Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00032
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 2001
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: VID00032
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Full Text

McCoy et al.: IPM of Diaprepes abbreviatus on Citrus


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

The root weevil, Diaprepes abbreviatus, as a larva, inflicts feeding injury to the bark of all
root parts of a citrus tree, thereby impairing root function and supplying infection courts for
soil-borne root rot diseases. Ideally, larvae should be controlled at the soil surface before they
reach the root zone. In screenhouse and field experiments conducted in central Florida from
1996-99, the synthetic pyrethroid, bifenthrin, at 0.54 g/m2 (0.554 kg ai/ha) and RPA107382,
an analog of fipronil, at 0.156 and 0.312 ml/m2 (0.242-0.466 kg ai/ha), were applied uniformly
to the soil surface beneath the tree to form a chemical barrier against neonates ofD. abbre-
viatus. By comparison to the control, larval populations were reduced by 80-100% within one
week and these reductions persisted for 4-8 weeks. In an open screenhouse, bifenthrin gave
excellent root protection of container-grown trees during a 22 week period when neonates
were added to containers weekly for 12 weeks. RPA107382 was highly effective for about 2
weeks but lacked residual effect. The accumulation of leaf litter beneath the tree impaired
coverage of the soil by bifenthrin resulting in reduced control. According to weekly baited
trap counts, both chemicals reduced non-target foraging ants, particularly Solenopsis in-
victa Buren. The reduction in S. invicta was temporary however, but it did allow time for
other foraging ants to re-establish and increase.

Key Words: Chemical control, bifenthrin, citrus root weevil, ant predators

El picudo Diaprepes abbreuiatus, en etapa de larva, inflige dano al comer de la corteza de to-
das las parties de la raiz de un arbol citrico, asi perjudicando la funci6n de las races y su-
pliendo sitios de infecci6n para enfermedades por suelo de podredumbre radical. Idealmente,
las larvas deberian ser controladas en la superficie del suelo antes de alcanzar la zona radi-
cal. En invernaderos y experiments hechos en la Florida central de 1996-99, el piretroide
sint6tico bifenthrin, a 0.54 g/m2 (0.554 kg ai/ha) y RPA107382, un andlogo de fipronil, a 0.156
y 0.312 mrl/m2 (0.242-00.466 kg ai/ha), fueron uniformemente aplicados a la superficie del
suelo bajo el arbol para formas una barrera qufmica contra neonatos de D. abbreviatus. Po-
pulaciones larvales fueron reducidas de 80-100% y estar reducciones persistieron de 4-8 se-
manas de acuerdo a bioensayos de varias pruebas de campo. En un a casa abierta de tela
metalica, bifenthrin dio excelente protecci6n radical a arboles criados en potes durante un
period de 22 semanas donde neonatos fueron anadidos a los potes semanalmente por 12 se-
manas. RPA107382 fue altamente efectivo por alrededor de 2 semanas pero careci6 efecto re-
sidual. La acumulaci6n de material de hojas bajo el arbol perjudico la cobertura del suelo por
el quimico resultando en control reducido. De acuerdo con cuentas semanales de trampas ce-
badas, ambos qufmicos redujeron hormigas forrajeras no-objetivo, particularmente Solenop-
sis invicta Buren. Sin embargo, la reducci6n de S. invicta fue temporera, pero si permiti6
tiempo para que otras hormigas forrajeras se reestablecieran e incrementaran. Este compor-
tamiento sugiere que control de S. invicta puede ser beneficioso.

Diaprepes abbreviatus L., a root weevil native
to the Caribbean islands (O'Brien & Wibmer 1984,
Woodruff 1985), has gradually become a major lo-
calized pest of citrus, many ornamental plants,
and some agronomic crops since its introduction
into Florida in 1964 (McCoy 1999). It can be a uni-
voltine species on citrus, however, the life cycle
can vary greatly in time. The adult, egg, and neo-
nate stage appear on the host plant above ground,
and all larval stages, the pupa and general adult
occur below ground. At hatch, neonates fall from
the leaf to the soil surface beneath the tree where

they enter the soil (Wolcott 1936). Numerous in-
stars feed on fibrous and woody roots, forming
deep grooves in the latter as they consume the
outer bark, including the cambium layer. Larvae
remain on the roots for 8-15 months, reaching 1.3-
2.5 cm in length (Wolcott 1936, Quintela et al.
1998). Injuries caused by D. abbreviatus larvae
serve as preferred infection courts for root rot dis-
eases of citrus caused by soil-borne fungal patho-
gens such as Phytophthora spp. (Graham et al.
1996). The interaction between root weevils and
soil-borne fungal pathogens of the roots results in

Florida Entomologist 84(3)

one of the most severe decline syndromes affecting
citrus. Therefore, pest management of D. abbre-
viatus can require treatment of both the insect
and soil-borne diseases.
Prior to the cancellation of the chlorinated hy-
drocarbon insecticides in the U.S.A. around 1980,
citrus root weevils were controlled with persis-
tent compounds, such as aldrin and dieldrin (Bul-
lock 1985). These chemicals were generally
broadcast on the soil beneath the tree as granules
in the dry fertilizer mix. By forming a chemical
barrier beneath the tree, invasive neonates were
killed before reaching the root system (Bullock
1985). Subsequently, organophosphate and car-
bamate substitutes for the chlorinated hydrocar-
bons were found to be less effective because of
their shorter residual.
In the past decade, entomopathogenic nema-
todes, infectious to all soil-inhabiting stages of
the weevils, have been applied as biopesticides
one or more times per year in Florida citrus (Mc-
Coy et al. 2000). Since larval control with nema-
todes varies in the field and likely occurs in the
rhizosphere after larvae are already feeding on
the roots (Duncan et al. 1999), various chemical
and non-chemical agents for combating neonates
at the soil surface have been under investigation.
Greenhouse and screenhouse studies using con-
tainer-grown citrus plants as indicators of neona-
tal feeding byD. abbreviatus have been conducted
with systemic and contact pesticides of various
formulations. Chlorpyrifos as a slow release gran-
ule, imidacloprid and bifenthrin have been shown
to be effective against neonates as soil drench or
soil-incorporated treatments (McCoy et al. 1995).
Bifenthrin is currently recommended for use in
citrus nurseries to prevent larval invasion of con-
tainer-grown plants (Simpson & McCoy 1996).
Bifenthrin is a pyrethroid with a broad-spec-
trum of activity against insects and mites. In view
of its toxicity to neonate root weevils and persis-
tence on the soil surface, screenhouse and field
studies were conducted during the past 4 yr with
two formulations, Brigade 10WSB 100% and
Capture 2L, to determine residual control of ne-
onate D. abbreviatus. Since bifenthrin is toxic to
the imported fire ant, Solenopsis invicta Buren,
an important predator of D. abbreviatus found in
citrus groves, ant populations were monitored us-
ing baited traps to assess non-target effects.


Screenhouse Experiment

Sixty-three, 3-yr-old Parson Brown orange trees,
grafted to Cleopatra mandarin rootstock, were
cleared of soil and pruned slightly for transplant
into 56.8 liter plastic containers with a 0.16 m2
surface area. Each tree was planted in sieved Can-
dler soil (Entisol type: 92% sand, 2.9% clay, 2.0%

silt) and placed on a bench in open sunlight. One
month after planting, trees were fertilized using
liquid 8:4:8 NPK at 60 ml per tree. Trees were wa-
tered only when rainfall failed to supply adequate
moisture to prevent wilt. Any weeds were periodi-
cally removed by hand. Leaf litter, collected from a
commercial grove, was scattered on the soil sur-
face to a depth of 1.27 cm of nine trees to compare
treatments with and without leaf litter.
Each treatment, Brigade 10WSB at 0.134,
0.269, and 0.54 g/m2 (high rate equivalent to field
rate) and Admire 2L at 0.54 g/m2, were applied
uniformly to the soil beneath the tree in 50 ml of
H20 with a B&G hand held sprayer. Two un-
treated controls were included in the experiment,
one with larval infestation and one without. The
treatment with leaf litter received Brigade
10WSB at 0.54 g/m2. The soil surface of all con-
tainerized trees was moistened just prior to treat-
ment and 100 ml of water per tree added
immediately after application using a sprinkling
can. Each treatment was replicated 9 times. All
treatments were applied on May 8, 1996.
At 2 days post-treatment and weekly there-
after, for 12 weeks, 25 neonatal D. abbreviatus
(48-h-old) were scattered on a moist soil surface of
each container. Each container received 325 lar-
vae over time. Neonates used in the study were
obtained from field-collected adult females held in
screened cages in the greenhouse at 27 + 2C.
On October 15, about 22 weeks after treat-
ment, each tree was carefully removed from the
container and the soil around the roots washed
through a sieve to recover larvae. Soil remaining
in the container was also wet-sieved to assure lar-
val recovery. After recording larval number, the
root symptoms of each tree were rated visually on
a scale from 1 to 5; 1 = no visible injury, 2 = nor-
mal fibrous root density, slight tap/lateral root
channeling, 3 = moderate fibrous root density and
tap/lateral root channeling, 4 = severe fibrous root
loss and tap/lateral root channeling, and 5 = no
fibrous roots and severe tap/lateral root channel-
ing and stem girdling. In addition, fibrous roots
were removed from each tree, dried, and weighed
to determine dry root weight. Differences in larval
survival/treatment and dry fibrous root weight
among different chemical treatments were com-
pared by ANOVA and Tukey's Studentized Range
(HSD) test (SAS Institute 1988). Root rates were
presented as treatment averages.

Field Test-1

The experiment was conducted in an irrigated
3-yr-old Flame grapefruit grove grafted to 'Swin-
gle' citrumelo rootstock. The grove was located
near Alturas, FL on Astatula fine sandy soil. The
tree spacing was 3.7 x 6.0 m. Any leaf litter was
removed prior to chemical applications with a
hand held blower.

September 2001

McCoy et al.: IPM of Diaprepes abbreviatus on Citrus

Brigade 10WSB was applied at 0.269 and 0.54
g/m2 (0.276-0.554 kg ai/ha) to a 1.5 m2 area beneath
the tree at 500 ml of finished spray mix at 20 psi
using a low pressure sprayer and a hand-held
spray wand. Each treatment was comprised of 4-
tree plots replicated 10 times and an untreated
check. Treatments were applied on April 12, 1996.
The grove was irrigated thoroughly before appli-
cation and for about 1 h after application.
At 0, 31, 59, 95, and 123 days post-treatment,
a laboratory bioassay was performed to deter-
mine the residual activity of the insecticide in the
soil over time. One soil sample of 1.5 ml by volume
per tree, 10 per treatment, was taken randomly
from the soil surface to a depth of 3.2 mm. Each
soil sample was assayed separately by placing
each sample in a micro-centrifuge tube (1.5 ml)
with 10 neonates. After a 7-day exposure at 27 +
2C, larval survival was recorded for each tube.
Larval survival representative of each post-treat-
ment bioassay of treated and untreated field soil
were made, after the proportions surviving were
transformed using arcsine JX, using a one way
ANOVA design followed by Tukey's Studentized
Range (HSD) test (SAS Institute 1988). Untrans-
formed means are shown in all figures.

Field Test-2

The experiment was conducted in a 5-yr-old
planting of Hamlin orange grafted to 'Swingle' cit-
rumelo rootstock planted in Candler soil, and lo-
cated on the University of Florida campus, Lake
Alfred, FL. The grove was set at a 3.7 x 6.0 m
spacing and had micro-sprinkler irrigation. Prior
to treatment, leaf litter was removed from the soil
beneath the trees. In this test, the residual con-
trol of Brigade 10SWS at 0.54 g/m2 was compared
to an untreated check. Plots consisted of 6 trees in
one row. Each plot was completely randomized
and replicated 4 times. Brigade was applied as a
band, 1.8 m in length, to the soil beneath the tree
at 369.3 liters/ha using a tractor-mounted herbi-
cide applicator set at 0.2 mpa and traveling at 4.4
kmph. Treatment was applied on September 17,
1998 under clear sky and an air temperature of
26C. Irrigation was applied for 3 h before appli-
cation to thoroughly wet the soil and immediately
after application for 1 h.
In this test, the bioassay method used to mea-
sure residual effect was changed to reflect more
typical field conditions. Soil cores were collected
randomly from beneath the tree to a depth of 2.54
cm midway between the trunk and dripline using
a cork borer with a diameter of 1.27 cm2. Soil sam-
ples were taken weekly, 5 samples per plot or a to-
tal of 20 samples per treatment. In the field, each
soil core was carefully placed intact, into a plastic
column with a screen base (20 mm mesh) just
large enough for the passage of larvae through the
soil into a well of a plastic tissue culture plate

(McCoy et al. 2000). In the laboratory, 10 vigorous
neonates (<48-h-old) were placed on the soil sur-
face within the column. After 72 h at 28C, the
number of larvae capable of moving through the
soil column into the well of the tissue culture plate
was recorded. The number of live, dead, and miss-
ing from the original inoculum also was recorded
by sorting through the soil. Statistical analysis
was performed as described in the previous test.

Field Test-3

The experiment was conducted in a one-yr-old
reset planting of Navel orange on 'Swingle' cit-
rumelo rootstock planted in Candler soil and lo-
cated on the University of Florida campus, Lake
Alfred, FL. The grove was close set at a 0.9 x 6.1
m spacing for research purposes and had micro-
sprinkler irrigation. Treatments were arranged
in a completely randomized block design, each
plot consisting of 6 trees/plot (4.6 x 0.5 m). No leaf
litter was found on the soil in the experimental
site. Treatments consisted of Brigade 10SWS at
0.54 g/m2 (0.56 kg ai/ha), RPA107382 (0.38 kg EC)
at 0.156 and 0.312 ml/m2 (0.10 kg and 0.19 kg ai/
ha) and an untreated check. RPA107382 is an an-
alog of fipronil, a phenyl pyrazole with toxicity to
neonatal D. abbreviatus (Nigg et al. 1999). Each
treatment was replicated 10 times. Chemicals
were applied on April 13, 1999 with a backpack
CO2 activated sprayer equipped with a hand held
boom with dual fan nozzles. Materials were
mixed in 2 liter quantities and applied at 387.2
LPH at 0.2 mpa. At the time of application, air
temperature was 18-20C and relative humidity
was 50%. Wind conditions were calm.
At various times, from 0 to 56 days post-treat-
ment, the previously described column bioassay
method was used to measure chemical residual on
the soil surface. Two soil cores per plot were taken
within 15.2 cm of the tree trunk to a depth of 2.54
cm. Bioassay data were analyzed using the proce-
dures described for tests 1 and 2.
Since bifenthrin and fipronil are toxic to S. in-
victa by contact or in bait form (Collins & Callcott
1998, Knapp 2000) and S. invicta is a general
predator ofD. abbreviatus (Whitcomb et al. 1982),
ant populations were monitored via baited traps
pre- and post-treatment to assess non-target ef-
fects.About 1.5 g of hamburger was placed on a fil-
ter paper strip that was, in turn, placed within a
40 mm plastic assay disk (Millipore) with a small
hole on the side to allow for ant entry. A single
trap was placed on the soil surface near each tree
trunk in each plot. After exposure for 90 min, the
trap was closed confining the ants within the disk.
In the laboratory, the number and species of ants
were recorded. Samples of ants were collected at 1
week pre-treatment and at 7, 14, 21, 29, 35, and
41 days post-treatment. Differences in the total
number of ants between treatments was analyzed

Florida Entomologist 84(3)


Mean number surviving Mean fibrous roots, Mean rootb
Treatment Rate (g/m2) larvae/unit + SD g dry wt/tree + SD rating (1-5)

Control (no larvae) -- 1133.9 + 184 a 1.0
Control (larvae) 15.1 9.0 bc 589.7 + 102 c 4.2
Brigade 10WSB 0.143 8.3 + 6.7 ab 816.5 + 197 c 2.5
Brigade 0.269 22.8 + 9.4 c 861.8 + 391 bc 2.7
Brigade (no litter) 0.54 8.8 + 4.1 ab 1179.3 + 304 a 1.7
Bridge (litter) 0.54 23.8 + 11.4 c 771.1 164 c 3.3
Admire 2F (no litter) 0.54 6.0 + 3.6 ab 997.9 + 201 ab 1.3

'Means followed by the same letter are not significantly different at the 5% level of probability using Tukey's Studentized Range (HSD) test.
'Root symptoms: 1-no visible injury, 2-slight tap/lateral root channeling, 3-moderate tap/lateral root channeling, 4-severe tap/lateral root channeling,
5-severe tap/lateral root channeling and stem girdling.

statistically for each sample date via ANOVA and
Tukey's Studentized Range (HSD) test.


