (Altieri 1995, Risch 1981).
ral enemies of arthropod pests (Hokkanen 1991). A major concern to Florida citrus producers is
Cover crops such as perennial peanut (Arachis the highly polyphagous Diaprepes root weevil, Di-
spp.) have been suggested for use in citrus groves aprepes abbreviatus (L.) (Simpson et al. 1996).
(Prine et al. 1981), but adoption of this practice Cover crops or trap crops could contribute to cit-
will depend on an unequivocal demonstration of rus productivity and control of damage from D.
benefits to grove managers. Cover crops should be abbreviatus. It is equally possible, however, that

Lapointe: Cover crops, citrus and Diaprepes abbreviatus

introduction into the citrus cropping system of an
additional plant resource for a highly polypha-
gous pest could result in higher pest population
density (Andow 1991), particularly in the case of
D. abbreviatus in Florida where natural enemies
are insignificant (Hall et al. 2001)
As a first attempt to study the potential influ-
ence of cover legumes on the biology of D. abbre-
viatus, I examined the response of larvae to 3
legume species in a greenhouse. Arachis pintoi
Krapovickas & Gregory is increasingly used as a
tropical forage and as a cover in diverse tropical
tree crops (de la Cruz et al. 1993) and has been
considered for use in Florida citrus groves. Caja-
nus cajan Millspaugh (pigeon pea) is widely
grown in Puerto Rico where anecdotal observa-
tions suggest a strong preference for this species
by D. abbreviatus. Crotalaria pallida Ait. (rattle-
box) has been used extensively in Florida as a
green manure and has become naturalized. I re-
port here the effect of these species alone and as-
sociated with citrus on development of
D. abbreviatus.


Trial I. Effect of Plant Associations in 3.8-L-pots.

Seed of C. cajan, A. pintoi and C. pallida were
planted in germination trays in a soilless potting
mix (Metromix 500, Scotts, Marysville, OH) and
transplanted at approximately 1 mo after germi-
nation. Seedlings of'Carrizo' citrange (C. sinensis
(L.) Osbeck x P trifoliata) were germinated in
fine, sterile sand (Bonsal Play Sand, W. R. Bonsal
Co., Charlotte, NC) and transplanted to 3.8-L-
pots at 4 mo after germination. All plants were
transplanted to 3.8-L-pots containing sterile
sand. The pots were lined with a nylon mesh cloth
to prevent escape of larvae through the drainage
holes. Pots were planted with 2 seedlings per pot
in the following combinations: two seedlings of ei-
ther 'Carrizo',A. pintoi, C. cajan, or C. pallida; or
one seedling of'Carrizo' and a companion plant of
A. pintoi, C. cajan, or C. pallida. An additional
combination consisted of one plant of C. cajan and
one plant ofA. pintoi. Treatments consisted of 3
weevil-infested and 3 noninfested pots of each of 8
plant combinations in a randomized block design.
A total of 48 pots were arranged randomly on
greenhouse benches within infested and nonin-
fested blocks. Ten early instar larvae ofD. abbre-
viatus weighing 20 + 5 mg each were added to
infested pots on 30 June 1998. Larvae were ob-
tained from a laboratory colony maintained by the
U.S. Horticultural Research Laboratory, Orlando,
FL and reared according to Lapointe & Shapiro
(1999). Early instars were used instead of neo-
nates to avoid escape or movement of larvae be-
tween pots. Citrus seedlings were one year old
and legume seedlings were 3 mo. old at the begin-

ning of the infestation period. Plants were main-
tained throughout the experiment on elevated
benches in a greenhouse with an average diurnal
temperature cycle of 35C maximum and 23C
minimum. No supplemental light was supplied.
Maximum photosynthetic photon flux in the green-
house was 800 mol.s-.m-2. Plants were watered
with a dilute fertilizer mix weekly using water-sol-
uble 20N-10P-20K at a rate of 150 mg.liter1N.
At the end of the infestation period, an OS5-FL
modulated chlorophyll fluorometer (Opti-Sci-
ences, Tyngsboro, MA, USA), was used to mea-
sure yield of chlorophyll fluorescence of light-
adapted leaves with saturation intensity of ~2.7
kuE for 0.8 sec. Yield (relative units) is an indica-
tor of quantum yield of photosynthesis and is of-
ten used in measuring plant stress (Schreiber &
Bilger 1987, van Kooten & Snel 1990). Chloro-
phyll fluorescence was measured during the late
morning hours (9:00-11:00). All readings (Y val-
ues) were taken from the center of a leaf, and al-
ways from the center leaflet in the case of
trifoliate leaves. Readings were taken from 3 po-
sitions on each plant:'top' was taken from the first
fully expanded leaf, 'bottom' was taken from the
lowest available intact leaf, and 'middle' was
taken from the leaf at the mid-point between'top'
and 'bottom'.
The plants and sand were removed from the
pots and sieved to recover larvae 37 d after infes-
tation on 6 August 1998. Recovered larvae were
counted and weighed before and after drying in an
analytical oven at 60C for >48 h to obtain fresh
and dry weights. Roots were washed and sepa-
rated from the above-ground portion by cutting at
the point where the first (uppermost) lateral root
emerged from the central root. Roots were al-
lowed to air-dry and weighed.
The effect of plant combination on total larval
weight per pot, and the effect of infestation and
plant combination on citrus root weight and chlo-
rophyll fluorescence were compared by ANOVA.
Where appropriate, means were compared by
Tukey's Honestly Significant Differences (HSD)
test (Abacus Concepts 1996).

Trial II. Effect of Plant Associations in 76-L-pots.

Eighteen 1-year-old seedlings of the rootstock
'Sun Chu Sha' mandarin (C. reticulata Blanco)
were transplanted to 76-L-pots containing sterile
sand in May, 1998. Availability of plants deter-
mined choice of rootstock for this trial. However,
'Carrizo' and'Sun Chu Sha' have been shown to be
equivalent in terms of larval weight gain of D. ab-
breviatus reared on potted seedlings, and in terms
of damage to roots of seedlings infested with
D. abbreviatus (Lapointe et al. 1999).
Plant combinations in the pots consisted of ei-
ther a single tree of'Sun Chu Sha' or a seedling of
'Sun Chu Sha' surrounded by 10 seedlings of

Florida Entomologist 86(1)

A. pintoi or 6 seedlings of C. cajan. Insect treat-
ments consisted of 3 infested and 3 noninfested
pots of each plant combination. Pots were infested
1 July 1998 with 50 larvae each with a mean (+
SEM) individual weight of 71.7 1.9 mg. Roots
and larvae were recovered 40 d later on 10 August
1998. To estimate larval number per pot, soil was
sifted and scanned for 15 min. Of the larvae recov-
ered, 15 were randomly selected from each pot,
weighed, dried in an analytical oven and weighed
again. Roots were allowed to air dry and then
Larval weights were summed for each pot. The
effect of plant combination on total larval weight
per pot, and the effect of infestation and plant
combination on citrus root weight were compared
by ANOVA. Experimental design was a 2 x 3 fac-
torial with 2 levels of infestation and 3 plant asso-
ciations. Where appropriate, means were
compared by Tukey's Honestly Significant Differ-
ences (HSD) test and groups of means (e.g., all
pots containing >1 plant of C. cajan) by post-hoc
orthogonal contrasts (Abacus Concepts 1996).


Trial I. Effect of Plant Associations in 3.8-L-pots.

Both companion plant and state of infestation
had a significant effect on chlorophyll fluores-
cence of'Carrizo' leaves. There was no significant
effect from leaf position (F = 2.4; df = 2, 72; P =
0.10), interaction between infestation and com-
panion plant (F = 1.8; df = 3, 72; P = 0.16) or be-
tween infestation and position (F = 0.5; df = 2, 72;
P = 0.59). For analysis of effects of infestation and
companion plant, the measures of chlorophyll flu-
orescence for top, middle, and bottom 'Carrizo'
leaves were pooled.
Infestation with larval D. abbreviatus reduced
chlorophyll fluorescence of the 'Carrizo' plants by
26 + 8% compared with noninfested controls (F =
23.7; df = 1, 72; P < 0.01). The species of compan-
ion plant also significantly affected chlorophyll
fluorescence of 'Carrizo' (F = 3.3; df = 3, 72; P =
0.03). 'Carrizo' planted with another 'Carrizo'
plant or planted with C. pallida had higher levels
of chlorophyll fluorescence compared with 'Carr-
izo' plants planted with either A. pintoi or C. ca-
jan. Association withA. pintoi or C. cajan reduced
photosynthesis fluorescence in 'Carrizo' by 20%
(Table 1).
There was neither an effect of companion (F =
0.9; df= 2, 66; P = 0.40) nor of infestation (F = 0.1;
df = 1, 66; P = 0.73) on the chlorophyll fluores-
cence of C. cajan. There was an effect of leaf posi-
tion on chlorophyll fluorescence (F = 9.1; df= 2,66;
P < 0.01). Top and middle C. cajan leaves had
higher levels of fluorescence (0.514 + 0.015 and
0.482 0.024, respectively) than lower leaves
(0.389 0.024) (Tukey's HSD, a = 0.05).


Companion Yield n

A. pintoi 0.302 + 0.028 a 18
C. cajan 0.308 + 0.032 a 18
Citrus 0.386 + 0.020 b 36
C. pallida 0.392 + 0.045 b 18

Means followed by the same letter are not significantly different at P
= 0.05 by Tukey's HSD after a significant ANOVA (F = 3.3; df = 3, 72; P=

The chlorophyll fluorescence ofA. pintoi was un-
affected by companion (F = 2.3; df= 2, 66; P = 0.11)
and infestation (F = 0.9; df = 1, 66; P = 0.35). As in
the case ofC. cajan, there was a significant effect of
leaf position (F = 4.5; df= 2,66;P = 0.01); top leaves
had a higher rate (0.433 0.019) compared with
middle and lower leaves (0.344 + 0.024 and 0.350 +
0.028, respectively) (Tukey's HSD, a = 0.05).
There was no effect of companion (citrus or an-
other C. pallida) (F = 0.54; df = 1, 49; P = 0.47) nor
of infestation (F = 0.004; df = 1, 49; P = 0.95) on
the chlorophyll fluorescence of C. pallida. There
was a significant effect of leaf position (F = 4.2; df
= 2,49; P = 0.02); top leaves had a higher rate
(0.465 0.027) than lower leaves (0.332 0.038)
and middle leaves were intermediate (0.372
0.032), not significantly different from either top
or lower leaves (Tukey's HSD, a = 0.05).
Both companion plant (F = 7.1; df = 3, 22; P <
0.01) and infestation (F = 66.9; df= 1, 22;P < 0.01)
significantly affected final root weight of'Carrizo'
plants in 3.8-L-pots, and there was a significant
interaction between companion plant and infesta-
tion (F = 3.9; df = 3, 22; P = 0.02). For this reason,
root weights were analyzed separately for in-
fested and noninfested groups (Table 2). Nonin-
fested 'Carrizo' plants in 3.8-L-pots had a smaller
root mass when associated with C. cajan, but not
when associated withA. pintoi or C. pallida, com-
pared with 'Carrizo' plants associated with an-
other 'Carrizo'. 'Carrizo' root weight was greater
when grown withA. pintoi than with C. pallida or
C. cajan (Table 2).
The root mass of all 'Carrizo' plants infested
with D. abbreviatus was significantly reduced. In
the case of'Carrizo' associated with'Carrizo', pots
contained 2 plants and therefore twice as much
citrus root mass was available compared with
'Carrizo' associated with a legume species. Appar-
ently, weevils fed equally on'Carrizo' regardless of
companion plant, i.e., none of the associations re-
sulted in reduced feeding damage to'Carrizo' (Ta-
ble 2). If the interaction term is ignored by setting
a at 1%, the main effect of infestation byD. abbre-
viatus reduced 'Carrizo' root weight by 68 7%

March 2003

Lapointe: Cover crops, citrus and Diaprepes abbreviatus


Citrus root weight (g)

Companion Infested" Noninfested" n

C. pallida 0.23 + 0.03 a 1.75 + 0.38 ab 3
C. cajan 0.39 + 0.17 ab 1.19 + 0.10 a 3
A. pintoi 0.49 + 0.19 ab 2.99 + 0.47 c 3
Citrus 1.12 + 0.21 b 2.21 + 0.18 bc 6

Means in a column followed by the same letter are not significantly different at P = 0.05 by Tukey's HSD after a significant ANOVA.
F = 4.6; df= 3, 11; P= 0.03.
F =6.0; df= 3,11; P= 0.01.

compared with noninfested 'Carrizo'. Similarly,
the main effect of companion plant was highly sig-
nificant (F = 7.1; df = 3, 22; P < 0.01). When C. ca-
jan or C. pallida were the companion plants, root
weight of'Carrizo' was reduced by 53 and 41%, re-
spectively. However, when associated with A. pin-
toi, root weight of'Carrizo' was equal to the root
weight of individual 'Carrizo' trees planted with a
second 'Carrizo' (a = 0.05, Tukey's HSD).
There was a significant effect of plant associa-
tion on total wet weight of larvae recovered from
each pot (F = 35.4; df = 7, 15; P < 0.01), total dry
weight of recovered larvae (F = 59.7; df = 7, 15; P
< 0.01) and on the number of larvae recovered per
pot (F = 8.3; df = 7, 16; P < 0.01). In the 3 associa-
tions that included C. cajan, more (F = 46.9; df =
1; P < 0.01) and larger (wet weight: F = 242.0; df
= 1; P < 0.01) larvae were recovered compared
with pots containing 2 'Carrizo' plants (Table 3).
The other 2 legume species did not affect the num-
ber or size of larvae recovered compared with pots
containing 2 'Carrizo' plants.

Trial II. Effect of Plant Associations in 76-L-pots.