Screenhouse Experiment

As expected, the control with neonates had
greater root injury than the control with no larvae
(Table 1). Fibrous root loss in the larval control

--I Brigade 0.54 g/m2
LZZI- Brigade 0.27 g/m2
100_ R Untreated Check
U a


2 40- b

a 20-

was nearly twice that of the plant only control
and overall root injury was severe. Only 5% of the
original larvae introduced were recovered at the
end of the test.
The presence of leaf litter on the soil surface of
the containerized citrus trees reduced larval con-
trol by Brigade suggesting that the soil barrier
was distorted by litter. As shown in Table 1, Bri-
gade at 0.54 g/m2, without leaf litter, resulted in
higher larval mortality and root protection com-

Sample Date

Fig. 1. Residual effect of two rates of Brigade 10SWS on neonate survival of Diaprepes abbreviatus based on sur-
face soil sampling at Alturas, FL-"bars with common letters on the same sampling date are not significantly dif-
ferent" at the P = 0.05 level.

September 2001

McCoy et al.: IPM of Diaprepes abbreviatus on Citrus

pared to the treatment with leaf litter; in fact,
there was no significant difference in larval sur-
vival or root protection between Brigade with lit-
ter, the control without larvae, and the Admire
standard. As the rate of Brigade decreased, root
protection also decreased (Table 1). In view of its
systemic action, Admire as a standard without lit-
ter performed as well as Brigade without litter
and the control without larvae.

Field Test-1

As shown in Fig. 1, the residual control of Bri-
gade against neonate D. abbreviatus, based on
surface soil bioassays, was significantly longer
(>59 days) at the highest rate (0.54 g/m2) than at
the lower rate (0.269 g/m2); but, the lower rate
was significantly different from the untreated
check (P = 0.05). At 95 days post-treatment or
longer, all treatments were the same and pro-
vided no control.

Field Test-2

In the fall study, the previously described soil
column bioassay method, more typical of what in-
vasive larvae would experience in the field, was


used to monitor residual control of the chemical
barrier to neonate Diaprepes. As shown in Fig. 2,
Brigade at 0.54 g/m2 gave 100% kill of neonates in
a bioassay performed within 72 h after treatment.
At 1 week post-treatment, larval survival increased
to about 20% but remained significantly lower than
the untreated check. Residual control varied from
40-58% compared to the untreated check during
the 8-10 weeks after treatment. Significant differ-
ence between treatments was measured via bio-
assay for about 3 months; however, larval survival
approached 80% by that time (Fig. 2).

Field Test-3

In the spring study, using the soil column bio-
assay, both Brigade at 0.54 g/m2 and the 2 rates of
RPA107382 at 0.156 and 0.312 ml/m2 significantly
reduced larval survival compared to the control
at day 1 post-treatment (Fig. 3). In the case of
Brigade, residual effect resulted in less than 30%
larval survival throughout the study (8 weeks),
whereas, the residual effect of RPA107382 began
to fail at 4 weeks post-treatment and was less
effective than Brigade thereafter, but signifi-
cantly better than the control through 6 weeks

9/17 9/22 9/24 10/1 10/8 10/15 10/22 10/29 11/9 11/15 11/29

Sample Day

Fig. 2. Residual effect of Brigade 10SWS applied in the fall on neonate survival of Diaprepes abbreviatus based
on soil column bioassay at Lake Alfred, FL-"bars with common letters on the same sampling date are not signifi-
cantly different" at the P = 0.05 level.

Florida Entomologist 84(3)

0 I L Y9C I alx I v i AM-Ld- I KO I ILM4I IS3-pl r J 4p i I i|
4/13 4/21 4/28 5/5 5/11 5/19 5/26 6/2 6/9

Sample Day
Fig. 3. Residual effect of Brigade 10SWS and two rates of RPA107382 on neonate survival of Diaprepes abbre-
viatus based on soil column bioassay at Lake Alfred, FL-"bars with common letters on the same sampling date are
not significantly different" at the P = 0.05 level.

The effect of the different chemicals applied to
the soil on non-target foraging ant populations
over time is depicted in Fig. 4 and Table 2. In the
control, Solenopsis invicta Buren was always most
frequently trapped using hamburger bait followed
by Dorymyrmex bureni (Trager), D. reginicula
(Trager), Brachymyrmex obscurior Forel, and Tet-
ramorium simillimum (F. Smith) in descending
order of abundance. As shown in Fig. 4, both Bri-
gade and RPA107382 suppressed S. invicta, at a
very low level, through 29 days post-treatment.
Thereafter, S. invicta gradually increased in abun-
dance. Other foraging ants mentioned above were
not eliminated and after 29 days both S. invicta
and other ants increased together at 35 and 41
days post-treatment (Fig. 4). As shown in Table 2,
D. bureni appeared to become the prevalent forag-
ing ant following the chemical treatments.


Screenhouse and field data collected at differ-
ent times of the year for 4 yr using two bioassay
methods all support the efficacy ofbifenthrin (Bri-
gade/Capture) as a chemical barrier against neo-
nate D. abbreviatus, if leaf litter does not interfere
with coverage of the soil. Since bifenthrin has the
propensity to bind strongly to soil particles after

drying, uniform application to a moistened sub-
strate free of debris appears vital for maximum
contact with the invasive larvae. Since young
trees (<5 yr old) do not accumulate leaf litter in
large quantity, a chemical barrier should be effec-
tive. Timmer et al. (2001) showed that leaf litter
accumulation beneath mature trees is greatest
from January through June. Unfortunately, this
coincides with the spring adult emergence of
D. abbreviatus. Removal of leaf litter by air blast
using a speed sprayer appears to be a feasible way
to redistribute the litter away from the tree.
According to bioassays, the most effective rate
of bifenthrin to reduce neonate populations enter-
ing the soil appears to be 0.54 g/m2 (0.554 kg ai/ha).
A rate of 0.269 g/m2 will reduce neonates within 7
days after soil application, however, residual effect
is lesser in the field. Effect of bifenthrin on other
instars and general adults is unknown. Obviously,
the adult stage would be exposed to the chemical
barrier at the time of emergence.
Although bifenthrin and RPA107382 reduced
S. invicta populations following soil application,
this reduction appeared to favor the re-establish-
ment of more diversity at baits in the ant fauna
among the treatments (Fig. 4). Data suggest that
this faunal shift at baits occurred following a sig-
nificant reduction of S. invicta. In view of its dom-

September 2001

McCoy et al.: IPM of Diaprepes abbreviatus on Citrus

120 = control Solenopsis invicta a
M Brigade (0.54 g/m2 )
E RPA 107382 (0.156 ml/m2) a a
100 E= RPA 107382 (0.312 ml/m2)

80 a
a a

60 a a

40 a a a

L.I b a a

C 140 -
Uo Other Species

+ 120

100 -

80 -
60 a a
a | aa
40 a 21 1?-a | a
20 b a
a a
8 0 a a s a

-1 7 14 21 29 35 41

Days (pre + post-treatment)

Fig. 4. Mean number of Solenopsis invicta and other foraging species trapped over time on the soil surface via
baited traps following the field application of Brigade and RPA107382 for root weevil larval control, Lake Alfred,
FL. Application made at zero days-"bars with common letters on the same sampling date are not significantly dif-
ferent" at the P = 0.05 level.

finance in disturbed habitats such as citrus groves re-establishment of a more diverse ant commu-
(Tschinkel 1988), its reduction could have caused nity improves predation on neonates. These data
the shift and therefore, might be beneficial, if the and other unpublished work show that S. invicta

Florida Entomologist 84(3)

September 2001


Treatment Species %

Control Solenopsis invicta 74.8
Dorymyrmex bureni 10.2
Dorymyrmex reginicula 10.3
Brachymyrmex obscurior 0.3
Tetramorium simillimum 4.4
Brigade 10SWS (0.54 g/m2) Solenopsis invicta 18.1
Dorymyrmex bureni 69.7
Brachymyrmex obscurior 3.8
Tetramorium bicarinatium 8.4
RPA 107382 (0.156 ml/m2) Solenopsis invicta 48.8
Dorymyrmex bureni 43.8
Brachymyrmex obscurior 2.7
Tetramorium simillimum 4.7
RPA 107382 (312 ml/m2) Solenopsis invicta 53.6
Dorymyrmex bureni 41.4
Brachymyrmex obscurior 1.1
Tetramorium bicarinatium 3.9

will recover however, and re-establish its domi-
nance at baits within a few weeks, suggesting
that periodic treatment for S. invicta could be a
positive management practice. Further research
is appropriate.
Hamburger was a successful bait for capturing
S. invicta and other foraging ants. It should be
pointed out however, that some selection of ant
species likely occurred because of a preference for
meat and aggressive dominance at baits by par-
ticular species.
Seasonal population dynamics of adult Di-
aprepes based on trapping data suggest that peak
emergence occurs in April through mid-June in ir-
rigated citrus groves in central Florida with some
emergence throughout the year (Stansly et al.
1997, McCoy & Duncan 2000). Since adults are
most abundant at this time of the year, one can
assume that oviposition and subsequent neonatal
invasion of the soil is also at a high level, particu-
larly in view of the fact that ovipositing adults ap-
parently live 3-4 months in the field and food is
abundant. According to this biological informa-
tion, it would appear that the use of bifenthrin
should begin in late April to early May to maxi-
mize its residual effect on invasive larvae. Stud-
ies are currently underway to evaluate this
strategy over time.


This research was partially supported by FMC Cor-
poration, Aventis Corporation and the Florida Citrus
Production Research Advisory Council. Florida Agricul-
ture Experiment Station Journal Series No. R-07834.


BULLOCK, R. C. 1985. Potential for controlling citrus
root weevil larvae and adults with chemicals. Flor-
ida Entomol. 68(3): 417-423.
COLLINS, H. L., AND A. M. A. CALLCOTT. 1998. Fipronil:
An ultra-low-dose bait toxicant for control of red im-
ported fire ants (Hymenoptera: Formicidae). Florida
Entomol. 81(3): 407-415.
GRAHAM. 1999. Entomopathogenic nematodes as a
component of citrus root weevil IPM, pp. 69-78 in
S. Polavarapu [ed.], Optimal use of insecticidal nem-
atodes in pest management. Rutgers University,
New Brunswick, NJ.
Insect-plant pathogen interactions: Preliminary stud-
ies of Diaprepes root weevil injury and Phytophthora
infections. Proc. Florida State Hort. Soc. 109: 57-62.
KNAPP, J. L. 2000. Florida Citrus Pest Management
Guide. Coop. Extension Service-IFAS, SP-43, Gaines-
ville, FL.
McCoY, C. W. 1999. Arthropod pests of citrus roots, pp.
149-156 in L. W. Timmer, and L. W. Duncan [eds.],
Citrus Health Management. APS Press, St. Paul, MN.
J. FOJTIK. 1995. Effect of surface-applied and soil-in-
corporated insecticides for the control of neonate lar-
vae of Diaprepes abbreviatus in container-grown
citrus. Proc. Florida State Hort. Soc. 108: 130-136.
McCoY, C. W., AND L. W. DUNCAN. 2000. IPM: An
emerging strategy for Diaprepes in Florida citrus,
pp. 90-104 in S. H. Futch [ed.], Diaprepes Short
Course. Coop. Extension Service, Florida Expt. Sta.
McCOY, C. W., D. I. SHAPIRO, AND L. W. DUNCAN. 2000.
Application and evaluation of entomopathogens for
citrus pest control, Chapter VII-II, pp. 577-595 in L. A.
Lacey and H. K. Kaya [eds.], Field Manual of Tech-
niques in Invertebrate Pathology. Kluwer Acad. Pubs.

McCoy et al.: IPM of Diaprepes abbreviatus on Citrus

K. NGUYEN. 2000. Entomopathogenic nematodes
and other natural enemies as mortality factors for
larvae ofDiaprepes abbreviatus (Coleoptera: Curcu-
lionidae). Biological Control 19: 182-190.
LIN, AND N. W. CUYLER. 1999. Fipronil for Diaprepes
abbreviatus (Coleoptera: Curculionidae) larval con-
trol in container-grown citrus. Proc. Florida State
Hort. Soc. 112: 77-79.
O'BRIEN, C. W., AND G. J. WIBMER. 1984. Annotated
checklist of the weevils (Curculionidae sensu lato)
of North America, Central America, and the West
Indies-Supplement 1. Southwestern Entomol. 9(3):
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(5): 1173-1179.
SAS Institute. 1988. SAS user's guide, version 6.03. SAS
Institute, Cary, NC.
SIMPSON, S. E., AND C. W. McCOY. 1996. Control of
Diaprepes root weevil with bifenthrin and other
pesticides. Proc. 1996 Ann. Japanese beetle review.
McMinnville, TN, January 24-25, 1996. 5 pp.

STANSLY, P. A., R. F. MIZELL, AND C. W. McCoY. 1997.
Monitoring Diaprepes abbreviatus (Coleoptera: Cur-
culionidae) with Tedders traps in Southwest Florida
citrus. Proc. Florida State Hort. Soc. 110: 22-26.
IBANEZ. 2001. Epidemiology and control of citrus
greasy spot in different citrus-growing areas in Flor-
ida. Plant Disease 84: (in press).
TSCHINKEL, W. R. 1988. Distribution of the fire ants So-
lenopsis invicta and S. geminata (Hymenoptera: For-
micidae) in northern Florida in relation to habitat
and disturbance. Ann. Entomol. Soc. Am. 81(1): 76-81.
1982. Predators of Diaprepes abbreviatus larvae.
Florida Entomol. 65(1): 150-158.
WOLCOTT, G. N. 1936. The life history of Diaprepes ab-
breviatus at Rio Piedros, P.R. J. Agric., University of
Puerto Rico 20(4): 883-914.
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(3): 370-379.

Florida Entomologist 84(3)

September 2001


Department of Neurobiology and Behavior, Cornell University, Seeley G. Mudd Hall
Ithaca NY 14853-2702, USA

'Present address: Department of Biological Sciences, Western Illinois University
1 University Circle, Macomb IL 61455, USA


Optical movement detectors are often used in laboratory studies of insect behavior. They of-
fer advantages of time resolution and ease of analysis compared with video. However, design
and construction have rarely been described in enough detail to allow the devices to be built
easily by others. We describe a simple optoelectronic system for measuring rapid movements
in one dimension, such as the protraction of an insect leg. The leg casts a bar-shaped shadow
onto a photodiode chip that is masked to expose a triangular area. Movement of the leg
changes the total area of the triangle that is shaded. A preamplifier converts the change in
photoelectric current to a voltage signal. The preamplifier includes an optional circuit for re-
moving 120 Hz ripple resulting from AC-powered light sources by subtracting the output of
a second, reference photodiode. We have used the system to quantify leg movements in an
acoustic startle response of a field cricket (Teleogryllus oceanicus LeGuillou). This system
could be adapted for a wide range of other applications in laboratory and field research.

Key Words: leg motion, position detector, photodiode, optoelectronic photodetector, cricket
acoustic startle response


Detectores 6pticos de movimiento han sido usados frecuentemente en studios de comporta-
miento de insects en el laboratorio, donde ofrecen ventajas de resoluci6n de tiempo y facili-
dad de andlisis comparado con video. Sin embargo, su diseno y construcci6n han sido
raramente descritos en suficiente detalle para permitir que otros puedan construir estos
aparatos facilmente. Describimos un sistema optoelectr6nico para medir movimientos rdpi-
dos en una dimension como la protracci6n de una pata de insecto. La pata forma una sombra
en forma de barra a un chip fotodiodo que esta cubierto para exponer una area triangular.
El movimiento de la pata cambia el area total del triangulo que esta sombreado. Un pream-
plificador convierte el cambio en corriente fotoel6ctrica a una serial de voltaje. El preampli-
ficador incluye un circuit para nulificar ondulaci6n de 120 Hz de fuentes de luz con
electricidad AC al restar la producci6n de fotodiodo de referencia. Hemos usado el sistema
para cuantificar movimientos de pata en una respuesta de susto acustica del saltamontes
Telegryllus oceanicu (LeGuilou). Elementos de este sistema pueden tambien ser adaptados
para otras aplicaciones en studios de laboratorio y de campo.

Optoelectronic position detectors may be used
to quantify behavior in a wide variety of settings.
Optoelectronic methods have two advantages
over video: (1) Temporal resolution is not limited
by video frame rate, and (2) The detector can mea-
sure a single parameter of movement or position
that would require extensive labor or computer
processing to extract from video records. Descrip-
tions of optoelectronic devices in the biological lit-
erature generally emphasize the unique features
of a particular detector without providing details
of design and construction. Potential users may be
discouraged from applying these methods if they
lack the practical electronics background to adapt
the circuits found in electronics cookbooks or tech-
nical literature from chip manufacturers. We de-

scribe a photodetector and amplifier designed for
monitoring leg protraction in a tethered flying
Polynesian field cricket (Teleogryllus oceanicus
LeGuillou), and readily adaptable to other uses as
described in Results and Discussion.
We are studying an ultrasound-induced escape
response in crickets, a flight turn that is a defense
against echolocating bats (Moiseff et al. 1978). A
cricket is flown on a tether and given pulses of ul-
trasound from a loudspeaker mounted to its left
or right. We monitor one component of the startle
response, a lateral outward swing of the metatho-
racic leg contralateral to the source of ultrasound
(May & Hoy 1990). This movement has ~30 ms la-
tency and 45 to 65 ms time-to-peak, too rapid to
quantify with conventional video. We designed an

Engel & Wyttenbach: Optoelectronic Movement Sensor

optoelectronic detector to convert leg position to a
continuous voltage signal that indicates position
without the need of further processing. The de-
sign uses a triangular detection surface in such a
way that movement of the leg's shadow produces
a change in illumination, and incorporates a
method for canceling the light ripple that is found
in AC powered light sources. These features are
demonstrated below (Results and Discussion).
This device has performed well in measure-
ments of the cricket acoustic startle response
(e.g., Engel & Hoy 1999). Movement of the met-
athoracic femur is faithfully represented as a
voltage signal, and ripple due to AC line power is
eliminated. The circuit is conservatively designed
using parts that are readily available in electron-
ics stockrooms or outlets such as Radio Shack
(electronics vendors and photodiode suppliers are
listed in Appendix A). The circuit and photodetec-
tor are described in sufficient detail to be built as
they are, and they could also be adapted to a vari-
ety of other purposes. It is our hope that descrip-
tions such as this will enable more widespread
use of optoelectronic methods in entomology and
in biological research in general.


This section describes the essential features of
design and construction. Additional notes are pro-
vided in Appendix A, along with a list of parts and
suppliers. For an introduction to electronic com-
ponents, schematic diagrams, and assembly tech-
niques see Mims (2000) or other basic texts.

Movement Detector

There are several approaches to optoelectronic
movement detection. (1) Use of small arrays of
discrete photodectors to determine the position of
a light or shadow (e.g., Kittmann 1991; Erber &
Kloppenburg 1995; Roberts 1995). (2) Use of a lin-
ear position-detector photodiode chip (e.g., Helv-
ersen & Elsner 1977; Hedwig 1988; Kelly &
Chapple 1988; Mayer et al. 1988; Hedwig &
Becher 1998). (3) Use of a simple photodetector
chip to measure a moving shadow (e.g., Meyer et
al. 1987; Rusch & Thurm 1989; Clark et al. 1990;
Gotz 1987; May 1990). The first two approaches
are relatively complicated, and reports have not
given sufficient details for circuits to be con-
structed readily. Our device is a refinement of the
third approach.
If a shadow has a single moving edge, then a
simple rectangular photodiode can act as a posi-
tion sensor because the shaded area varies linearly
with position, resulting in a linear change in the
output of the photodiode (Meyer et al. 1987; Riisch
& Thurm 1989; Clark et al. 1990). However, when
the shadowing object has both leading and trailing
edges, as an insect appendage does, the shadow's

area does not change with position. In this case, a
mask with a triangular or crescent-shaped open-
ing can be interposed into the light path between
the appendage and the photodetector (Gotz 1987;
May 1990; May & Hoy 1991). As the appendage's
shadow moves to a wider part of the triangular
opening, the shaded area of the photodetector in-
creases. This makes a simple and effective position
indicator, as we show below (Results and Discus-
sion). In the cricket leg position detector described
here, a triangular mask is affixed directly to the
photodetector surface, as in May (1990), eliminat-
ing the need for optics between the mask and the
photodetector, as in Gotz (1987).
Our photodetector uses a 10 x 20 mm unpack-
aged silicon photodiode chip (EG&G Vactec, St.
Louis, MO, VTS 3081). A predecessor of our design
used a CdS photoconductor chip in a voltage-
divider circuit (May 1990). We chose a silicon pho-
todiode because of its superior frequency response
and uniform surface geometry compared with
CdS photoconductors (Photodiodes, Hamamatsu,
Bridgewater, NJ, 1997). For structural support, the
chip is fastened with double-sided foam tape to an
IC (integrated circuit) mount (Fig. 1C) with the
chip lead wires soldered to the pins. A second IC
mount serves as a socket, with wires leading to the
amplifier. In our setup this socket is attached to a
swivel ball joint mounted on a rod. The small size
of the detector assembly allows it to be placed
around electrodes for simultaneous neural and be-
havioral recording from a flying cricket (J.E.E., un-
published data) and minimizes disturbance of the
acoustic field.
The photodiode chip is masked with black
graphics tape (Chartpak, Leeds MA) to leave a
triangular area exposed (Fig. 1C). The tethered
cricket is illuminated from above with a fiber-op-
tic light guide, so that the metathoracic femur
casts a bar-shaped shadow onto the photodetec-
tor, which is 1 to 2 cm below the leg. As the leg piv-
ots laterally, its shadow moves to a wider part of
the triangle. The area of the shadow on the trian-
gle, and the resulting change in photocurrent
from the photodetector, are proportional to the
magnitude of lateral movement (see Results and

Position Amplifier

A photodiode produces a current directly pro-
portional to the amount of illumination (Photo-
diodes, Hamamatsu, Bridgewater NJ, 1997, p. 5).
The amplifier (Fig. 1A) converts this current sig-
nal to a more conveniently analyzed voltage sig-
nal with the desired level of gain. The first
operational amplifier (op amp), IC1, of the posi-
tion amplifier converts current (i) to voltage (u).
The current-to-voltage gain is determined by feed-
back resistance R, with the relationship v = iR.
The value of R is best chosen by trial and error

Florida Entomologist 84(3)

i position +vcc I

B battery




No_ +15V

R common



Fig. 1. Position detector design. A. Amplifier circuitry. The two components, the position amplifier and the ripple-
compensation amplifier, are outlined in dashed boxes. If the position amplifier is built alone, the points indicated by
asterisks are connected. Each IC is half of a 1458 dual op-amp package, with numbers indicating pin assignments.
In R6, R8, and R13, CCW indicates the lead that has zero resistance when the potentiometer is turned fully coun-
terclockwise. Abbreviations: C, capacitor; CCW, counterclockwise; IC, integrated circuit; NC, normally closed; NO,
normally open; PD, photodiode; R, resistor; SW, switch; Vcc, power supply voltage. See Appendix A for a list of parts.
B. Power can be provided by two 9 V batteries (left) or an external DC power supply not exceeding +18 V (right).
Open circles labeled +Vcc and -Vcc are nodes to be connected to the op-amp packages (pins 8 and 4, respectively)
and to the terminals of R6. C. Photodetector. A photodiode is masked to expose a triangular area and mounted to an
8-pin integrated circuit (IC) socket for physical support. A second IC socket serves as a connector to the amplifier.

September 2001

Engel & Wyttenbach: Optoelectronic Movement Sensor

because the appropriate gain depends upon the
strength of the light and the size and sensitivity
of the photodiode. In our setup, R is 10 or 100 k"
set by R1 or by R1 and R2 in parallel (R = R1 x R2/
(R1 + R2)). If lower light levels or a smaller pho-
todiode are used, an R of several megaohms might
be required. Switch SW1 allows switching be-
tween two gain settings during operation.
The second op amp (IC2) adds a DC offset to
the signal, allowing the baseline of the output
voltage to be adjusted to the middle of the input
range of a recording device or oscilloscope. Poten-
tiometer R6 controls the amount of DC offset,
functioning as a voltage divider together with re-
sistor R5. IC2 also inverts the voltage signal. In
our setup the anode of the photodiode (red lead) is
connected to the positive input of IC1 so that a leg
swing away from the body (which increases
shadow area) causes a negative photocurrent
fluctuation leading to a negative voltage signal.
IC2 inverts this so that the lateral leg swing of an
escape response produces a positive output sig-
nal. If a negative-going signal were desired, the
photodiode leads would be connected in the re-
verse orientation. Either orientation is permissi-
ble because the photodiode is not voltage-biased
in this circuit.

Ripple Compensation

Light from AC-powered lamps can have a pro-
nounced ripple at twice the line frequency (120 Hz
in North America). In our setup a standard fiber
optic light source (Dolan-Jenner, Woburn MA,
Series 180) provides strong illumination with low
heat and allows the lamp housing to be kept out-
side of the Faraday cage. However, the optic rip-
ple is considerable (Fig. 2B).
This ripple can be cancelled electronically (Fig.
2B). A reference photodiode is placed near the
movement detector but out of the path of the
cricket's shadow. Both photodiodes pick up the
light ripple, and the movement detector also
senses a change in illumination as the leg moves.
Ripple is eliminated by subtracting the reference
signal from the movement signal. The reference
photodiode need not be identical to the movement
detector photodiode, nor need their levels of illu-
mination be matched, because the gain of the ref-
erence signal can be adjusted over a large range.
Appendix B shows a straightforward method for
adjusting the gain so that the reference signal ex-
actly cancels the ripple in the movement signal.
The first op amp (IC3) of the ripple compensa-
tion amplifier converts the reference photocur-
rent to a voltage signal (Fig. 1A). Coarse gain
control is adjusted at this stage by potentiometer
R8, and fine gain control is adjusted at the second
op amp (IC4) by potentiometer R13. The reference
signal is subtracted from the movement signal by
feeding it into the positive input of the position

amplifier's second op amp (IC2). To build the posi-
tion amplifier only, without ripple compensation,
connect the points marked with asterisks in Fig. 1A
and omit all components in the ripple section.
An alternative to electronic subtraction of light
ripple is use of a DC-powered light source (Fig.
2B). We converted our Dolan-Jenner fiber optic
light source to DC power by cutting the wires that
connect the variable transformer to the lamp, di-
verting the transformer output to a DC converter
in a separate housing, and feeding the DC power
back to the lamp. The DC converter consists of a
full-wave rectifier protected with a heat sink,
with the output leads connected across 93,000 gF
of capacitance in parallel to the lamp. Because the
cost of capacitors increases with voltage rating,
15 V capacitors were used and the power supply
transformer is kept in the lower end of its range.
This provides ample light for our purposes.

Testing Linearity and Ripple Compensation

To show that the voltage signal is a linear func-
tion of shadow area, a metal rod (2 mm diameter)
was mounted on a motor so that it swung across
the photodetector 1.5 times per second (Fig. 2A).
The far edge of the photodiode was 108 mm from
the axle, giving the rod a speed of 1 m/s as it
passed over the photodiode. The output of the am-
plifier was recorded without ripple compensation
and with electronic compensation (Fig. 2B, left and
center traces). Then, with electronic compensation
inactivated, the DC converter was switched into
the lamp power supply without otherwise alter-
ing the setup (Fig. 2B, right trace).
To demonstrate adjustment of ripple compen-
sation (Fig. 3), the position photodetector was set
up to monitor the left metathoracic femur of a
tethered flying cricket. The reference photodetec-
tor was a fragment of a broken solar cell (~0.15
cm2 lit area). Voltage outputs were digitally sam-
pled at 5 kHz and 0.3 mV resolution. To show es-
cape responses in both directions (Fig. 3E), a
second photodetector and amplifier (also with rip-
ple compensation) monitored the right metatho-
racic femur. Both photodetectors were in place
throughout the series of records in Figure 3. Ul-
trasound pulses were 20 kHz carrier frequency,
10 ms duration. The cricket preparation and
acoustic setup have been described elsewhere
(May & Hoy 1991; Wyttenbach & Hoy 1997); as
previously, hind wings were cut short so they
would not interrupt the light path.


Linearity of Movement Detection

Photodiode chips have a uniform photosensitive
surface (in contrast to photoconductors, which
have a zigzag ribbon of photosensitive material).

Florida Entomologist 84(3)

piano wire ,

axle ..- V


0.5 V

10 ms



angular position




ripple subtracted

5 mm


DC-converted lamp

angular position

Fig. 2. System performance. A. Linearity testing. A 2mm thick rod passed across the photodiode at a speed of 1 m/
s. The shadow's area is proportional to its length on the triangular photodetector. The graph shows the calculated
length of the rod's shadow as a function of its angular position. B. Ripple compensation. Using the above setup, records
were made without compensation for light ripple (left), with electronic compensation (center), and with the light source
converted to DC operation (right). The two compensated traces are linear, as predicted in part A. Onset and offset are
not instantaneous because of the 2 ms time required for the full width of the shadow to enter or leave the triangle. C.
Calculations for a cricket leg. The metathoracic femur is essentially rectangular when viewed from above. As long as
its shadow crosses the entire triangle, the area of the shadow on the triangle is proportional to its midline length. The
calculated midline length is an approximately linear function of angular position over a range of at least 45.

The photocurrent output of a photodiode should For the output signal to faithfully indicate leg
be a linear function of the area that is illuminated deflection, however, the shaded area itself must
(or the area that is shaded). To test this, a rod be a linear function of the angular movement of
about as thick as a cricket hindleg femur was the femur. This assumption holds rather well,
moved across the triangular detector such that provided that the photodetector is positioned so
the area of the rod's shadow on the photodetector that the femur is perpendicular to the axis of the
increased uniformly with time (Fig. 2A). The out- triangle when the femur is in the middle of its
put signal increased uniformly as well (Fig. 2B), range of motion (Fig. 2C). This was demonstrated
indicating that the output signal is a linear func- with a trigonometric model. The shadow of the fe-
tion of shaded area. mur on the photodetector can be represented as a

September 2001

Engel & Wyttenbach: Optoelectronic Movement Sensor

A no ripple compensation B coarse gain adjustment

(reference signal)

L leg 50 ms 0.1 V 200ms button SW2 depressed

C before fine D after fine E acoustic startle responses
adjustment adjustment lateral

Ll medial

50 ms 50 ms T
I leg I 500 ms 0.2 V

IR sound pulse IL sound pulse

Fig. 3. Adjustment of ripple compensation. The position photodetector monitored the left metathoracic femur of
a tethered flying cricket. Voltage scale of Panel A applies to all panels except E. A. Signal without ripple compensa-
tion (SW3 open). Fluctuations are a combination of light ripple (120 Hz) and leg vibration due to wing beats (~35
Hz). B. Coarse gain (R8) is adjusted to minimize the jump of the trace when the reference signal is switched into
the movement amplifier path using pushbutton SW2. C. Ripple compensation is engaged (SW3 is closed) but fine
gain has not been properly balanced to eliminate ripple. D. Fine gain (R13) has been adjusted to minimize ripple.
Remaining fluctuations reflect real leg movements due to wing vibration. E. Escape responses in the same prepa-
ration. A second photodetector and amplifier monitored the right metathoracic femur. For both legs, a positive volt-
age signal indicates lateral deflection. Ultrasound pulses from the right and then the left side (arrows) evoked
outward lateral movements of the contralateral hind legs. Differences in amplitudes of vibration and escape re-
sponse between the two legs were real (not an artifact of using two different detectors and amplifiers).

bar of uniform thickness (Fig. 2C). As long as this
bar crosses two sides of the triangle, the area of
the shadow on the triangle will always be propor-
tional to the length of the shadow at its midline.
The trigonometric model shows that with optimal
positioning of the photodetector, this midline
length is an approximately linear function of an-
gular movement over a range of 450 (Fig. 2C) (leg
movement in the escape response rarely exceeds
20, Miles et al. 1992).
In our experiments it is not necessary to deter-
mine the relationship between the degree of
movement and the amplitude of the voltage sig-
nal. We are concerned with relative changes in
the amplitude of the leg response in a habituation
paradigm. The habituated responses are simply
normalized to control responses within the same
trial. However, the position-to-voltage scale for a
trial could be determined by moving the cricket's
femur to different angles using forceps, recording
the resulting voltage output levels, and measur-
ing the leg angles from videotape recordings
made during the same manipulations.