There was a significant effect of association (F
= 9.1; df = 2, 12; P < 0.01) and of infestation (F =

11.1; df= 1, 12;P < 0.01) on weight of citrus roots.
There was also a significant interaction between
these effects (F = 4.3; df = 2, 12; P = 0.04). Citrus
root weight differences between noninfested and
weevil-infested trees differed between citrus
alone and citrus with A. pintoi compared with cit-
rus associated with C. cajan. There was no signif-
icant difference in citrus root weights between
infested or noninfested citrus in association with
C. cajan, whereas citrus root weights of nonin-
fested citrus alone or in association with A. pintoi
were increased by 47-55% over infested citrus
(Fig. 1).
The effect of infestation was therefore tested
for each treatment (association) separately, and
treatment means were compared within infested
and noninfested groups. Uninfested 'Sun Chu
Sha' plants associated with C. cajan had much
smaller root mass compared with 'Sun Chu Sha'
alone or 'Sun Chu Sha' associated with A. pintoi
(Fig. 1). Root mass of'Sun Chu Sha' was equiva-
lent for all 3 infested treatments ('Sun Chu Sha'
alone, 'Sun Chu Sha'/A. pintoi, and 'Sun Chu Sha'
C. cajan).
The effect of plant association on final larval
fresh weight of 15 larvae recovered from each pot
was significant at P = 0.06 (ANOVA). However,


Larval weight (mg)

Association No. larvae recovered Wet" Dry'

Citrus/Citrus 1.7 + 0.3 a 70.4 + 23.4 a 16.2 + 7.7 a
C. pallida/C. pallida 1.7 + 0.9 a 87.8 + 23.4 a 13.6 + 4.2 a
Citrus/C. pallida 1.7 + 0.3 a 161.1 + 24.0 a 38.8 + 8.2 b
A. pintoi/A. pintoi 3.3 + 0.9 ab 166.5 + 47.0 a 30.0 + 9.4 a
Citrus/A. pintoi 3.3 + 0.3 ab 210.7 + 7.1 a 40.1 + 0.6 a
Citrus/C. cajan 5.0 + 0.6 b 852.8 + 53.7 b 280.8 + 32.4 b
C. cajan/C. cajan 5.0 + 0.6 b 966.9 + 145.8 b 311.2 + 42.5 bc
C. cajan/A. pintoi 5.7 + 0.3 b 1027.0 + 95.0 b 399.4 + 7.3 c

Means in a column followed by the same letter are not significantly different at P = 0.05 by Tukey's HSD after a significant ANOVA.
F = 8.3; df= 7,16; P < 0.01
F = 35.4; df = 7,15; P < 0.01
F = 59.7; df= 7,15; P < 0.01

Florida Entomologist 86(1)


1.5 -

1.0 -

0.5 -



Fig. 1. Interaction plot of effect of infestation
with larval D. abbreviatus on fresh root weight
(+SE, n = 3) of citrus trees grown in 76-L-pots
alone or in association with 2 legumes. Means fol-
lowed by the same letter are not significantly dif-
ferent (a = 0.05, Tukey's HSD); percent reduction
of root weight for infested vs. noninfested associa-
tions was significant (*) or not (N.S.) by unpaired
t test (a = 0.05).

the effect of plant association on final larval dry
weight of 15 larvae was highly significant (F =
23.5; df = 2, 6; P < 0.01). The dry weight of larvae
recovered from pots containing 'Carrizo'/C. cajan
was 38% greater than that of larvae from the
other 2 treatments (Table 4).


Damage by subterranean larvae is difficult to
detect and plant damage often is not evident until
days or weeks after feeding occurs. In this trial,
we were unable to visually detect differences be-
tween infested and noninfested plants over the
period of infestation (1 mo). Indeed, it is not un-

O Citrus
* Citrus/A. pintoi
O Citrus/C. cajan

55% -
/ a

N.S. -19% N.S.


Larval weight (mg)

Association Initial' Final wet" Final dry'

Citrus/A. pintoi 73.1 + 2.6 a 2480.3 + 185.3 a 714.3 + 60.4 a
Citrus/Sun Chu Sha 71.1 + 2.8 a 2604.4 + 218.1 a 767.3 + 66.2 a
Citrus/C. cajan 70.9 + 5.1 a 3207.3 + 128.8 a 1191.3 + 27.3 b

"Means are not significantly different by one-way ANOVA (F = 0.1; df = 2, 6; P= 0.90).
'Means are not significantly different by one-way ANOVA (F = 4.5; df = 2, 6; P = 0.06).
'Means in a column followed by the same letter are not significantly different at a = 0.05 by Tukey's HSD following a significant ANOVA (F = 23.5; df
=2, 6; P < 0.01).

common in these tests to find seedlings with ex-
tensive root damage without any visual foliar
symptoms. However, we were able to detect a sig-
nificant reduction in chlorophyll fluorescence due
to larval feeding using a fluorometer. This method
may be useful in citrus groves as an indicator of
tree health in general and presence of root weevils
in particular.
Ground covers offer advantages and disadvan-
tages when incorporated into agricultural produc-
tion systems. In recent years, the tropical forage
A. pintoi has shown potential for use as a cover
crop in tropical tree crops such as coffee, banana,
oil palm, macadamia, and heart-of-palm (de la
Cruz et al. 1993) and has been proposed for use in
Florida's subtropical citrus groves (Prine et al.
1981). While slow to establish, A. pintoi is effec-
tive at weed suppression, has a non-twining
growth habit, and is efficient at fixation of atmo-
spheric nitrogen (Thomas 1993). For citrus, these
attributes must be considered in relation to the
potential for nutrient competition, ease of man-
agement, and effect on pests and diseases. This
study indicates that A. pintoi is the most appro-
priate of the 3 species studied here as a cover crop
in citrus in terms of its effect on a major pest,
D. abbreviatus.
C. pallida did not decrease chlorophyll fluores-
cence or root growth by the citrus rootstock. The
larvae recovered from pots containing C. pallida
associated with rootstock, however, weighed sig-
nificantly more than larvae reared on rootstock
alone. This, combined with the upright, annual
growth habit of C. pallida make this legume a less
desirable option as a cover crop compared with A.
C. cajan (pigeon pea) has been reported to be
attractive to D. abbreviatus (Barrow 1924). In the
tests reported here, C. cajan was a superior host
for development ofD. abbreviatus compared with
'Carrizo'. More and larger larvae survived in pots
when C. cajan was present, regardless of associa-
tion with another plant species (Tables 3 and 4).
In addition, the root mass of noninfested rootstock
seedlings was greatly reduced when grown in as-

-4 .

March 2003

Lapointe: Cover crops, citrus and Diaprepes abbreviatus

sociation with C. cajan compared with rootstock
grown alone or in association withA. pintoi (Table
2, Fig. 1). The combination of increased growth of
Diaprepes root weevil and apparent allelopathic
effects on citrus makes C. cajan a particularly in-
appropriate choice for a cover crop.
When A. pintoi was grown in close association
with a citrus rootstock in 3.8-L-pots, the rootstock
produced the same amount of root mass as citrus
plants grown alone (Table 2, Fig. 1). Although
none of the 3 legume species tested reduced the
feeding damage caused by D. abbreviatus to the
rootstock, larvae reared in associations that in-
cludedA. pintoi gained weight at the same rate as
larvae reared on citrus alone (Table 3). Similarly,
larvae reared in pots with A. pintoi alone gained
the same amount of weight as larvae reared on
citrus alone (Table 3). Simpson et al. (1996) re-
ported larvae feeding on the roots of peanut (Ara-
chis hypogaea) and the results presented here
indicate thatA. pintoi is also a host ofD. abbrevia-
tus. This presents the danger of increased pest
populations in citrus/A. pintoi polycultures, but
also the possibility of diversion of larval infesta-
tion from the principal crop (citrus) to the cover
crop. Andow (1991) surveyed published reports of
the effect of crop diversity on pest density and
found that a minority (15%) of species were more
abundant in polycultures while 52% were less
abundant compared with monocultures. Attempts
to establish cover crops in citrus should monitor
key pests such as D. abbreviatus. Although none
of the species tested here had negative effects on
D. abbreviatus larvae, legumes are known to be a
rich source of phytochemicals with diverse insect
antifeedant and toxic properties (Simmonds et al.
1990) and should be surveyed for their activity
against D. abbreviatus for possible inclusion in
citrus production systems


I thank Hunter Smith for technical assistance with
glasshouse bioassays. Drs. Richard Mayer, Jeffrey Sha-
piro, Clayton McCoy and an anonymous reviewer pro-
vided helpful comments on the manuscript. Mention of a
trademark or proprietary product does not constitute a
guarantee or warranty of the product by the U.S. Depart-
ment of Agriculture and does not imply its approval to
the exclusion of other products that may also be suitable.