Compensation for Light Ripple

When illumination is provided by an AC light
source, the optic ripple adds substantial "noise" to
the movement signal (Fig. 2B; Fig. 3A). We com-
pared two methods for eliminating this ripple,
subtracting it electronically or adding a DC con-
verter to the light source. Both approaches
worked well (Fig. 2B). Electronic subtraction is
the most flexible method because any source of
light can be used. The gain of the reference signal
must be calibrated with each use, but this can be
done by a simple procedure (Appendix B, Fig. 3).
DC illumination is the more direct method. DC il-
luminators may not be commonly found in many
laboratories, but an AC unit can be converted to
DC operation as described above.

Other Applications

This design includes a photodetector that con-
verts angular position to a photoelectric signal, a
photodiode amplifier with DC offset compensa-

tion, and a second amplifier incorporating contin-
uous gain adjustment and subtraction of its
output from the first amplifier. These three com-
ponents could be adapted to a variety of uses. (1)
The photodetector could monitor other append-
ages such as wings or antennae, and the shape of
the mask could be altered as needed to track the
movement of a particular appendage (Gotz 1987).
Photodiodes come in several sizes, from 1 x 3 to 20
x 20 mm, and can be connected in parallel if
larger areas are needed (larger solar cells may
also be used). (2) The position amplifier could be
used with an unmasked photodiode in applica-
tions where the timing of movement is of more
interest than its spatial characteristics. For ex-
ample, in the laboratory an unmasked detector
could indicate the wing beat frequency of a teth-
ered insect or detect an animal's transit past a
point in a cage. The latter application could be
adapted in the field for counting visits to a colony
or lure. (3) The ripple compensation amplifier
could serve as the basis for other applications re-
quiring differential processing, such as automatic
compensation for variation in ambient light lev-
els. Another potential application is dual-photo-
diode movement detection (Crawford & Fettiplace
1985; Iwazumi 1987), in which a shadow overlaps
two adjacent photodiodes and moves onto one as
it moves off of the other.
The advantages of fine temporal resolution
and simplicity of analysis mentioned in the Intro-
duction make optoelectronic movement detection
attractive for a variety of applications. We hope
that this description will provide a point of entry
for workers without much electronics back-
ground, and a starting point for those with more
experience who can modify the design to suit their
particular applications.


We thank Bruce Land for comments on circuit de-
sign, and Ronald R. Hoy, in whose laboratory we carried
out this project (both are of Cornell University).

Calibration of photodiode measurements of cell mo-
tion by a transmission optical lever method. Pfluiigers
Arch. 415: 490-493.
chanical properties of ciliary bundles of turtle co-
chlear hair cells. J. Physiol. (London) 364: 359-379.
ENGEL, J. E., AND R. R. HOY. 1999. Experience-depen-
dent modification ofultrasound auditory processing in
a cricket escape response. J. Exp. Biol. 202: 2797-2806.

September 2001

ERBER, J., AND P. KLOPPENBURG. 1995. The modulatory
effects of serotonin and octopamine in the visual sys-
tem of the honey bee (Apis mellifera L.): I. Behav-
ioral analysis of the motion-sensitive antennal
reflex. J. Comp. Physiol. A 176: 111-118.
GOTZ, KG. 1987. Course-control, metabolism and wing
interference during ultralong tethered flight in
Drosophila melanogaster. J. Exp. Biol. 128: 35-46.
HEDWIG, B. 1988. Activation and modulation of auditory
receptors in Locusta migratoria by respiratory
movements. J. Comp. Physiol. A 162: 237-246.
HEDWIG, B., AND G. BECHER. 1998. Forewing move-
ments and intracellular motoneurone stimulation in
tethered flying locusts. J. Exp. Biol. 201: 731-744.
IWAZUMI, T. 1987. High-speed ultrasensitive instrumen-
tation for myofibril mechanics measurements. Am. J.
Physiol. 257: C253-C262.
KELLY, T. M., AND W. D. CHAPPLE. 1988.An inexpensive,
microcomputer-based system for recording move-
ments in real time. J. Neurosci. Methods. 23: 35-42.
KITTMANN, R. 1991. Gain control in the femur-tibia
feedback system of the stick insect. J. Exp. Biol. 157:
MAY, M. L. 1990. Biomechanics of Ultrasound-Induced
Steering in Tethered, Flying Crickets. Doctoral The-
sis, Cornell University.
MAY, M. L., AND R. R. HOY. 1990. Leg-induced steering
in flying crickets. J. Exp. Biol. 151: 485-488.
MAY, M. L., AND R. R. HOY. 1991. Habituation of the ul-
trasound-induced acoustic startle response in flying
crickets. J. Exp. Biol. 159: 489-499.
MAYER, M., K. Vogtmann, B. Bausenwein, R. Wolf, and
M. Heisenberg. 1988. Drosophila flight control during
"free yaw turns". J. Comp. Physiol. A 163: 389-399.
1987. Photoelectric recording of mechanical responses
of cardiac myocytes. Pfluigers Arch. 408: 390-394.
1992. Multisegmental analyses of acoustic startle in
the flying cricket (Teleogryllus oceanicus): Kinemat-
ics and electromyography. J. Exp. Biol. 169: 19-36.
MIMS, F. M. 2000. Getting Started in Electronics (3rd
ed.). Radio Shack U. S. A. 128 pp. (Radio Shack item
MOISEFF, A., G. S. POLLACK, AND R. R. HOY. 1978. Steer-
ing responses of flying crickets to sound and ultra-
sound: Mate attraction and predator avoidance.
Proc. Natl. Acad. Sci. USA 75: 4052-4056.
ROBERTS, W. M. 1995. Hummingbird licking behavior and
the energetic of nectar feeding. Auk 112: 456-463.
RUSCH, A., AND U. THURM. 1989. Cupula displacement,
hair bundle deflection and physiological responses in
the transparent semicircular canal of young eel.
Pfluigers Arch. 413: 533-545.
VON HELVERSEN, 0., AND N. ELSNER. 1977. The stridu-
latory movements of acridid grasshoppers recorded
with an opto-electronic device. J. Comp. Physiol. A
122: 53-64.
WYTTENBACH, R. A., AND R. R. HOY. 1997. Spatial acuity
of ultrasound hearing in flying crickets. J. Exp. Biol.
200: 1999-2006.

Florida Entomologist 84(3)

Engel & Wyttenbach: Optoelectronic Movement Sensor


Photodiode chips in sizes from 1 x 3 mm to 20
x 20 mm are available from Advanced Photonix
(Camarillo, CA, 805-987-0146, www.advanced-
photonix.com), Perkin Elmer (formerly EG&G
Vactec; St. Louis, MO, 314-423-4900, www.perki-
nelmer.com) and Hamamatsu (Bridgewater, NJ,
908-231-0960, usa.hamamatsu.com). Solar cells
of 20 x 40 mm are available from Edmund Scien-
tific (Barrington, NJ, 800-728-6999, www.ed-
sci.com) and Radio Shack (Fort Worth, TX, 800-
843-7425, www.radioshack.com). Allied Electron-
ics (Fort Worth, TX, 800-433-5700, www.alliede-
lec.com), Newark Electronics (Chicago, IL, 800-
463-9275, www.newark.com), and Radio Shack
are comprehensive vendors that carry all the re-
maining parts.
In addition to the two photodiodes, the circuit
shown in Fig. 1 A requires the following parts:
fixed 1/4 or 1/8 Watt resistors of 100 k2 (Rl), 11
kQ (R2), 10 kQ (R3, R4, Rll), 20 kQ (R5, R10), 5
kQ (R7, R12), and 1 kQ (R9); variable resistors of
10 kQ (R6), 1 MQ (R8), and 50 kQ (R13); 100 pF
capacitors (C1-2), type 1458 dual op-amp pack-
ages (IC1-4); SPST switches (SW1, SW3); SPDT
momentary pushbutton switch (SW2); DPST
switch (power on/off, not shown in Fig. 1). Note
that R6 and R13 should be 10-turn potentiome-
ters for greater precision in setting DC offset and
ripple gain. Other parts needed to house the cir-
cuit include a case, circuit board, connection jacks
(banana or BNC), and so on.
We used bipolar 1458 dual op amps (Radio
Shack 276-038) because they are resilient and
readily available. For biological applications re-
quiring exceptional high-frequency or low-noise
performance, this design could be refined by se-
lecting high-performance op amps, by optimizing
the feedback capacitance (C1), or by voltage-bias-
ing the photodiode. Guidelines can be found in
technical literature from photodiode manufactur-
ers (e.g., Photodiodes, Hamamatsu, Bridgewater,
NJ, 1997).
Resistors R1 and R2 should be at least 1 kQ to
prevent exceeding the op amp current rating, yet
small enough to avoid saturating the op amp at
high light intensities. The ideal values of R1 and
R2 for a particular setup are best determined by
trial and error. This process is made easier if R1
and R2 are plugged into an IC socket instead of
being soldered directly to the circuit board. The
range of gains available during normal use could
be extended over several orders of magnitude by
making SW1 a rotary switch and installing addi-
tional resistors. Capacitor C1 is included to pre-
vent ringing. However, it should be noted that the
feedback circuit is a low pass filter with a cutoff

frequency of f = [2m R1 Cl] 1; therefore C1 should
be small enough to avoid filtering out biological
signals of interest.
The maximum useful gain of the first op amp
of the position amplifier is limited because the DC
baseline resulting from overall illumination is
amplified along with the signal due to movement.
If movements of the leg shadow are small relative
to the lighted area of the photodiode, the final out-
put signal after DC compensation will also be
small. This can be countered to some extent by re-
ducing the unused lighted area of the photodetec-
tor as much as possible. At the second op amp, the
baseline signal is subtracted using DC offset com-
pensation. Therefore, an additional amplifier
stage could be placed after the second op amp if
more gain were needed. We have not found this to
be necessary.
In the DC offset compensation circuit, R5
should not be much greater than R4 because the
ratio R4/R5 is a gain factor that limits the avail-
able range of offset. At the same time, R5 must be
greater than R6 to give R6 sufficient linearity as
a variable voltage divider. A "voltage-follower"
amplifier could be added as a buffer between R6
and R5 if desired. This would make the R6 voltage
divider linear without regard to R5, and would
also allow a larger R6 to be used (to reduce the
current drain on batteries, for instance).

For electronic subtraction to be effective, ripple
in the reference and position signals must have
the same amplitude. This is achieved by adjusting
the gain of the ripple compensation amplifier
through the following simple procedure:
1. Adjust position detector amplifier. Set oscilloscope to
DC mode and confirm that ripple compensation is
not engaged (SW3 is open). Set the position amplifier
gain (SW1) and adjust DC offset (R6) to center the
signal (Fig. 3 A).
2. Adjust coarse gain of reference signal. Use the mo-
mentary pushbutton (SW2) to switch the reference
signal into the movement amplifier path (Fig. 3 B).
Adjust R8 to minimize the jump in the oscilloscope
trace as SW2 is pressed and released.
3. Adjust fine gain to minimize ripple. Switch the oscil-
loscope to AC mode and engage ripple compensation
by closing switch SW3 (Fig. 3 C). Adjust R13 to min-
imize the size of the ripple (Fig. 3 D). In this example
(Fig. 3 D), light-source ripple has been effectively
eliminated. The remaining "noise" is biological in or-
igin (leg vibration due to wing beats).
4. Readjust DC offset. Return the oscilloscope to DC
mode and adjust DC offset (R6) to center the trace.

Florida Entomologist 84(3)

September 2001


1USDA-ARS, Kika de la Garza Subtropical Agricultural Research Center, 2413 E. Highway 83, Weslaco, TX 78596

2USDA-APHIS-PPQ, 1913 SW 34th St., Gainesville, FL 32608.

'USDA-ARS Subtropical Horticulture Research Station, 13601 Old Cutler Road, Miami, FL 33158

4USDA-APHIS-PPQ, P.O. Box 2140, Moore Air Base, Mission TX 78572


Trap/lure combinations were tested against populations ofAnastrepha suspense (Loew) and
Anastrepha ludens (Loew) as substitutes for the traditional glass McPhail trap. Open-bot-
tom, plastic traps baited with a two component synthetic lure (ammonium acetate and pu-
trescine) caught as many and sometimes more fruit flies than the McPhail trap baited with
torula yeast. Sex ratio of flies caught with the synthetic lure was similar to that caught with
torula yeast, i.e., generally female biased, but variable among seasons and locations. The
synthetic lure attracted fewer non-target insects giving a substantial time savings in trap
maintenance. Moreover, the synthetic lure was effective for ten weeks without replacement.
Propylene glycol antifreeze increased captures significantly and improved preservation of
specimens when used as the trap liquid compared to water. Dry jar traps and cardboard
sticky traps were ineffective in comparison with the liquid baited traps.

Key Words:Anastrepha, traps, synthetic lure, fruit flies, pest detection


Combinaciones de varias trampas con diferentes cebos fueron evaluadas contra poblaciones
deAnastrepha suspense (Loew) yAnastrepha ludens (Loew) para substituir para la tradicio-
nal trampa vidrio de McPhail. Trampas de plastic con un cebo syntetico de dos componien-
tes (acetato amoniaco y putrescina), capturaron igual o mas moscas de fruta que la trampa
McPhail cebada con torula en agua. La proporci6n sexual de las moscas capturadas con el
cebo synt6tico fue igual que las capturadas con torula; generalmente hubo mas hembras,
pero, variable con respect a ubicaci6n y temporada. El cebo synt6tico atrayeron menos in-
sectos de otros tipos por su mejor efici6ncia resultando en menos tiempo manejando las
trampas. Ademas, el cebo synt6tico fue efectivo por diez semanas sin recebar. El anti-conge-
lante (glycol propilico), mezclado con el agua, aumenta las captures y preserve mejor los es-
pecimenes capturadas. Trampas secas y laminas pegajosas no fueron efectivas en
comparaci6n con las trampas cebadas con liquid.

McPhail traps baited with an aqueous slurry of
torula yeast have long been the industry standard
for tephritid fruit fly surveillance programs (Bur-
ditt 1982; Cunningham 1989). However, trap-back
studies using marked flies have shown that at the
usual trap densities, around 2-4 traps per kn2,
McPhail traps recapture substantially less than
one percent of the released individuals (Plant &
Cunningham 1991; Thomas et al. 1999). When one
considers this degree of trap efficacy in the context
of early detection of wild fly infestations, there is
an obvious need to increase trap densities, or, to
develop a more effective trap. Because the former
option is the less desirable for reasons of cost, the
latter potential has been investigated.
The McPhail trap is a bell-shaped, invaginated
glass jar, designed to be suspended in fruit trees,

with an opening at the bottom and a reservoir for
fluid of about 0.5 1 capacity. The fluid serves as
both the attractant and the catch mechanism,
with the flies attracted into the trap by the food
odor, then drowning in the liquid. In an early de-
sign, McPhail (1937) employed fermenting sugar
solutions, but later found success with protein
based lures (McPhail 1939; Steyskal 1977). This
led to the present standard of torula yeast hy-
drolysate with borax (Lopez et al. 1971).
Robacker & Warfield (1993), Heath et al.
(1995), and Robacker & Heath (1997), found that
amino acid metabolites associated with bacteria
and fermenting host fruits were highly attractive
to Anastrepha fruit flies. However, delivery sys-
tems designed to contain and release these chem-
icals are not easily inserted in the solid glass of

Thomas et al: Trap-lure Combinations for Anestrepha

the McPhail trap. Therefore, trap devices compat-
ible with the new lures have been designed and
tested. In experiments conducted in Guatemala,
Heath et al. (1997) found that a sticky trap baited
with three components, ammonium acetate, trim-
ethylamine and putrescine, caught about as
many and sometimes more Medflies, Ceratitis
capitata (Weidemann), and Mexflies Anastrepha
ludens (Loew), than did the McPhail/torula yeast
trap. Katsoyannos et al. (1999) reported that a
plastic, open-bottom, trap, baited with water and
the three component lure was five times more ef-
fective for catching Medflies than the same plas-
tic trap containing protein hydrolysate. The
protein based liquids are attractive to a broad
range of insects which is not the case with the
synthetic lures (Aluja 1999; Heath et al. 1995;
Katsoyannos et al. 1999). Because far fewer of the
non-target insects are caught, time spent servic-
ing the traps is substantially reduced.
Based on these preliminary experiences, we con-
ducted a series of tests in Florida against wild pop-
ulations of Caribfly, Anastrepha suspense (Loew),
in Texas against released, radiosterilized, Mexflies,
and in Nuevo Leon, Mexico against wildA. ludens
in their native habitat. It is known that species of
Anastrepha are not equally responsive to the tra-
ditional McPhail trap (McPhail 1939; Aluja et al.
1989). Ideally, one combination of trap and attrac-
tant could be deployed effectively against several
different pest tephritids. The purpose of these
tests was to provide direct comparisons of the
more promising trap/lure combinations.


Trap Devices

The standard glass McPhail trap was included
for comparison with the new trap/lure devices.
Two new plastic traps were tested which in size,
liquid reservoir capacity, and bottom opening di-
ameter, were similar to the McPhail trap. One is
manufactured by Florence Agri Investment Inc.
(Miami FL), hereinafter referred to as the FAI
trap, and the other by International Pheromone
(South Wirral, UK), hereinafter referred to as the
IPM trap. Both traps are of two piece construction
consisting of transparent upper halves separable
from yellow lower halves. They differed slightly in
conformation, the IPM trap being cylindrical with
a flat top, whereas the FAI trap is cylindrical but
with a rounded top. Also, in the IPM trap the top
half inserts within the bottom half, whereas in
the FAI trap the bottom half inserts into the top.
A sticky trap, called the ChamP' trap (Seabright,
Albany, CA) was also included in the tests. This
trap consisted of a square (15 x 15 cm), folding,
double-sided, perforated, yellow cardboard with
glue on the outside surfaces. Developed for use in
combination with fruit fly sex pheromones, this

trap has a metal hook for suspension in the target
host tree.


The standard aqueous torula yeast slurry,
three 5 gm yeast pellets (2% borax, manufactured
by ERA International, Freeport NY) dissolved in
350 ml water, was used in the McPhail traps, and
in some tests, the plastic traps. The yeast slurry
was renewed weekly when the traps were ser-
viced. A two component lure consisting of ammo-
nium acetate and putrescine was tested in the
ChamP traps and in both plastic traps. The lure is
marketed as Mediterranean fruit fly lure dual-
paks by CONSEP Inc. (Bend, OR). The capture
liquid was either 350 ml of water with 2% borax
and five droplets of Triton X-100R synthetic de-
tergent (Fisher Scientific, Pittsburgh, PA) to
break water tension, or, antifreeze (propylene gly-
col). The liquid, but not the lure, was renewed
each week when the traps were serviced. A dry
version of these traps included a 14 inch plastic
strip impregnated with pesticide (DDVP) and
baited with the two component lure.

Study Sites and Test Protocols

A series of pairwise tests, or in some cases, 3-
way tests, were conducted in Florida against pop-
ulations of A. suspense between February and
July, 1998, comparing different trap/lure combi-
nations. The duration of the tests varied from
three to 12 weeks, depending on fly activity, with
each test including five of each trap/lure tested,
with one of each combination in the same tree fol-
lowing Aluja et al., 1989. It was reasoned that
proximity would intensify the competition among
the traps. Tests were conducted at Labelle FL
with the traps hung in loquat trees, Eriobotrya
japonica (Thunb.) with a minimum distance of 1
m separating each trap. All traps were serviced
and their position within the tree alternated
weekly. At Ft. Pierce FL the target host was a
hedge, Eugenia uniflora (L.) (Surinam Cherry),
with the traps spaced at 3 m and rotated weekly.
In the Rio Grande Valley of Texas, McPhail
traps with torula yeast slurry, IPM traps with two
component lure and antifreeze, and IPM traps
with antifreeze alone, were compared during six
weeks of November and December 1998. Thirty of
each trap type were placed in individual trees in
a large commercial citrus grove (mainly grape-
fruit but some oranges) targeted by the weekly
sterile release program. The traps were posi-
tioned randomly with a minimum distance of 30
m (3 trees) separating each trap. The traps were
rotated when the traps were serviced weekly.
Near the town of Linares in the state of Nuevo
Leon, Mexico, traps were placed in yellow chapote
trees, Sargentia greggi (Wats.), for testing against

Florida Entomologist 84(3)

wild populations ofAnastrepha ludens. This test
was conducted during the spring of 1999 at five
sites with one of each of five trap/lure combina-
tions. These were: 1) the standard glass McPhail
trap with torula yeast, 2) the plastic IPM trap
baited with the two component lure and water, 3)
the plastic IPM trap with two component lure and
20% propylene glycol, 4) the dry version IPM trap
with the two component lure and a vapona strip,
and, 5) the ChamP sticky trap with the lure pack-
ets inside. Although the sticky traps were re-
placed weekly, the lure packets were simply
removed from the old sticky trap and placed in-
side the new trap each week.
Within each site a minimum distance of at
least 50 m was maintained between traps. The
trap positions were designated A-B-C-D-E and
each week the traps were rotated so that the trap
at A was moved to B, the trap at B went to C, etc.
The Mexican test continued for ten consecutive
weeks so that each trap was at each position twice
during the course of the test.

Statistical Analysis

Because populations and activity changed over
the course of the season, the numbers of flies
trapped tended to vary greatly from week to
week. Because this variation could mask differ-
ences in trap efficacy statistical comparisons were
made by converting the numbers captured to per-
cent of total weekly captures following Heath et
al. (1995). The mean percent weekly values were
then compared by a pairwise students t-test, The
t-score probabilities calculated by the software
program TPROB (Speakeasy Computing 1987).


Florida Tests

The results, including statistical analysis of all
trap-lure combinations, are shown in Table 1 and
summarized below.
Plastic vs. Glass McPhail Traps. The FAI plas-
tic trap was tested against the McPhail trap in an
area where A. suspense was breeding in loquat.
The test was run for eight consecutive weeks; both
traps baited with aqueous torula yeast with borax
as preservative. The McPhail traps caught
slightly more flies on average with a capture rate
of 18 flies vs. 13 flies per trap-week. The difference
in captures expressed as a percent of total was not
statistically significant between the traps.
Synthetic Lure vs. Torula Yeast. The two com-
ponent lure was tested in the FAI plastic trap
against the McPhail trap containing the torula
yeast slurry. The synthetic lure trap was equiva-
lent in effectiveness to the traditional trap cap-
turing a weekly mean of 42 flies vs. 37 flies per
trap; rates that were not significantly different.
The IPM trap gave somewhat better results. At
Labelle FL in loquats the IPM trap with the two
component lure caught many more flies than the
McPhail/torula trap, 77 vs. 34 flies per trap-week.
The same result was obtained at Ft. Pierce FL in
surinam cherry with the IPM trap taking 64 vs.
52 flies per trap-week. These differences when ex-
pressed as a percentage of flies caught gave a bor-
derline t-score of 1.93 which has a p of 0.06.
Importantly, the synthetic lure was effective
throughout the ten week test period. The longest
previous test of these lures was four weeks (Kat-


Trap-Lure N Mean + s.d. Trap-Lure N Mean + s.d. t df p

MAC-TY-WB 725 53.9 + 10.5 FAI-TY-WB 535 46.1+ 10.5 1.50 14 0.078
MAC-TY-WB 551 43.3 + 8.6 FAI-AP-WB 622 51.1 + 4.6 1.37 4 0.121
FAI-AP-WB 622 51.1 + 4.6 FAI-AP-dry 62 5.5 + 4.4 12.32 4 <0.001
FAI-AP-dry 62 5.5 + 4.4 MAC-TY-WB 551 43.3 + 8.6 6.73 4 0.001
MAC-TY-WB 509 40.5 + 14.2 IPM-AP-WB 1165 57.5 + 5.4 1.93 4 0.063
IPM-AP-dry 288 17.0 + 13.7 MAC-TY-WB 509 40.5 + 14.2 3.57 4 0.012
IPM-AP-WB 1165 57.5 + 5.4 IPM-AP-dry 288 17.0 + 13.7 4.76 4 0.004
MAC-TY-WB 4880 42.6 + 22.7 IPM-AP-WB 3835 57.4 + 22.7 1.22 12 0.123
MAC-TY-WB 975 31.8 + 10.1 IPM-AP-PG 204 68.2 + 10.1 5.70 8 <0.001
IPM-AP-WB 2616 38.0 + 5.9 IPM-AP-PG 4315 62.0 + 5.9 7.08 10 <0.001
IPM-AP-WB 6333 33.4 + 9.0 IPM-AP-PG 11029 66.6 + 9.0 7.37 14 <0.001
IPM-AP-WB 335 80.0 + 7.8 FAI-AP-WB 105 20.0 + 7.8 15.38 14 <0.001
IPM-AP-PG 5262 54.4 + 5.4 FAI-AP-PG 4450 45.6 + 5.4 2.84 10 0.009
IPM-AP-PG 5703 92.7 + 2.4 ChamP-AP 443 7.3 + 2.4 56.18 8 <0.001

September 2001

Thomas et al: Trap-lure Combinations for Anestrepha

soyannos et al. 1999). Also, similar to previous ex-
perience, we noted fewer non-target insects in the
synthetic lure traps.
Dry vs. Wet Traps. Flies captured in aqueous
solutions tend to decompose, constraining the
amount of information that can be recovered from
these specimens. This problem is exacerbated by
evaporation due to dry or windy weather, or de-
lays in servicing the traps. Another concern is
that the aqueous liquid may not be an effective
capture mechanism and that flies entering might
escape. A series of tests were conducted by substi-
tuting insecticide strips for the liquid as the kill-
ing agent. However, these test results were not
encouraging. The FAI trap with water captured
ten times as many flies per week on average (41.5
+ 6.4) as the dry insecticide version (4.1 2.4).
Similarly, the IPM trap with water captured four
times as many flies weekly (77.0 43.2) as the dry
version IPM trap (19.2 10.7).
Capture Liquid. Another alternative to the
preservation and evaporation problem is the use
of antifreeze instead of water as the capture liq-
uid. IPM plastic traps with the two component
lure and propylene glycol was tested against the
standard McPhail trap with torula yeast slurry.
The result was a marked improvement. The IPM
trap with antifreeze captured nearly twice as
many flies per trap: 89 vs. 39 mean flies weekly.
Another pair of tests was conducted using all IPM
plastic traps and synthetic lures so that the only
variable in the design was the capture liquid. In
both tests the water based traps captured only
half as many flies weekly as did the antifreeze
traps. The rate of capture expressed as a percent-
age of captures and compared by the t-test were
found to be significantly different (Table 1).
A series of tests were conducted to determine
the best antifreeze concentration. The results
were suggestive but inconclusive. A 50% solution
was tested against a 10% solution in April and
again in May. In the first test the 10% solution
was significantly better than the 50% solution
(45.1% vs. 35.7% of the flies per week). But, the re-
sults reversed in the May test where the 50% so-
lution caught slightly more flies, 40.9% vs. 35.4%,
although the difference was not statistically sig-

nificant. Also, in Texas an 8 week comparison was
made between the 10 and 20% concentration
against sterile Anastrepha ludens with no signifi-
cant difference in captures (33.2% vs. 26.9%).
IPM vs. FAI Traps. Two pairwise tests were
conducted to compare the plastic traps; one with
water as the capture liquid and one with anti-
freeze. The two component lure was used in all
traps in both tests. With water as the trap liquid
the IPM trap outperformed the FAI trap with a
rate of 7 vs. 2 flies per trap-week. With antifreeze
as the trap liquid the difference was less dramatic
but still significant when the data was converted
to percentages with the IPM trap catching 54.4%
vs. 45.6% of the flies.
Sticky Traps vs. Liquid Traps. The ChamP
traps were tested at Labelle FL against the IPM
traps baited with two component lure and 10%
propylene glycol as the capture liquid. The
ChamP trap was ineffective, capturing an order of
magnitude fewer flies, only 18 vs. 228 flies per
trap-week over 5 weeks.

Mexico Test

The experiment in Mexico differed from the
Florida testing in that all traps were tested simul-
taneously instead of pairwise, and the targeted
insects were native populations of A. ludens. Ta-
ble 2 provides compilation of the results which
were similar to those obtained in Florida with
A. suspense. The IPM trap with synthetic lure
and antifreeze was the best trap for capturing
wild Mexflies by a wide margin. It caught the
most flies at all five sites. Moreover, it was the
best trap in seven of the ten week test period and
was never worse than second best in the other
weeks. The next best trap/lure combination was
the synthetic lure in the IPM trap with water and
borax which caught about half as many flies as
the antifreeze version but twice as many as the
McPhail-torula trap. The IPM trap with synthetic
lure outperformed the McPhail trap in nine out of
the ten weeks tested and was the second best
trap, after the antifreeze trap, at all five sites. The
sticky trap was the least effective trap. It caught
the fewest flies at all five sites, catching none at


Trap-Lure Flies Mean + s.d. t df p

IPM-AP-PG 558 47.8 + 17.4 2.41 18 0.013
IPM-AP-WB 295 29.1 17.2 2.40 18 0.014
MAC-TY-WB 177 14.2 + 9.5 2.45 18 0.012
IPM-AP-dry 81 6.3 + 3.7 2.43 18 0.013
ChamP-AP 33 2.6 + 3.1

Florida Entomologist 84(3)

two sites. It was the worst trap in six of the ten
weeks and was never better than next to worst in
the other weeks. All of these differences in trap-
ping rates were statistically significant (Table 2).