ABACUS CONCEPTS. 1996. StatView Reference. Berkeley,
ALTIERI, M. A. 1995. Biodiversity and biocontrol: les-
sons from insect pest management. Adv. Plant
Pathol. 11: 191-209.

ANDOW, D. A. 1991. Vegetational diversity and arthro-
pod population response. Annu. Rev. Entomol. 36:
BARROW, E. H. 1924. White grubs, Lachnosterna sp.,
and larvae of the weevil root-borer, Diaprepes
spengleri L., attacking sugar cane in the Guanica
district of Porto Rico, and methods practised for
controlling them. J. Dept. Agric. Puerto Rico 8: 22-
The contribution of Arachis pintoi as a ground cover
in some farming systems of tropical America, pp.
102-108. In P. C. Kerridge and B. Hardy [eds.], Biol-
ogy and agronomy of forageArachis. Centro Interna-
cional de Agricultura Tropical (CIAT), Cali,
AND B. BULLOCK. 2001. Status of biological control
by egg parasitoids of Diaprepes abbreviatus (Co-
leoptera: Curculionidae) in citrus in Florida and Pu-
erto Rico. BioControl 46: 61-70.
HOKKANEN, H. M. T. 1991. Trap cropping in pest man-
agement. Annu. Rev. Entomol. 36: 119-138.
LAPOINTE, S. L., AND J. P. SHAPIRO. 1999. Effect of soil
moisture on development of Diaprepes abbreviatus
(Coleoptera: Curculionidae). Florida Entomol. 82:
1999. Identification of sources of plant resistance to
Diaprepes abbreviatus (Coleoptera: Curculionidae)
by three bioassays. J. Econ. Entomol. 92: 999-1004.
ROUSH. 1981. 'Florigraze' rhizoma peanut: A peren-
nial forage legume. Univ. of Fla., Gainesville, Circu-
lar S-275, 22 pp.
RISCH, S. J. 1981. Insect herbivore avundance in tropi-
cal monocultures and polycultures. An experimental
test of two hypotheses. Ecology 61: 1325-1340.
SCHREIBER, U., AND W. BILGER 1987. Rapid assessment
of stress effects on plant leaves by chlorophyll fluo-
rescence measurements, pp. 27-53. In J. D. Ten-
hunen et al. [eds.], Plant response to stress, NATO
ASI Series, Vol. G15. Springer-Verlag, Berlin.
AND G. B. MARINI BETTOLO. 1990. Insect antifeedant
activity associated with compounds isolated from
species of Lonchocarpus and Tephrosia. J. Chem.
Ecol. 16: 365-380.
ADAIR 1996. Diaprepes abbreviatus (Coleoptera:
Curculionidae): host plant associations. Environ.
Entomol. 25: 333-349.
THOMAS, R. J. 1993. Rhizobium requirements, nitrogen
fixation, and nutrient cycling in forage Arachis, pp.
84-94. In P. C. Kerridge and B. Hardy [eds.], Biol-
ogy and agronomy of forage Arachis. Centro Inter-
nacional de Agricultura Tropical (CIAT), Cali,
VAN KOOTEN, O., AND J. F. H. SNEL. 1990. The use of
chlorophyll fluorescence nomenclature in plant
stress physiology. Photosynthesis Research 25: 147-

Florida Entomologist 86(1)


'El Colegio de la Frontera Sur (ECOSUR), Carretera Antiguo Aeropuerto km 2.5,
Tapachula, 30700 Chiapas, Mexico

2Imperial College at Silwood Park. Ascot Berkshire, SL5 7PY, UK

The establishment or colonization of natural
enemies is a critical period of adjustment for the
introduced individuals in the target area. Suc-
cessful colonization and its effectiveness in con-
trol, depend on the intrinsic capabilities of the
species and the interaction of physical and biotic
factors (Callan 1969). Although the concept of es-
tablishment is simple, in practice it is a difficult
task. From 4,769 introductions of predators and
parasitoids made up to 1990, only 1,445 (30.3%)
were established (Greathead & Greathead 1992).
Diverse reasons have been mentioned of the traits
likely to reduce the establishment of biological
control agents. Among the most important are:
adverse climatic condition, insufficient genetic
variation, poor searching capacity, interference
from chemicals or cultural practices, inadequate
numbers of individuals introduced, lack of syn-
chrony with the host and interference by native
organisms (Hopper et al. 1993, Hopper 1996).
Prorops nasuta Waterston is an African parasi-
toid of the coffee berry borer (CBB) Hypothenemus
hampei (Ferrari) (Coleoptera: Scolytidae), which
is considered as the main pest of coffee worldwide
(Le Pelley 1968, Baker 1999). P nasuta females
usually spend most of its life inside of a coffee
fruit infested by the CBB. The wasp feeds on all
juvenile stages, but only paralyses and oviposits
on the full-grown larvae and pupae. The larva
starts to feed externally after hatching and one
host is sufficient for the development of each
wasp. When the parasite larva is fully developed,
it spins a cocoon and enters to the pupal stage
(Hargreaves 1935; Infante 1998). The life cycle of
P nasuta form egg to adult lasts 28 days at 22C
(Infante 2000), but adults remain in the coffee
fruit for a few days more in order to copulate. As
with other bethylids, this species has males
emerging before their sisters with which they
mate. The new generation of wasps leave the
berry during the day, searching for infested coffee
fruits (Hargreaves 1935, Murphy & Moore 1990).
In the last few years this parasitoid has been
introduced to Mexico, Guatemala, El Salvador,
Honduras, Ecuador, Colombia, Jamaica, Indone-
sia and India. In all these countries P nasuta has
been released and is presently under evaluation
as a biological control agent (Klein-Koch et al.
1988, Barrera et al. 1990, Baker 1999). In the case

of Mexico, repeatedly releases of parasitoids have
been carried out in Chiapas since 1992. Recovery
surveys indicated that up to now, there is no evi-
dence of the establishment of P nasuta in the
country (Infante et al. 2001). Identifying the fac-
tors that interfere in the establishment of P na-
suta is essential to the success of biological control
programmes for the CBB. For that reason, the
present paper reports on some native arthropods
that were found predating on this species in coffee
plantations of Chiapas.
The wasps used in releases were reared in the
laboratory on borer infested coffee fruits. Parasi-
toids were taken to the field as adults and re-
leased from one liter plastic jars directly onto the
branches of coffee trees. From 1992 to 1996, ca.
156,000 individuals were released in the field
usually in the morning (before 12:00 hr). Coloni-
zation sites for releases consisted of 22 locations,
each approximately one-fourth of a hectare. Be-
cause few people observe insects under field con-
ditions (P. S. Baker, personal communication), we
stayed in the field for about 1-2 hours following
parasitoid liberations, to observe the wasps and
any possible interaction with other organisms.
The organisms detected as interacting with P na-
suta were collected and taken to the laboratory for
A list of arthropods preying on adults of P na-
suta is presented in Table 1. Meantime females
were searching for infested fruits, they were easy
prey for ants, which normally were present on cof-
fee trees. It was possible to detect at least five spe-
cies of ants preying on adults of P nasuta. Also six
species of spiders caught parasitoids on their
webs. As spiders were collected in alcohol, it was
not possible to verify their predation on P nasuta.
However, we would assume they do, considering
they are generalist predators. Recent studies
have reported that spiders, such as, Cyclosa car-
oli, Leucauge mariana and L. vetusta capture and
consume individuals of the CBB and its bethylid
parasitoid Cephalonomia stephanoderis (Henaut
et al. 2001).
Because of the small size of P. nasuta (2mm),
observations were especially difficult to carry out.
We could not measure the intensity in which pred-
ators were acting, but presumably they are only
important during releases, or maybe when the