Texas Test

The Texas test was conducted in a commercial
grove against aerially released, radiosterilized,A.
ludens. Temperatures varied sharply during this
test such that although an equal number of flies
was released weekly, the numbers trapped back
also varied sharply. The McPhail trap averaged
from 2.2 to 33.2 flies weekly for a mean of 17.0 +
10.5 flies per trap per week. The IPM traps aver-
aged from 4.4 to 24.8 flies weekly for a mean of
13.7 8.6 flies per trap per week. Because of the
variation in weekly captures, these numbers were
converted to percentage rates for comparison.
Nonetheless, the difference in means, 46.3%
(McPhail) vs. 41.3% (IPM), was not statistically
significant (t = 0.62, 10 d.f., p = 0.274). Interest-
ingly, the control traps containing only antifreeze
succeeded in capturing an average of 5.0 3.9
flies per trap per week. By comparison, Heath et
al. (1994) found that traps containing water alone
were not attractive to the Mexfly.
During this test three experienced trappers
were separately observed and timed with a stop-
watch as they serviced the traps. Servicing in-
volved emptying the trap, separating the fruit
flies from the other insects and placing them in vi-
als containing preservative, and recharging the
trap liquid. The average time required for one
person to service the McPhail trap was 150.7 sec
(n = 90). The average time required to service the
IPM traps was 107.7 sec (n = 90). It was judged
that the average difference, 43 sec, was due to the
lesser time it took to separate the fruit flies from
the IPM trap due to the presence of fewer non-tar-
get insects. It should be noted that had these
studies been conducted in an urban setting a
greater differential might have been found. Trap-
pers have reported incidents wherein house flies
have filled McPhail traps requiring up to 15-20
minutes in service time.
We attempted a cost-benefit analysis to compare
operating expenses of a program with the plastic
traps and synthetic lure versus a program using
the traditional McPhail trap with torula yeast.
However, because costs vary regionally, especially
for labor, and some materials are not universally
available, these, cost estimates are relative. As of
this writing the plastic traps range about 2-3 times
more expensive than the glass McPhail trap. But,
because the plastic traps are stackable and light in
weight there is a substantial savings in shipment
costs over the bulky glass traps. The cost of the
commercially available synthetic lure packets is
about ten times the cost of three torula yeast pel-
lets. However, whereas the yeast slurry must be re-

newed every week, the packets need be renewed
only once every ten weeks, giving an equivalent
cost for lure over the season. The use of 20% propy-
lene glycol incurs an additional cost which would
about double the weekly expense in expendable
materials, except that this liquid can be recycled
and reused three to four times before replacement.
The replacement rate varies because it is due
mainly to loss in handling (spillage and absorption)
rather than deterioration.

Sex Ratio and Reproductive Stage of Trapped Flies

Gender bias is an important concern because
surveillance trapping is often combined with SIT
suppression programs. Under these conditions
there could be an advantage to a female biased
trap (Katsoyannos et al. 1999). Trapping A. sus-
pensa in Florida, Calkins et al. (1984) reported a
strong female bias in McPhail traps baited with
protein by a ratio of 2:1 over males. Our Florida
results were similar. TheA. suspense females out-
numbered the males in all trap-lure combina-
tions. In five tests the McPhail trapped flies
ranged from 66.4 to 82.9% females with a mean of
75.0 + 6.2 percent. Captures in the plastic traps
baited with ammonium acetate and putrescine
with water ranged from 61.6 to 78.7% females for
a mean of 75.0 4.4 (n = 8). Traps with the two
component lure, but using propylene glycol as the
trap fluid, obtained the same result: a range of
62.9 to 85.3% females over six tests for a mean of
77.7 + 8.02 percent.
However, we obtained very different results with
A. ludens in Mexico. The McPhail traps caught
exactly the same number of males and females.
But, the synthetic lure traps were strongly male
biased during the ten week study. The flies caught
by the plastic traps baited with ammonium ace-
tate and putrescine with water were 71.3% males.
The flies caught with the same lure but with pro-
pylene glycol were 68.4% males. This result might
be explained by the studies of Lopez & Hernandez
(1967) withA. ludens who found that traps baited
with corn protein tended to catch more females,
while traps with fermenting bait (sugar and
yeast) tended to catch more males. Monitoring
populations ofA. ludens in Belize, Houston (1981)
found that sex ratio varied over the season, and
from place to place, although the variation was
between unity and a skewness in favor of females.
Likewise, Robacker (1999) reported gender differ-
ences in attraction to the synthetic lures by loca-
tion and season. Thus, sex ratio of trapped flies is
influenced by confounding factors which include
changes in the population structure and corre-
sponding changes in the response of the adults to
the attractant over the season. Our previous ex-
perience withA. ludens trapping has been equally
ambiguous. Annual surveys for Mexflies in citrus
groves in Texas with McPhail traps are consis-

September 2001

Thomas et al: Trap-lure Combinations for Anestrepha

tently female biased at a ratio of 3:1 (1,387 fe-
males vs. 411 males from January 1997 through
May, 2000). But traps in the chapote mots of
Nuevo Leon are generally male biased. From
1995 through 1998 we trapped 3,794 males to
2,711 females, a ratio of approximately 3:2. One
explanation is that there may be a stronger influ-
ence from lekking behavior in the chapote mots
compared to the citrus groves such that male cap-
tures are favored in the traps. Robacker (1993)
demonstrated experimentally that the male sex
pheromone strongly inhibits the attraction of the
immature females to the chapote trees with leks.
Using lab bioassays, Robacker & Warfield
(1993) found no significant difference between the
sexes ofA. ludens in attraction to torula yeast or to
synthetic lure. But, in further refining these tests,
Robacker (1999) demonstrated that male response
was strongly influenced by age, with older males
being significantly more attracted to the synthetic
lure. Our data (Table 3) shows the change in re-
productive status of the females over the course of
the Mexican field test in 1999. The larger numbers
captured in the last two weeks of the test was evi-
dently due to an influx of immature flies, indicat-
ing a strong local emergence of new adults. The
weeks with the largest numbers of flies were also
the weeks with the least skew in sex ratio. This
suggests that a factor contributing to the bias in
sex ratio over most of the test period was a general
absence of immature females. Thus, our results in-
dicate that with the two species of Anastrepha
tested there is a general bias for female captures
with the synthetic lure, as there is with the
McPhail-torula trap, but that this is subject to lo-
cal and seasonal variation in population structure
and prevailing environmental conditions.


The efficacy of a fruit fly trap is influenced by
weather (Cunningham et al. 1978, Gazit et al.

1998), by the habitat surrounding the tree with
the trap (Aluja et al. 1996), and even the position
of the trap within the tree (Hooper & Drew 1979;
Robacker et al. 1990). By rotating the traps
weekly we hoped to minimize these effects, but
because weather conditions also vary from week
to week, it was impossible to completely neutral-
ize the influence of the environment. It was also
known that different species of tephritids respond
differentially to the traps and lures (Aluja et al.
1989). Thus, we deemed it important to apply our
studies in different locations, in different habi-
tats, and against different species of fruit flies.
The results of the tests in Florida against A.
suspense suggest that the plastic versions of the
open-bottom trap can be substituted for the tradi-
tional glass McPhail trap without significant loss
in effectiveness. Likewise, the artificial lures based
on ammonia and putrescine can be used in the
plastic traps and catch as many, and often more
Caribflies, but with fewer non-target insects, than
the McPhail trap. Another advantage of the plastic
traps over the McPhail traps is their yellow color.
Greany et al. (1977) and Robacker (1992) found
that yellow to orange hues are visually attractive
to Caribfly and Mexfly respectively. Of the plastic
traps tested, the IPM trap outperformed the FAI
trap. Tests in Mexico against wild populations ofA.
ludens likewise demonstrated the superiority of
the plastic traps with synthetic lures over the tra-
ditional glass McPhail with torula yeast. Tests in
Texas against sterile flies only demonstrated
equivalence between the trap configurations in
terms of numbers captured, but a greater selectiv-
ity on the part of the synthetic lures, resulting in a
reduction of handling time by about one-third.
One might conjecture why there was such
large differences among the three configurations
of IPM traps and synthetic lure but different pre-
servatives. A concern with the open bottom trap is
that flies can exit the trap without getting caught
(Aluja et al. 1989). The flies have to fall into the


Julian week Gravid females Non-gravid females Males Percent males

14 15 3 50 73.5
15 29 5 50 59.5
16 32 2 72 67.9
17 25 1 68 72.3
18 65 0 100 60.6
19 14 1 22 59.5
20 14 3 19 52.8
21 8 6 48 77.4
22 6 4 55 84.6
23 7 43 67 59.0

Totals 215 68 551 66.1

liquid and drown to be actually trapped, an essen-
tially passive catch mechanism. The use of a
knockdown insecticide or a sticky contact surface
might have been expected to solve that problem,
yet the two traps with those active trap advan-
tages underperformed all of the liquid traps.
The greater captures by the wet IPM trap com-
pared to the McPhail trap can be explained as re-
sulting from the superior lure in the former.
However, it is much more difficult to explain the
better catch in the antifreeze trap compared to
the water trap, having the same lure and configu-
ration. The capture of flies in traps containing
only antifreeze suggests that the antifreeze has
an attractance. Inasmuch as propylene glycol it-
self is unlikely to be attractive to insects there
may be an impurity or breakdown product in the
commercial formulation which is attractive to
flies. For whatever reason, the antifreeze greatly
improved captures when used as the trap fluid in
combination with the two component lure.
Among the liquid based configurations we can-
not make absolute recommendations for one trap
over another. Ultimately, program managers
must decide which trap is appropriate for their
situation, among which, efficacy is but one consid-
eration. In programs where traps are rotated
among sites to follow fruit phenology, the porta-
bility of the traps may be an overriding factor.
Our studies provide information on the character-
istics of some of the trap-lure designs now avail-
able among those most likely to be useful in fruit
fly surveillance programs. For some programs the
detection of new infestations is the objective, as
opposed to the monitoring of existing populations,
and the trapping protocol will vary accordingly.
Having uniformity in trapping protocols among
programs is a consideration in that it facilitates
comparisons across regions. In some cases, for ex-
ample, where quarantine restrictions are trig-
gered by fly finds, the requisite trap design may
be codified (e.g., Nilakhe et al. 1991). Lastly, it
might also be noted that the improvement in effi-
cacy of the synthetic lures over the torula yeast is
incremental. Until the degree of improvement
reaches an order of magnitude, one has to expect
that further enhancements will be discovered,
and thus, even the best trap-lure designs now
available could be outmoded in the near future as
research in this area continues.

We are grateful to David C. Robacker, USDA-ARS,
Weslaco, TX, David Lance, USDA-APHIS, Otis, MA,
Richard L. Penrose, California Department of Food &
Agriculture, Sacramento, CA, and two anonymous re-
viewers, for advice and comments on the manuscript.
Mr. Danny Gates, USDA-APHIS, suggested the experi-
mental design for the Mexican portion of the testing.
Manuel Beltran of Florence Agri Investments, Miami
FL, provided the multilure traps. Pat Minyard of the

September 2001

California Department of Food & Agriculture provided
the ChamP traps. Dan Flores and Santiago Moreno, Jr.
assisted in the trapping study in Texas. Celestino Cer-
vantes, Ronay Riley and Francisco Daniel were diligent
in the servicing of the trap lines in Mexico.

ALUJA, M. 1999. Fruit fly (Diptera: Tephritidae) re-
search in Latin America: myths, realities and
dreams. An. Soc. Entomol. Brasil 28: 565-594.
AND F. AYORA. 1989. Behavior ofAnastrepha ludens,
A. obliqua and A. serpentina (Diptera: Tephritidae)
on a wild mango tree (Mangifera indica) harbouring
three McPhail traps. Insect Sci. Applic. 10: 309-318.
population fluctuations and ecological implications
for management of Anastrepha fruit flies (Diptera:
Tephritidae) in commercial mango orchards in
southern Mexico. J. Econ. Entomol. 89: 654-667.
BURDITT, A. K. 1982. Anastrepha suspense (Loew)
(Diptera: Tephritidae), McPhail traps for survey and
detection. Florida Entomol. 65: 367-373.
1984. Probability of detecting Caribbean fruit fly,
Anastrepha suspense (Loew) (Diptera: Tephritidae),
populations with McPhail traps. J. Econ. Entomol.
77: 198-201.
CUNNINGHAM, R. T. 1989. Population detection. pp. 169-
173, In A. S. Robinson and G. Hooper (eds.). World
Crop Pests 3B. Fruit Flies: their biology, natural en-
emies and control. Elsevier, Amsterdam.
URAGO. 1978. Tephritid fruit fly trapping: liquid food
baits in high and low rainfall climates. J. Econ. En-
tomol. 71: 762-763.
1998. Trapping females of the Mediterranean fruit
fly (Diptera: Tephritidae) in Israel: comparison of
lures and traps. J. Econ. Entomol. 91: 1355-1359.
CHAMBERS. 1977. Field studies on color preferences of
the Caribbean fruit flyAnastrepha suspense (Diptera:
Tephritidae). Entomol. Exp. & Appl. 21: 63-70.
fect on the attractiveness of a corn hydrolysate to the
Mediterranean fruit fly and severalAnastrepha spe-
cies (Diptera: Tephritidae). J. Econ. Entomol. 87:
F. JERONIMO. 1997. Adding methyl-substituted am-
monia derivatives to a food-based synthetic attrac-
tant on capture of the Mediterranean and Mexican
fruit flies (Diptera: Tephritidae). J. Econ. Entomol.
90: 1584-1589.
A. MANUKIAN, AND W. L. MEYER. 1995. Development
of a dry plastic insect trap with food based synthetic
attractant for the Mediterranean and Mexican fruit
flies (Diptera: Tephritidae). J. Econ. Entomol. 88:
HOOPER, AND DREW. 1979. Effect of height of trap on
capture of tephritid fruit flies with cuelure and me-
thyl eugenol in different environments. Environ. En-
tomol. 8: 786-788.

Florida Entomologist 84(3)

Thomas et al: Trap-lure Combinations for Anestrepha

HOUSTON, W. W. K. 1981. Fluctuations in numbers and
the significance of the sex ratio of the Mexican fruit
fly,Anastrepha ludens, caught in McPhail traps. En-
tomol. Exp. & Appl. 30: 140-150.
N. D. EPSKY, AND J. HENDRICHS. 1999. Field evalua-
tion of Mediterranean fruit fly (Diptera: Tephritidae)
female selective attractants for use in monitoring
programs. J. Econ. Entomol. 92: 583-589.
LOPEZ, F., AND 0. HERNANDEZ. 1967. Sodium borate in-
hibits decomposition of two protein hydrolysates at-
tractive to the Mexican fruit fly. J. Econ. Entomol.
60: 137-140.
new yeast hydrolysate-borax bait for trapping the
Caribbean fruit fly. J. Econ. Entomol. 64: 1541-1543.
MCPHAIL, M. 1937. Relation of time of day, temperature
and evaporation to attractiveness of fermenting
sugar solution to Mexican fruitfly. J. Econ. Entomol.
30: 793-799.
MCPHAIL, M. 1939. Protein lures for fruit flies. J. Econ.
Entomol. 32: 758-761.
DAVIDSON. 1991. Mexican fruit fly protocol helps ex-
port Texas citrus. Subtrop. Plant Sci. 44: 49-52.
PLANT, R. E., AND R. T. CUNNINGHAM. 1991. Analyses of
the dispersal of sterile Mediterranean fruit flies
(Diptera: Tephritidae) released from a point source.
Environ. Entomol. 20: 1493-1503.
ROBACKER, D. C. 1992. Effects of shape and size of col-
ored traps on attractiveness to irradiated, labora-

tory-strain Mexican fruit flies (Diptera: Tephriti-
dae). Florida Entomol. 75: 230-241.
ROBACKER, D. C. 1993. Understanding olfactory attrac-
tion in Anastrepha using A. ludens as a model sys-
tem. pp. 201-206. In M. Aluja and P. Liedo (eds.).
Fruit Flies: biology and management. Springer-Ver-
lag, New York, NY.
ROBACKER, D. C. 1999. Attraction of wild and labora-
tory-strain Mexican fruit flies (Diptera: Tephritidae)
to two synthetic lures in a wind tunnel. Florida En-
tomol. 82: 87-96.
ROBACKER, D. C., AND R. R. HEATH. 1997. Decreased at-
traction ofAnastrepha ludens to combinations of two
types of synthetic lures in a citrus orchard. J. Chem.
Ecol. 23: 1253-1262.
BARGER. 1990. Effects of trap color, height, and place-
ment around trees on capture of Mexican fruit flies
(Diptera: Tephritidae). J. Econ. Entomol. 83: 412-419.
ROBACKER, D. C., AND W. C. WARFIELD. 1993. Attraction
of both sexes of Mexican fruit fly,Anastrepha ludens,
to a mixture of ammonia, methylamine and pu-
trescine. J. Chem. Ecol. 19: 2999-3016.
STEYSKAL, G. C. 1977. History and use of the McPhail
trap. Florida Entomol. 60: 11-16.
SPEAKEASY COMPUTING. 1987. Speakeasy Manual,
Speakeasy Computing. Chicago, IL.
VLASIK, AND J. L. DAVIDSON. 1999. Mexican fruit fly
population suppression with the sterile insect tech-
nique. Subtrop. Plant Sci. 51: 61-71.