March 2003

Scientific Notes


Species Family Location Observations

Pseudomyrmex sp. Formicidae El Encanto There were three species. Individuals of one
species foraged inside coffee fruits. Possible
predator of the CBB, as well.
Azteca sp. Formicidae S. Enrique Predator of adults only.
Tapinoma sp. Formicidae Laboratory Ants foraged inside fruits and predated on ju-
veniles of wasps and borers. Possibly the
wasp was dead before ants entered.
Uluborus nr campestratus Uloboridae La Gloria Adult parasitoids were found in the web.
Cyclosa caroli Araneidae S. Anita Adult parasitoids were found in the web.
Dolichognatha sp. Tetragnathidae R. Izapa Adult parasitoids were found in the web.
Leucauge sp. Tetragnathidae R. Izapa Adult parasitoids were found in the web.
Theridion nr nudum Theridiidae R. Izapa Some spiders make their web in the base of
the bunch of fruits.
Chrysso cambridgei Theridiidae Maravillas Adult parasitoids were found in the web.

new progeny of wasps leave the coffee fruits in
searching for new hosts. Arthropods preying on P.
nasuta were associated with old coffee trees and
abundant shade. It is possible there are other spe-
cies of ants and spiders predating on P. nasuta, as
there is a great diversity and abundance of these
organisms associated with coffee trees in Chiapas
(Ibarra-Nufiez 1990, Ibarra-Nufiez & Garcia-
Ballinas 1998). Callan (1969) emphasized the
jeopardy due to ants when colonizing natural en-
emies for classical biological control. Notwith-
standing, Le Pelley (1968) stated that this sort of
predation is incidental and only has a transitory
effect, since there is no continuing association be-
tween predator and prey. According to the poor re-
sults in the establishment of P nasuta in Mexico
(Infante et al. 2001), is possible that predation is
only a part of several factors that impede the es-
tablishment of P nasuta. On the other hand, re-
leasing parasitoids in the adult stage may not be
favorable for this species. Because of that, it
would be worth trying to release other biological
stage. For instance, parasitoids could be taken
from the laboratory to the field, in the pupal stage
while they are still inside the coffee fruits. Coffee
fruits could be placed inside a small cage hanging
on a branch of coffee with a thread covered with
grease to avoid ants and non-flying organisms. In
this way, the wasps would emerge when favorable
conditions exist. This sort of release might over-
come some problems with predators and it would
have the additional advantage of giving refuge
and shelter to the wasps.
We are grateful to G. Hernandez and J. L. Bar-
rera for technical assistance. Milimo Mebelo
made helpful comments on an earlier version of
this manuscript.

Prorops nasuta is an African parasitoid that
has been imported into Mexico for the biological

control of the coffee berry borer, Hypothenemus
hampei. After being released for several years, the
establishment of this parasitoid was never
achieved. In the present paper we report some na-
tive arthropods that were found interfering with
P nasuta in coffee plantations of Chiapas during
the release of parasitoids. Presumably predation
by ants and spiders on adults of P nasuta is only
a component of several factors that impede the es-
tablishment of P. nasuta in this country. Releas-
ing P nasuta in the pupal stage instead of adults
is discussed in order to enhance the potential of
this species as a biocontrol agent.


BAKER, P. S. 1999. La broca del caf6 en Colombia. In-
forme Final del proyecto MIP para el caf6 DFID-CE-
NICAFE -CABI. Chinchind, Colombia. 154 pp.
troducci6n de dos species de parasitoides africanos
a M6xico para el control biol6gico de la broca del caf-
eto Hypothenemus hampei (Ferrari) (Coleoptera: Sc-
olytidae). Folia Entomol. Mex. 79, 245-247.
CALLAN, E. MCC. 1969. Ecology and insect colonization
for biological control. Proc. Ecol. Soc. Australia 4, 17-
GREATHEAD, D. J., AND A. H., GREATHEAD. 1992. Biolog-
ical control of insect pests by insect parasitoids and
predators: the BIOCAT database. Biocontrol News
Inform. 13: 61N-68N.
HARGREAVES, H. 1935. Stephanoderes hampei, Ferr.,
coffee berry borer in Uganda. East African Agr. J. 1:
IAMS. 2001. Retention, capture and consumption of
experimental prey by orb-web weaving spiders in
coffee plantations of Southern Mexico. Ent. Exp.
Appl. 98, 1-8.
HOPPER, K. R. 1996. Making biological control introduc-
tions more effective. In: Biological control introduc-
tions: opportunities for improved crop protection.
Ed. By J. K. Waage. BCPC, London. 144 pp.

88 Florida Ento

Hopper, K. R., and R. T. Roush. 1993. Mate finding, dis-
persal, number released, and the success of biologi-
cal control introductions. Ecol. Entomol. 18, 321-
IBARRA-NUNEZ, G. 1990. Los artr6podos asociados a caf-
etos en un cafetal mixto del Soconusco, Chiapas,
Mexico. I. Variedad y abundancia. Folia Entomol.
Mex. 79, 207-231.
versidad de tres families de aranas tejedoras (Ara-
neae: Araneidae, Tetragnathidae, Theridiidae) en
cafetales del Soconusco, Chiapas, Mexico. Folia En-
tomol. Mex. 102, 11-20.
INFANTE, F. 1998. Biological control of Hypothenemus
hampei (Coleoptera: Scolytidae) in Mexico, using the
parasitoid Prorops nasuta (Hymenoptera: Bethyl-
idae). PhD Dissertation. Imperial College, Univer-
sity of London, U. K. 173 pp.


ologist 86(1) March 2003

INFANTE, F. 2000. Development and population growth
rates of Prorops nasuta (Hym., Bethylidae) at con-
stant temperatures. J. Appl. Entomol. 124: 343-348.
recovery of Prorops nasuta (Hymenoptera: Bethyl-
idae) an imported parasitoid of the coffee berry borer
(Coleoptera: Scolytidae) in Mexico. Southwest. Ento-
mol. 26, 159-163.
NEROS, AND D. DELGADO. 1988. Factores naturales de
regulaci6n y control biol6gico de la broca del caf6 (Hy-
pothenemus hampei Ferr.). Sanidad Vegetal 3, 5-30.
Le PELLEY, R. H. 1968. Pests of coffee. Longmans Green
and Co. London. 590 pp.
MURPHY, S. T., AND D. MOORE. 1990. Biological control
of the coffee berry borer, Hypothenemus hampei
(Ferrari) (Coleoptera: Scolytidae): previous pro-
grammes and possibilities for the future. Biocontrol
News and Information 11: 107-117.