Florida Entomologist 84(3)

September 2001


1Universidad Aut6noma de Yucatan, Facultad de Medicina Veterinaria y Zootecnia, Departamento de Zoologia
Apartado Postal 4-116 Itzimnd, M6rida, Yucatan, M6xico

'Instituto de Ecologia, A. C. Departamento de Entomologia. Km 2.5 antigua carretera a Coatepec
Apartado Postal 63, 91000, Xalapa, Veracruz, M6xico


Examination of horseflies deposited in the Colecci6n Entomol6gica Regional of Universidad
Aut6noma de Yuca Merida, Yucatan, Mexico (CER-UADY) and Instituto Nacional de Diag-
nostico y Referencia Epidemiol6gicos, Mexico City, Mexico (INDRE) collections revealed a
significant number of species and new localities from the Peninsula of Yucatan (PY). Previ-
ously published information is summarized, and new information about tabanid species re-
ported for PY is presented, with emphasis on the most important protected areas within the
biotic province of Yucatan: Celestun, Cuxtal, Dzilam and Ria Lagartos (Yucatan), Calakmul
(Campeche), Sian Ka'an and El Eden (Quintana-Roo). Over 5,000 specimens collected by
netting, human bait, Malaise traps and light traps were examined. A final list of 29 species,
17 representing state records and three representing PY records, is provided. One species is
also reported for the first time from Mexico. Species diversity by state is as follows:
Campeche, 19 species, 10 new state records; Quintana Roo, 23 species, 2 new state records;
Yucatan, 22 species, 9 new state records. The 29 species reported for the biotic province of
Yucatan represents more than 14% of the species known from Mexico. Most of these species
have Neotropical or amphitropical affinities. Species showed wide distribution ranges
within the biotic province of Yucatan, probably related to climatic and orographic homoge-
neity, which define the limits of the province.

Key words: Tabanidae, Peninsula of Yucatan, Mexico


Al revisar los tabanos depositados en las colecciones Colecci6n Entomol6gica Regional of
Universidad Aut6noma de Yuca Merida, Yucatan, Mexico (CER-UADY) e Instituto Nacional
de Diagn6stico y Referencia Epidemiol6gicos, Mexico City, Mexico (INDRE), encontramos un
numero important de species y nuevos registros de localidades para la Peninsula de Yuca-
tan (PY). Compilamos y aportamos nueva informaci6n de tabanos reportados para la PY, con
enfasis en las areas protegidas mas importantes ubicadas en la Provincia Bi6tica de Yuca-
tan: Celestun, Cuxtal, Dzilam y Ria Lagartos (Yucatan), Calakmul (Campeche), Sian Ka'an
y El Eden (Quintana Roo). Se revisaron 5,067 ejemplares recolectados con red entomol6gica,
cebo human, trampas Malaise y de luz. Incluyendo registros previous y nuevos, obtuvimos
un listado final de 29 species, 17 nuevos registros estatales y tres nuevos registros para la
PY. Una especie es nuevo registro para M6xico. La diversidad de species por estado es la si-
guiente: Campeche (19 species, 10 nuevos registros estatales); Quintana Roo (23 species,
2 nuevos registros estatales); Yucatan (22 species, 9 nuevos registros estatales). Las 29 es-
pecies reportadas para la Provincia bi6tica de Yucatan representan mas del 14% de las es-
pecies conocidas para M6xico. La mayoria de las species tiene afinidad neotropical o
anfitropical. Las species mostraron un amplio rango de distribuci6n dentro de la provincia
bi6tica de Yucatan, probablemente relacionado con la homogeneidad climatica y orografica
que define los limits de la provincia.

The family Tabanidae includes approximately cies noted as widely distributed in Mexico (e.g,
4,290 species worldwide, of which nearly one- Diachlorus ferrugatus) might occur in PY.
third are Neotropical. In the most recent cata- Based on an extensive literature review, we re-
logue of Tabanidae south of the USA, 207 species corded 25 species that had been specifically re-
were reported for Mexico, but only 14 species ported for PY from Towsend (1897) to Fairchild &
were specifically reported for one or more of the Burger (1994): Catachlorops fulmineus var. ocel-
states included within the Peninsula of Yucatan latus Enderlein, Chlorotabanus mexicanus (Lin-
(PY) (Fairchild & Burger 1994). However, it was naeus), Chrysops auroguttatus Krober, C. flavidus
considered quite possible that some of the 22 spe- Wiedemann, C. pallidefemoratus Krober, C. sca-

Manrique-Saide et al.: Horseflies from the Yucatan

laratus Bellardi, C. variegatus (De Geer), Di-
achlorus ferrugatus (Fabricius), Esenbeckia illota
(Williston), Leucotabanus canithorax Fairchild,
L. exaestuans (Linnaeus), L. itzarum (Bequaert),
Scione aurulans (Wiedemann), Stenotabanus in-
dotatus Ibanez-Bernal, S. jamaicensis (New-
stead), S. littoreus (Hine), S. pechumani Philip,
Tabanus campechianus Towsend, T colombensis
Macquart, T commixtus Walker, T haemagogus
Williston, T occidentalis var. dorsovittatus Mac-
quart, T oculus Walker, T vittiger ssp.guatemala-
nus Hine and T yucatanus Towsend.
Furthermore, Ibanez-Bernal & Coscar6n (2000)
reported (without specific names) 9 species for
Campeche, 20 for Quintana Roo, and 10 for Yuca-
tan. Our review found 9 species for Campeche, 22
for Quintana Roo, and 13 for Yucatan.
The only previous faunistic studies of the ta-
banid fauna within PY were a general list of
Diptera for Sian Ka'an (Ibanez-Bernal et al. 1990)
and a detailed faunistic work of tabanids for this
protected area (Ibanez-Bernal 1992). These two
papers listed 17 species (Catachlorops fulmineus
var. ocellatus, Chlorotabanus mexicanus, Chrysops
flavidus, C. scalaratus, C. variegatus, Diachlorus
ferrugatus, Leucotabanus canithorax, L. itzarum,
Stenotabanus indotatus, S. jamaicensis, S. lit-
toreus, Tabanus campechianus, T colombensis,
T commixtus, T occidentalis var. dorsovittatus,
T oculus and T vittiger ssp. guatemalanus).
These 17 species, plus four unconfirmed records
by Ibanez-Bernal (C. auroguttatus, E. illota,
S. aurulans and T yucatanus) gives a total of 22
species for Quintana Roo. Only 10 of the species
had been previously reported.


Study Area

PY is situated in southeast Mexico and has a
surface area of 142,523 km2. PY inland vegetation
is tropical, mostly covered by short or medium
sized dry deciduous forests, although there are
some patches of high perennial forest. Coastal
vegetation includes dunes, mangroves, marshes
and petenes. The peten includes mangrove, short
and medium sized deciduous forests and swamp
vegetation elements. A more detailed description
of dominant vegetation can be found in Flores &
Espejel (1994). An interesting feature is the lack
of surface running water, although there are some
temporary ponds and cenotes (depressions in the
karst landscape filled with groundwater) (Bren-
ner et al. 1995).
PY includes the Mexican states of Campeche,
Quintana Roo and Yucatan. PY includes two bi-
otic provinces: biotic province of Yucatan (north-
ern PY) and Peten (southern PY) (Barrera 1962;
Alvarez & de Lachica 1991). The present report
deals with the biotic province of Yucatan, de-

scribed as an area including low level carstic soils,
without surface running water, situated in the
north of PY up river from the mouths of the
Champoton and Hondo rivers (Alvarez & de Lach-
ica 1991).
More than 5,000 specimens deposited at CER
and INDRE were examined. These specimens
were collected by netting, human bait, Malaise
traps and light traps. The authors and Adriana
Godinez, Her6n Huerta, Carmen Martinez, Leti-
cia Miranda, Carlos Navarro, Rafael Paz, Cresen-
cio Perez, from INDRE and UADY collected the
tabanids. Dr. Atilano Contreras and his team
from Instituto de Biologia UNAM, also provided
specimens from Calakmul, Campeche.
All specimens were collected within seven pro-
tected areas (defined as terrestrial areas repre-
senting different ecosystems and their biodiversity
under special governmental regime of protection
conservation, restoration and development; SE-
MARNAP, 1996), Celestun, Cuxtal, Dzilam and
Ria Lagartos (Yucatan), Calakmul (Campeche),
Sian Ka'an and El Eden (Quintana-Roo) repre-
senting almost 10% of PY land area (Fig. 1). For
details about species and collection localities of
Sian Ka'an readers are referred to Ibahez-Bernal
(1992). There are few records from El Eden, but
these were included because of their importance
although it is clear that its tabanid diversity is
poorly known. Other localities of the states were
also included. Even though political borders do
not have any biological meaning, they are useful
to reference localities where species are recorded.
The localities in PY are shown in Fig. 1.


INDRE = Instituto Nacional de Diagn6stico y
Referencia Epidemiol6gicos; CER-UADY = Colec-
ci6n Entomol6gica Regional of Universidad Au-
tonoma de Yucatan; CD = Coastal dune; DTF =
Deciduous tropical forest; M = Mangrove; P = Pe-
ten; PY = Peninsula of Yucatan.


Identification of 5, 067 specimens provided a
list of 23 species (Table 1) including 17 state
records (see species accounts), and 3 records for
PY (Lepiselaga crassipes, Phaeotabanus longiap-
pendiculatus, Tabanus pungens). One species
(Chrysops varians) is recorded for the first time
from Mexico. Seventeen state records were con-
firmed and six species previously reported for PY
(Catachlorops fulmineus var. ocellatus, Chrysops
auroguttatus, Esenbeckia illota, Scione aurulans,
Stenotabanus indotatus and S. pechumani) were
not found in this study.
At present, including previous and new
records, PY has the following tabanid species:

Florida Entomologist 84(3)

Fig. 1. Protected areas sampled in the Peninsula of Yucatan (redrawn from Bezaury et al. 1995). The map shows
the borders of the states of Tabasco and Campeche and with the countries Guatemala and Belize.

Campeche: 19 species, including the following
new state records: Chrysops scalaratus, C. varians,
C. variegatus, Lepiselaga crassipes, Leucotabanus
canithorax, L. exaestuans, Tabanus colombensis,
T commixtus, T occidentalis var. dorsovittatus and
T oculus.
Quintana Roo: 23 species, including 2 new
state records (Lepiselaga crassipes and Tabanus
Yucatan: 22 species, including the following new
state records: Chrysops flavidus, C. scalaratus,
C. varians, Lepiselaga crassipes, Phaeotabanus
longiappendiculatus, Tabanus campechianus,
T colombensis, T occidentalis var. dorsovittatus
and T pungens.

Annotated List of Tabanidae from Peninsula of Yucatan

Tribe Pangoniini
Esenbeckia illota (Williston), 1901

Known Distribution: From Mexico (Chiapas,
Campeche, Nuevo Leon, Quintana Roo and Yuca-
tan) to Belize and Honduras (Williston 1901;
Bequaert 1931; 1933; Pearse 1945; Philip 1954a,
1978b; Fairchild & Burger 1994). Within PY, Philip
(1978b) recorded one female from Campeche.
Bequaert (1931) reported one female from Quin-
tana Roo (cited as Territory of Quintana Roo) and
Philip (1954a) reported the existence of specimens

September 2001

Manrique-Saide et al.: Horseflies from the Yucatan


Campeche Quintana Roo Yucatan

Species CA Others ED SK (*) Others CE CU DZ RL Others

Catachlorops fulmineus var. ocellatus +
Chlorotabanus mexicanus + + +
Chrysops auroguttatus +
C. flavidus + + + +
C. pallidefemoratus + + + + + + +
C. scalaratus + + + +
C. varians + +
C. variegatus + + + + + +
Diachlorus ferrugatus + + + + + + + + +
Esenbeckia illota + + +
Lepiselaga crassipes + + +
Leucotabanus canithorax + +
L. exaestuans + + +
L. itzarum + + + + + +
Phaeotabanus longiappendiculatus +
Scione aurulans +
Stenotabanus indotatus +
S. jamaicensis + + +
S. littoreus + + + + + +
S. pechumani +
Tabanus campechianus + + + + + +
T colombensis + + + + + +
T commixtus + + + + + + + + +
T haemagogus + + + + +
T occidentalis var. dorsovittatus + + + + + +
T oculus + + + + + + + +
T pungens + +
T vittiger ssp. guatemalanus + + + + + +
T yucatanus + + + +

from Yucatan, without more detailed information.
None of the reports for Campeche and Yucatan
mentioned the collection date.

Tribe Scionini
Scione aurulans (Wiedemann), 1830

Known Distribution: From Mexico (Chiapas,
Oaxaca, Tabasco and Veracruz) to Guatemala, Be-
lize, Honduras, Costa Rica and Colombia (Philip
1954a, 1978a; Fairchild & Burger 1994). This spe-
cies was reported from PY (Bequaert 1931, 1933;
Pearse 1945) for Quintana Roo (cited as Territory
of Quintana Roo).

Tribe Chrysopsinae
Chrysops auroguttatus Kr6ber, 1930

Known Distribution: From Mexico (Veracruz),
to Costa Rica, Panama, northern Colombia and

Trinidad (Fairchild 1986; Fairchild & Burger
1994). Within PY this species was reported from
an unknown locality in Quintana Roo (cited as
southeast of Peto) by Bequaert (1931) (as
Chrysops incisa Macquart, 1845). However, it was
not found in our collections, nor was it collected by
Ibanez-Bernal (1992) in Quintana Roo. This spe-
cies is very similar to C. pallidefemoratus and it is
possible that the previous records were misiden-

Chrysops flavidus Wiedemann, 1821

This species is widely distributed from the
southern USA to Mexico, as well in Cuba and Ba-
hamas (Cruz & Garcia 1974; Fairchild & Burger
1994). It has been reported from the Mexican
states of Hidalgo, Tabasco, Veracruz (Fairchild &
Burger 1994) and Quintana Roo (Sian Ka'an)
(Ibanez-Bernal 1992). It is reported here for the
first time in Yucatan, mainly from coastal areas

Florida Entomologist 84(3)

within coastal dune and mangrove, but also from
a deciduous tropical forest close to the coast and
Material Examined: YUCATAN, Celestun Re-
serve, 11 Celestun (CD), 4-XII-1995, 1 9 Eco-
paraiso (CD), 10-IV-1997, 3 9 Rancho Loma
Bonita (P), 7-IX-1995, 14 9 Ibid. 8-XII-1995, 1 9
Rancho Loma Bonita (DTF), 19-VII-1996, 2 9
Ibid. 5-IX-1995, 1 9 Ibid. 23-X-1996, 23 9 Ibid. 5-
XII-1995; Dzilam Reserve, 2 9 Dzilam, 20-1-1993,
net, 1 9 2-II-1993, net; Ria Lagartos Reserve, 2 9
Rio Lagartos, 6-III-1994, Malaise trap, 1 9 El
Cuyo, 23-V-1995, net, 1 9 EL Cuyo (M), 18-VII-
1996, 1 9 Ibid. 28-XI-1997, 3 9 Ibid. 29-XI-1995,
13 9 Ojo de Agua (P), 8-VII-1996, 5 9 Peten
Tucha, 31-XI-1995, 3 9 Ibid. 1-XII-1995, 4 9 En-
trada a Sac-Bo (DTF), 9-VII-1996.