Scientific Notes


'Department of Entomology, University of California, Riverside, CA 92521

2California Department of Food and Agriculture, Mt. Rubidoux Field Station
4500 Glenwood Drive, Riverside, CA 92501

The glassy-winged sharpshooter, Homalodisca
coagulata (Say) is the focus of a major classical bi-
ological control program in California. This insect
presents a serious threat to several agricultural
commodities and potentially native plants as well
because of its ability to vector the xylem-inhabit-
ing bacterium Xylella fastidiosa, the causative or-
ganism of "scorch like" diseases such as Pierce's
Disease of grapes and oleander leaf scorch, a seri-
ous malady of oleanders (Purcell & Saunders
1999). Homalodisca coagulata is an invasive pest
in California and its native range is the southeast-
ern and northeastern regions of the USA and Mex-
ico, respectively (Triapitsyn & Phillips 2000).
Homalodisca coagulata probably was translocated
to southern California as egg masses via the move-
ment of ornamental plants in the late 1980's (So-
rensen & Gill 1996) and without an accompanying
natural enemy fauna; inordinate populations of
glassy-winged sharpshooters have resulted.
During foreign exploration by MSH and SVT
for H. coagulata and associated egg parasitoids in
Florida in August 2001, the authors visited the
Florida State Collection of Arthropods, Bureau of
Entomology, Florida Department of Agriculture
and Consumer Services in Gainesville. Following
discussion with colleagues there, specimen re-
ceipt vouchers for H. coagulata were provided
that had been sent in for identification by lay peo-
ple, ornamental, horticultural, and agricultural
growers from around Florida. A total of 229 re-
ceipts were catalogued for adult H. coagulata over
the period 1958-2001 inclusive, and chits con-
tained information on date of collection, locality,
host plant, and sex of specimens. These data were
used to determine possible host plant records, dis-
tribution densities, and submission frequencies
for H. coagulata for different areas of Florida.
Homalodisca coagulata was collected from at
least 72 plant species in 71 genera contained in
37 families and Citrus spp. were the most com-
mon plants from which adult H. coagulata were
captured (Table 1). Of these plant association
records in Table 1 it is uncertain which can sup-
port development of H. coagulata from egg to
adulthood. Adult H. coagulata are vagile and
known to be highly polyphagous while the rela-
tively immobile immature stages have a nar-
rower host range (Turner & Pollard 1959). Citrus
may be over-represented in this dataset because
of regular pest surveys in this economically im-

portant crop. To determine if regional differences
in numbers ofH. coagulata specimens sent in for
identification existed, Florida was divided into
thirds: (1) top third was north of 290 Latitude; (2)
middle third was 27-29; and (3) the bottom third
was south of 29. Specimen receipts for each
county in each section of the state were assumed
to have been submitted for identification accord-
ing to a poisson distribution and proportions were
compared using a Log-likelihood Ratio Test (i.e.,
G-test). Pair-wise comparisons between regions
from which specimens were received were made
using x2 as sample sizes were large (Sokal & Rohlf
1995). The G-test was also used to determine if the
frequency with which samples were submitted
from each region significantly differed. Significant
differences in the number of specimens received
by region existed (X2 = 11.03; df= 2; P = 0.004). Sig-
nificantly more specimens were received for iden-
tification from north Florida, intermediate
numbers from central Florida, and fewest speci-
mens came from south Florida (Fig. 1). No signifi-

North Central South
Area of Florida

1OSpecimens ESubmissions

Fig. 1. Proportion (+95% CI) of specimens and sub-
missions of adult Homalodisca coagulata from north,
central, and south Florida to the Florida State Collec-
tion of Arthropods in Gainesville for identification over
the period 1958-2001. Bars with the same letters (upper
case [specimen comparisons] or lower case [submission
comparisons]) are not significantly different from each
other at the 0.05 level of significance.

Florida Entomologist 86(1)

March 2003


Plant family Plant species No. of collected H. coagulata specimens













Acer rubrum
Sansevieria sp.
Yucca aloifolia
Mangifera indica
Schinus terebinthifolis
Nerium oleander
Ilex spp.
Brassaia actinophylla
Eupatorium capillifolium
Helianthus annuus
Solidago altissima
Begonia sp.
Catalpa sp.
Spathodea campanulata
Ditremexa occidentalis
Casuarina spp.
Clusia sp.
Bucida buceras
Concarpus erectus
Ipomaea spp.
Cycas sp.
Aleurites fordii
Codiaeum variegatum
Ricinus comminis
Albizia julibrissin
Bauhinia punctata
Caesalpinia pulcherrima
Cercis sp.
Glycine max
Medicago sativa
Mimosa sp.
Parkinsonia aculeata
Pisum sp.
Psophcarpus tetragonolobous
Tetragonolobous sp.
Wisteria sp.
Quercus laevis
Q. virginiana
Pennisetum purpureum
Carya illinoensis
Juglans regina
Persea americana
Lagerstroemia indica
Magnolia grandiflora
Abelmoschus esculentus
Hibiscus rosa-sinensis
Swietenia mahagoni
Ficus benjamin
Callistemon viminalis
Eucalyptus spp.
Melaleuca quinquenervia
Myrtus communis
Psidium spp.
Mirabilis jalapa
Olea sp.
Hemionitis arifolia
Leucadendron sp.

Scientific Notes


Plant family Plant species No. of collected H. coagulata specimens

Rosaceae Malus spp. 2
Photinia sp. 1
Prunus persica 1
Pyracantha sp. 1
Pyrus sp. 1
Rosa sp. 1
Rubiaceae Gardenia sp. 1
Rutacaea Citrofortunella microcapra 1
Citrofortunella mitis 1
Citrus spp. 68
Fortunella sp. 3
Salicaceae Populus sp. 1
Salix spp. 2
Sapotaceae Manilkara roxburghiana 1
Vitaceae Ampelopsis arborea 1
Vitis spp. 4

cant differences (X2 = 3.23; df = 2; P = 0.20) in
frequency of submissions from each region were
observed (Fig. 1).
When taken together, these data suggest that
more H. coagulata were caught and submitted for
each identification event from North and Central
Florida but the rate of submission was similar
across the entire state. These data support MSH
and SVT's observations that H. coagulata is more
abundant and easier to collect in northern Florida
in comparison to central and southern Florida.
Possible constraints on the southern distribution
of H. coagulata could be related to temperature,
humidity, and rainfall dines or interspecific com-
petition with other proconiine sharpshooters (e.g.,
Oncometopia nigricans [Walker] [Hemiptera: Ci-
cadellidae: Cicadellinae: Proconiini]) that have
similar habitat requirements.
This work was supported in part by the Cali-
fornia Department of Food and Agriculture. We
thank Ruth Vega (UCR) for assistance with data
entry. Susan Halbert at the Florida State Collec-
tion of Arthropods, Bureau of Entomology, Florida
Department of Agriculture and Consumer Ser-
vices in Gainesville helped with locating H. coag-
ulata identification records.


Information from identification receipt vouch-
ers prepared by the Florida State Collection ofAr-
thropods, Bureau of Entomology, Florida

Department of Agriculture and Consumer Ser-
vices in Gainesville over the period 1958-2001 for
Homalodisca coagulata were analyzed for infor-
mation on host plants and distribution in Florida.
Homalodisca coagulata was recorded from at
least 72 plant species in 37 families and greater
numbers ofH. coagulata were sent in for identifi-
cation from northern Florida even though there
were no significant difference in specimen sub-
mission frequencies from north, central, and
south Florida.


PURCELL, A. H., AND S. R. SAUNDERS. 1999. Glassy-
winged sharpshooter expected to increase plant dis-
ease. California Agric. 53: 26-27.
SOKAL, R. R., AND F. J. ROHLF. 1995. Biometry: the prin-
ciples and practice of statistics in biological research.
Third Edition. W. H. Freeman and Company, New
SORENSEN, J. T., AND R. J. GILL. 1996. A range exten-
sion of Homalodisca coagulata (Say) (Hemiptera:
Clypeorrhyncha: Cicadellidae) to southern Califor-
nia. Pan Pacific Entomol. 72: 160-161.
TRIAPITSYN, S. V., AND P. A. PHILLIPS. 2000. First record
of Gonatocerus triguttatus (Hymenoptera: My-
maridae) from eggs of Homalodisca coagulata (Ho-
moptera: Cicadellidae) with notes on the
distribution of the host. Florida Entomol. 83: 200-
TURNER, W. F., AND H. N. POLLARD. 1959. Life histories
and behavior of five insect vectors of phony peach
disease. USDA Tech. Bull. 1988: 28.

Florida Entomologist 86(1)


USDA-ARS Invasive Plant Research Laboratory, Ft. Lauderdale, FL 33314

Salvinia molesta D. S. Mitchell, an invasive
floating fern, has invaded 12 states in the U.S. and
is now well established in Texas and Louisiana
(Jacono 1999). This plant quickly colonizes the
surface of slow moving, fresh water bodies caus-
ing severe ecological and economic problems
(Harley & Mitchell 1981). Successful biological
control programs targeting this weed have been
conducted in at least 13 countries worldwide us-
ing a small weevil, Cyrtobagous salviniae Calder
and Sands (Coleoptera: Curculionidae) (Julien &
Griffiths 1998). This species was first collected in
southeastern Brazil in 1979 (Forno & Bourne
1984) and, after extensive host range testing, was
released in Australia in 1980 where it reduced a
large infestation of S. molesta by more than 95%
after 13 months (Room et al. 1981). Since then C.
salviniae has been introduced into different coun-
tries with infestations of S. molesta where it suc-
cessfully reduced the weed to acceptable levels in
most cases (Julien & Griffiths 1998).
Although C. salviniae has been present in Flor-
ida since at least 1960 (Kissinger 1966) where it
reproduces on common salvinia, S. minima Baker,
attempts to transfer it to Texas and Louisiana
sites infested with S. molesta in 1999 met with
mixed results (Tipping, unpublished data). A com-
parison of gene sequence data between the Flor-
ida and Brazilian populations (ex Australia)
revealed some differences, the biological signifi-
cance of which remain under study (Goolsby et al.
2000). Consequently, further releases of the Flor-
ida population were discontinued in favor of the
Brazilian population that had been imported from
Australia. After additional host range testing con-
firmed its specificity to S. molesta, a general re-
lease permit was obtained for a designated area in
eastern Texas and western Louisiana.
The first releases were conducted during Octo-
ber 10-11, 2001 when a total of 880 C. salviniae
adults was released at four sites (220 per site). No
more releases were done until March 2002 in order
to determine if the insects could survive the winter.
Subsequent visits to these sites plus four control
sites were conducted in December 2001 and March
2002. During December 4-6, 2001, one hundred
plants from the mat of S. molesta confined within a
floating 1 m2 square of PVC pipe frame, were se-
lected without bias and hand-searched and the
number of adults found was recorded and left in
place. Adults were recovered from three of the four
sites with up to nine adults detected at one site
from the 100 plant sample. Hand-searching was
used because it is non-destructive. However, de-

structive methodologies like Berlese funnels usu-
ally extract 5-10 times more adults than hand
searching (Tipping, unpublished data) so the wee-
vil density may have been up to 90 adults in the
sampled area at one site.
Further sampling was halted during the win-
ter and resumed on March 25-27, 2002. Two of the
four release sites yielded adult C. salviniae during
hand searches with three and four adults found at
those sites. It is unknown if these weevils were
the original adults or progeny from the October,
2001 release. Sands et al. (1986) found that adults
held under constant temperatures of 23, 27, and
31C lived an average of 163.1, 116.9, and 101.5
days, respectively. These adults were found inside
the original release square but sampling within
paired 0.1 m2 quadrats, one along either side of a
transect line, 1, 5, and 10 meters away from the
release square, yielded no C. salviniae. It is note-
worthy, therefore, that adults of C. salviniae were
able to persist or propagate over a period of 166
days despite air temperatures that reached as low
as -9.1 C and 7.1C at Toledo Bend reservoir and
Lake Texana, respectively. The former site is the
most northern and the latter the most southern.
Temperature data were not available in the
immediate vicinity of these research sites so the
nearest recording stations were used (Table 1). In
addition, unlike air temperatures, water temper-
atures were only available on selected dates
which varied between the Toledo Bend and Lake
Texana sites. In the area around Toledo Bend res-
ervoir, there were 26 days during October 2001
through March 2002 when the minimum air tem-
peratures were below 0C. In contrast, only 7 days
had minimal air temperatures below 0C in the
Lake Texana area. These temperatures probably
were not maintained for long; in every case the
daily maximum temperature was above 0C.
Whiteman & Room (1991) found that Salvinia
molesta was killed when its buds were exposed to
temperatures less than -30C for 2-3 h. In many lo-
cations at these two sites, the plants were in coves
and backwater areas sheltered from direct winds
and protected by overhanging vegetation from the
shoreline or adjacent floating plants like water
hyacinth, Eichhornia crassipes (Mart.) Solms.
These conditions likely buffered any negative
temperature effects on the plants and insects. As
expected, water temperatures lagged behind the
colder air temperatures (Table 1).
This finding indicates that the Brazilian popula-
tion of C. salviniae can survive the winter climate
where S. molesta is extant in Texas and Louisiana.

March 2003

Scientific Notes


Toledo Bend Reservoir Lake Texana

Air Water Air Water
Temperature' Temperature2 Temperature3 Temperature4

Date Max Min Surface Date Max Min Surface

10/09/01 25.9 17.2 22.7 10/13/01 26.7 15.6 25.3
11/13/01 24.3 12.3 20.7 11/08/01 28.8 15.5 23.2
12/11/01 10.9 0.4 15.2 12/06/01 26.5 19.5 20.5
01/08/02 17.8 -2.3 9.5 01/15/02 18.8 3.4 12.9
02/12/02 15.2 -1.3 9.5 02/14/02 20.0 3.7 13.6
03/12/02 14.4 6.6 12.5 03/28/02 28.7 11.7 17.1

Texas A&M University Agricultural Research and Extension Center at Overton, TX.
Texas Sabine River Authority, station #15659.
'Texas A&M University Crop Weather Program Victoria County weather station.
'Lavaca-Navidad River Authority, site #13985.


Cyrtobagous saluiniae survived the winter of
2001-2002 in eastern Texas and western Louisi-
ana after its release in October 2001 on giant sal-
vinia, Salvinia molesta. Adults were recovered
from Toledo Bend reservoir and Lake Texana in
March 2002, up to 166 days after they were re-
leased. Although minimum air temperatures
were recorded at or below 0C on at least 26 days
at Toledo Bend reservoir, water temperatures
likely buffered these extreme conditions. Weather
conditions at Lake Texana, a more southern site,
were more benign.


FORNO, I. W., AND A. S. BOURNE. 1984. Studies in South
America of arthropods on the Salvinia auriculata
complex of floating ferns and their effects on S. mo-
lesta. Bull. ent. Res. 74: 609-621.
DRIVER 2000. Evidence of a new Cyrtobagous spe-

cies (Coleoptera: Curculionidae) on Salvinia minima
Baker in Florida. Southwest. Entomol. 25: 299-301.
HARLEY, K. L. S., AND D. S. MITCHELL. 1981. The biology
of Australian weeds. 6. Salvinia molesta D. S. Mitch-
ell. J. Aust. Inst. Agric. Sci. 47: 67-76.
JACONO, C. C. 1999. Salvinia molesta (Salviniaceae),
new to Texas and Louisiana. Sida 18: 927-928.
JULIEN, M. H., AND M. W. GRIFFITHS. 1998. Biological
Control of Weeds. A World Catalogue of Agents and
Their Target Weeds. Fourth Edition. CABI Publish-
ing, Wallingford.
KISSINGER, D. G. 1966. Cyrtobagous hustache, a genus
of weevils new to the United States fauna (Co-
leoptera: Cuculionidae: Bagoini). Coleopt. Bull. 20:
A. SANDS. 1981. Successful biological control of the
floating weed salvinia. Nature 294: 78-80.
comparative study on the intrinsic rates of increase
of Cyrtobagous singularis and C. salviniae on the
water weed Salvinia molesta. Entomol. exp. appl. 42:
WHITEMAN, J. B., AND P. M. ROOM. 1991. Temperatures
lethal to Salvinia molesta Mitchell. Aquatic Bot. 40:

Florida Entomologist 86(1)


1USDA, ARS Biological Control of Pests Research Unit, P.O. Box 67, Stoneville, Mississippi 38776

2USDA, ARS Biological Control and Mass Rearing Research Unit, P.O. Box 5367
Mississippi State, Mississippi 39762

Pitfall traps are commonly used to sample ants
and other ground-crawling arthropods. In some
cropping systems, pitfalls are the best means of
collecting arthropods such as red imported fire
ants (Solenopsis invicta Buren) that are active on
the soil surface (Kharboutli & Mack 1993). One
interesting application for pitfall traps may be to
collect ant species as biological indicators (Peck et
al. 1998).
Pitfall traps usually consist of a vial or similar
container buried up to the rim in the soil. A killing
agent (e.g., ethanol or propylene glycol) is placed
in the container to capture crawling insects for
study. Pitfall traps can yield species richness, spe-
cies composition, and relative abundance of forag-
ing ants (Bestelmeyer et al. 2000).
Some pitfall traps consist of a container in-
stalled permanently (or semi-permanently) in the
soil, into which the actual trapping container is
placed for easy removal. In larger studies involv-
ing hundreds of pitfall traps, however, this may
not be possible. For large-scale, area-wide studies
of black imported fire ant (Solenopsis richteri
Forel) and native ants in Mississippi, we use
small (2.54 cm I.D.) plastic vials. A cordless drill
with a 2.86 cm diameter auger bit is used to drill
a hole into the soil, and the vial is placed snugly
into the hole with the top flush to the soil surface.
This method presents 3 problems. First, vials can
be very difficult to remove when soil dries around
them. Secondly, soil is frequently brushed into the
vials during removal. Finally, stooping/kneeling is
necessary to remove the vials.
A device was constructed to address the prob-
lems listed above (Fig. 1). A shaft (61 cm long),
handle (14 cm long), and trigger (14 cm long) were
constructed of 1.27 cm diameter stainless steel
tubing. The lower handle was welded to the main
shaft, and the trigger was articulated on a short,
upright piece of tubing welded to the lower han-
dle. The upright tubing was split and flattened at
the top to accept the trigger. A hole was drilled
through the upright tubing and the trigger at the
point of articulation to accommodate a hex head
cap screw. A threaded rod (0.48 cm diameter) was
inserted into the main shaft and through a rubber
head constructed of four, 0.64 cm thick rings cut
from sheets of pure gum rubber. The upper end of
the threaded rod was secured to a nut welded to
the end of the trigger, and a second nut was used
to lock the threaded rod into position once it was


61 cm


Fig. 1. Schematic of pitfall retriever. When the
trigger (A) is depressed, the rubber head (B) expands
to grip the inside of the pitfall trap. The trigger has a
nut welded to one end (C), through which a threaded
rod is placed, and locked in position with a second nut
(D). An additional nut (E) secures the lower end of the
threaded rod to the rubber head. Figure is not drawn
to scale.

March 2003


Scientific Notes

properly adjusted. A washer (1.8 cm diameter) se-
cured with a nut to the bottom of the threaded rod
and another washer (3.4 cm diameter) welded to
the bottom of the shaft held the rubber head in
place. The upper washer should be larger than
the inner diameter of the pitfall trap to prevent
the device from being inserted too far, and the
lower washer must be smaller in diameter than
the rubber head.
When the rubber head of the device is inserted
into a pitfall trap, the trigger is squeezed, the
head is compressed vertically and expands later-
ally, and the trap is pulled free. The rubber head
expands and grips traps tightly without harming
them, preventing debris from entering the trap
during removal, and the 61 cm shaft reduces
stooping and effort. The trigger is squeezed until
the collector brings the trap up to grasp it, then
the trigger is released, and the trap can be
capped. Researchers interested in using this de-
vice for vials of different size could easily alter the
diameter of the rubber head to fit their needs.
Some adjustment may be necessary to get suffi-
cient expansion and grip without cracking plastic
vials. As configured, our device works well with 9-
dram, crystal plastic vials (Bioquip, Gardena, CA,
Other uses for this device might include place-
ment and retrieval of traps in hard to reach places
(e.g., below plant canopies, in crevices, etc.), pro-
vided a long drilling instrument could be used to
drill an appropriate hole for placing the traps.

The length of the device could be altered to suit
the needs of the researcher.
We thank John R. Davis for assisting with
testing of the prototype. R. L. Brown and D. D.
Hardee provided helpful reviews of a previous
version of the manuscript.


A device is described for rapid, easy removal of
pitfall traps embedded in the soil. The device pre-
vents debris from entering traps during removal
and reduces stooping and effort involved in pitfall
trap retrieval.


R. SILVESTRE. 2000. Field techniques for the study of
ground-dwelling ants. pp. 122-144. In D. Agosti, J. D.
Majer, L. E. Alonso, and T. R. Schultz (eds.), Ants:
Standard Methods for Measuring and Monitoring
Biodiversity. Smithsonian Institution Press, Wash-
ington, D. C.
KHARBOUTLI, M. S., AND T. P. MACK. 1993. Comparison
of three methods for sampling arthropod pests and
their natural enemies in peanut fields. Journal of
Economic Entomology 86: 1802-1810.
Using ant species (Hymenoptera: Formicidae) as a
biological indicator of agroecosystem condition. En-
vironmental Entomology 27: 1102-1110.

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