Chrysops pallidefemoratus Krober, 1930

Known Distribution: From Trinidad, Belize
and the three states of PY (Bequaert 1931, 1944;
Fairchild 1942a; Fairchild 1971; Fairchild 1986;
Ibanez-Bernal 1992; Fairchild & Burger 1994).
Material examined: CAMPECHE, Calakmul
Reserve, 3 Ejido Nuevo Becan, El Chorro, 30-V-
1997, Malaise trap, 7 El Refugio, 2-1-1993, net.
YUCATAN, Cuxtal Reserve, 9 Xmatkuil (DTF),
28-31-1-1994, Malaise trap, 3 Xmatlkuil, 24-28-II-
1994, Malaise trap; Ria Lagartos Reserve, 2 El
Cuyo, 23-V-1995, net; Ria Lagartos Reserve, 1 Pe-
ten Tucha, 8-VII-1996, 1 La Darcena (CD), 10-
VII-1997, 1 Ibid. 28-VIII-1995.

Chrysops scalaratus Bellardi, 1859

The type locality is Mexico. Known distribu-
tion from Panama to southern Mexico including
the states of Chiapas, Quintana Roo (Sian Ka'an)
and Veracruz (Ibanez-Bernal 1992; Fairchild &
Burger 1994). It is reported here for the first time
for Campeche and Yucatan.
Material Examined: CAMPECHE, Calakmul
Reserve, 1 archaeologic zone, 3-V-1997, Malaise
trap, 1 9 Ejido Nuevo Becan, El Chorro, 30-V-
1997, Malaise trap. YUCATAN, Celestun Re-
serve, 9 9 Celestun (CD), 4-IX-1995, 58 9 Rancho
Loma Bonita (P), 7-IX-1995, 10 9 1 6 Rancho
Loma Bonita (DTF), 20-1-1997, 233 9 Ibid. 5-IX-
1995, 9 9 Ibid. 23-X-1996, 9 9 Ibid. 5-XII-1995;
Ria Lagartos Reserve, 4 9 El Cuyo (CD) 23-V-
1995 net, 4 9 Ibid. 5-V-1994 Malaise trap, 1 9 La
Darcena, 13-1-1997, 1 9 Rio Lagartos, 6-III-1994,
Malaise trap, 4 9 El Cuyo (M), 29-XI-1995, 1 9
San Felipe, 26-V-1995, Malaise trap, 2 9 Rio
Lagartos (P), 4-IV-1997, Malaise trap, 1 9 Ojo de
Agua (P), 27 9 1 6 Ibid. 8-VII-1996, 1 9 Ibid. 31-
VIII-1995, 179 1 6 Ibid. 14-X-1996, 8 9 Ibid. 1-
XII-1995, 17 9 Entrada a Sac-Bo (DTF), 21-I-
1997, 2 9 Ibid. 1-IV-1997, 5 9 Ibid. 9-VII-1996, 1
9 Ibid. 9-VII-1997, 1 9 Ibid. 29-VIII-1995, 7 9

Peten Tucha, 7-VII-1997, 2 9 Ibid. 31-VIII-1995,
1 9 1 6 Ibid. 1-IX-1995, 6 9 Ibid. 30-XI-1995.

Chrysops varians Weidemann, 1828

Known Distribution: Costa Rica, Panama, Co-
lombia Venezuela, Guyana, Brazil, Ecuador, Peru,
Paraguay, Argentina and Trinidad (Fairchild &
Burger 1994). This is the first time that C. varians
is reported for Mexico.
Material Examined: 8 9 MEXICO, YUCATAN,
Ria Lagartos Reserve, El Cuyo, 23-V-1995, P. Man-
rique coll., net; 2 9 CAMPECHE, Calakmul Re-
serve, archaeologic zone, 3-V-1997, Contreras,
Gonzalez, Ibarra and Martinez colls., Malaise trap.

Chrysops variegatus (De Geer), 1776

Known Distribution: Includes most of the Neo-
tropics from Mexico to Argentina, as well as Cuba
and the West Indies (Fairchild 1971, 1986; Cruz &
Garcia 1974; Fairchild & Burger 1994). In Mexico,
it has been reported from Jalisco, Tabasco, Quin-
tana Roo (Sian Ka'an), Veracruz and Yucatan
(Chichen Itza) (Bequaert 1931, 1933; Pearse
1945; Ibanez-Bernal et al. 1990; Ibanez-Bernal
1992). It is reported here for the first time for
Material Examined: CAMPECHE, Calakmul
Reserve, 10 Y archaeologic zone, 3-V-1997, Mal-
aise trap, 2 Y Ejido Nuevo Becan, El Chorro, 30-
V-1997, Malaise trap, 2 Y El Refugio, 2-1-1993,
net. YUCATAN, Celestun Reserve, 1 Y Rancho
Loma Bonita (P), 26-III-1996, 2 Y Ibid. 8-VII-
1996, 1 Y Ibid. 4-XII-1995, 1 Y Rancho Loma Bo-
nita (DTF), 1 Y Ibid. 28-XI-1995, 1 Y Ibid. 5-XII-
1995, Dzilam Reserve, 7 Y Dzilam, 2-2-1993, net;
Ria Lagartos Reserve, 1 Y San Felipe, 26-5-1995,
net, 2 9 Rio Lagartos, 6-III-1994, Malaise trap, 4
9 Ojo de Agua (P), 8-VII-1996, 1 9 Peten Tucha,
2-IX-1995, 9 1 Ibid. 15-X-1996, 1 9 Ibid. 30-XI-
1995, 1 9 Entrada a Sac-Bo (DTF), 19-III-1996, 2
9 Ibid. 9-VII-1996, 2 9 Ibid. 9-VII-1997.

Tribe Diachlorini
Stenotabanus indotatus Ibanez-Bernal, 1992

Described by Ibanez-Bernal (1991) and re-
ported (Ibanez-Bernal 1992) for Quintana Roo
(Sian Ka'an), Mexico. It is apparently confined to
that area (Ibanez-Bernal 1992; Fairchild &
Burger 1994).

Stenotabanus jamaicensis (Newstead), 1909

Known Distribution: Jamaica, Cuba, Hispani-
ola, Puerto Rico, Cayman Islands and the Baha-
mas (Bequaert 1940; Philip 1958; Cruz & Garcia
1974). Its continental distribution is restricted to
Mexico and Belize (Fairchild & Burger 1994).

September 2001

Manrique-Saide et al.: Horseflies from the Yucatan

Within PY it has been reported for Quintana Roo
(Sian Ka'an) (Ibianez-Bernal 1992) and Yucatan
(Fairchild & Burger 1994).
Material Examined: YUCATAN, Ria Lagartos
Reserve, 2 Y Rio Lagartos (CD), net.

Stenotabanus littoreus (Hine), 1907

Known Distribution: From Mexico to Panama
(Fairchild 1953, 1971; 1986; Fairchild & Burger
1994). It has been previously reported in Mexico
for Quintana Roo (Colonia Santa Maria and Sian
Ka'an) (Bequaert 1931; Pearse 1945; Ibanez-Ber-
nal 1992) and Yucatan (Puerto Progreso) (Philip
1978a). Pearse (1945) erroneously cited Colonia
Santa Maria, which is near to Puerto Morelos, as
Puerto Morelos. The species is widely distributed
along the coast of Quintana Roo and Yucatan.
Material Examined: YUCATAN, Celestun Re-
serve, 59 Y Celestun (CD), 4-IX-1995, 1 Y Dumac
(M), 3-VII-1997, 7 Y Ibid. 6-IX-1995, 1 Y Rancho
Loma Bonita (M), 8-VII-1996, 1 Y Ibid. 7-IX-
1995, 2 Y Rancho Loma Bonita (DTF), 15-VII-
1996, 1 Y Ibid. 5-IX-1995; Dzilam Reserve, 14 Y
Dzilam, 15-VIII-1993, Malaise trap; Ria Lagartos
Reserve, 1 Y El Cuyo (M), 8-VII-1997, 1 Y Ibid.
23-V-1995, 3 Y Ibid. 11-VII-1996, 2 Y La Darcena
(DC), 10-VII-1997, 4 Y Ibid. 28-VIII-1995, 12
Ibid. 17-X-1996; 1 Y Ojo de Agua (P), 31-VIII-
1995, 3 Y Peten Tucha, 8-VII-1996.

Stenotabanus pechumani Philip, 1966

Known Distribution: From Texas, USA to Mex-
ico in the states of Veracruz and Campeche
(Philip 1966; Fairchild & Burger 1994). We could
not confirm its presence in PY.

Catachlorops fulmineus var. ocellatus Enderlein, 1925

Known Distribution: From Mexico to Colombia
(Fairchild & Burger 1994). In Mexico, it has been
collected in Chiapas, Tabasco and Veracruz (Fair-
child 1940). Only one specimen has been collected in
Quintana Roo (Sian Ka'an) (Ibianez-Bernal 1992).

Diachlorus ferrugatus (Fabricius), 1805

Known Distribution: From the southern USA,
Bahamas, Mexico and Central America as far
south as Costa Rica (Fairchild 1971). In Mexico, it
has been found in Tabasco and PY, in Campeche,
Yucatan (Puerto Progreso) (Bequaert 1931;
Pearse 1945) and Quintana Roo (Sian Ka'an and
Cancun) (Ibanez-Bernal 1992). This species is dis-
tributed around the Gulf of Mexico.
Material Examined: CAMPECHE, Calakmul
Reserve, 14 9 archaeologic zone, 3-V-1997, Mal-
aise trap, 5 9 El Refugio, 2-1-1993, Malaise trap.
QUINTANA ROO, 1 9 Vallehermoso, Rancho 3,
19-VII-1993, Malaise trap, 15 9 Vallehermoso,

24-VII-1993, Malaise trap. YUCATAN, Celestun
Reserve, 1 Y Celestun (CD), 15-1-1997, Malaise
trap, 19 Y Peten, 8-9-IV-1997, Malaise trap, 6 Y
Peten, 1-VII-1997, Malaise trap; 1 Y Celestun
(CD), 3-VII-1997, 1 Y Ibid. 4-IX-1995, 2 Y Ibid. 8-
XII-1995, 3 Y Dumac (M), 28-III-1996, 3 Y Ibid.
4-VII-1997, 4 Ibid. 6-IX-1995, 1 1 6 Ibid. 16-X-
1996, 96 Y 6 6 Ecoparaiso (CD), 21-1-1997, 4Y
Ibid. 10-IV-1997, 15 Y Ibid. 17-VII-1996, 3 Y Pe-
ten 2, 20-1-1997, 16 Y Ibid. 8-IV-1997, 53 Y Ibid.
30-VI-1997, 234 Y 6 6 Rancho Loma Bonita (P),
25-III-1996, 16 Y Ibid. 16-VII-1996, 67 Y Ibid. 7-
IX-1995, 13 Y Ibid. 24-X-1996, 23 Y Ibid. 4-XII-
1995, 163 Y Rancho Loma Bonita (DTF), 26-III-
1996, 71 Y Ibid. 1-IV-1997, 162 Y Ibid. 16-VII-
1996, 22 Y Ibid. 5-IX-1995, 28 Y 1 6 Ibid. 23-X-
1996, 47 Y Ibid. 5-XII-1995; Cuxtal Reserve, 1 Y
Dzununcan (DTF), 8-IV-1993, net, 5 Y Xmatkuil
(DTF), 17-1-1996, net; Dzilam Reserve, 5 Y Dz-
ilam, 5-V-1993, net; Ria Lagartos Reserve, 74 Y
Rio Lagartos, Peten 4, 7-VII-1997, Malaise trap, 1
Y El Cuyo, 5-V-1994, Malaise trap, 1 Y El Cuyo,
23-V-1995, Malaise trap, 2 Y San Felipe, 26-V-
1995, net, 7 Y Ojo de Agua (P), 21-III-1996, 39 Y
Ibid. 8-IV-1997, 2 Y Ibid. 30-VI-1997, 348 Y 4 6
Ibid. 8-VII-1996, 2 Y 3 6 Ibid. 31-VIII-1995, 5 Y
Ibid. 14-X-1996, 4 Y Ibid. 30-XI-1995, 1 Y Ibid. 1-
XII-1995, 111 Y Peten Tucha, 16-1-1997, 32 Y
Ibid. 4-IV-1997, 179 Y Ibid. 8-VII-1996, 4 Y 1 6
Ibid. 1-IX-1995, 12 Y Ibid. 30-XI-1995, 25 Y Ibid.
1-XII-1995, 7 Y Camino a Nuevo Tekal (DTF), 9-
VII-1997, 1 Y Entrada a Sac-Bo (DTF), 13-1-1997,
3 Y Ibid. 19-III-1996, 94 Y Ibid. 1-IV-1997, 32 Y
1 6 Ibid. 9-VII-1996, 1 Y Ibid. 29-VIII-1995, 1 Y
Ibid. 5-IX-1995, 1 Y Ibid. 15-X-1996, 2 Y Ibid. 28-
XI-1995, 1 Y La Darcena (CD), 13-1-1997.

Chlorotabanus mexicanus (Linnaeus),1758

Known Distribution: From southern Mexico to
northern Brazil (Fairchild and Burger 1994). Ac-
cording to Fairchild (1986) and Ibanez-Bernal
(1992), it has been found in Campeche, Chiapas,
Quintana Roo (Sian Ka'an) and Veracruz. This
species has been reported from rain forest areas,
which probably restricts its distribution in PY to
a few southern areas.
Material Examined: CAMPECHE, Calakmul
Reserve, 9 Y archaeologic zone, 3-V-1997, Malaise
trap, 1 Y El Refugio, 2-1-1993, light trap.

Phaeotabanus longiappendiculatus Macquart, 1855

Known Distribution: From Mexico to Panama.
The first report for Mexico was Bellardi (1859) (as
Tabanus luteoflavus Bellardi 1859) and the last
was Towsend (1897) (as Tabanus limonus var.
mexicanus Towsend 1897). It is reported here for
the first time for PY and Yucatan State.
Material Examined: YUCATAN, Celestun Re-
serve, 1 Y Rancho Loma Bonita (P), 7-IX-1995,

Florida Entomologist 84(3)

Malaise trap, 2 Entrada a Sac-Bo (DTF), 9-VII-
1996, Malaise trap, 4 Ibid. 15-X-1996.

Catachlorops fulmineus var. ocellatus Enderlein, 1925

Known Distribution: From Mexico to Colombia
(Fairchild & Burger 1994). In Mexico, has been
collected in Chiapas, Tabasco and Veracruz (Fair-
child 1940). Only one specimen has been collected
from PY in Quintana Roo (Sian Ka'an) (Ibanez-
Bernal 1992).

Leucotabanus canithorax Fairchild, 1941

This species has been reported from Colombia,
Trinidad, Guyana, Panama and Belize (Fairchild
1985; Fairchild and Burger 1994). In Mexico, it
has been reported from Chiapas (Fairchild 1985)
and Quintana Roo (Sian Ka'an) (Ibanez-Bernal
1992; Fairchild & Burger 1994). It is reported
here for the first time for Campeche.
Material Examined: CAMPECHE, Calakmul
Reserve, 1 9 archaeologic zone, 3-V-1997, Malaise

Leucotabanus exaestuans (Linnaeus), 1758

According to Fairchild (1986) and Fairchild &
Burger (1994), the range of this species appears to
cover the entire Neotropics, from Mexico to Ar-
gentina, and Trinidad. This species was reported
from Yucatan (Chichen Itza) by Bequert
(Bequaert 1931, 1933; Pearse 1945) (as Tabanus
leucaspis Wiedemann, 1928). It is reported here
for the first time for Campeche.
Material Examined: CAMPECHE, Calakmul
Reserve, 1 9 El Refugio, 2-IX-1993, Malaise trap.
YUCATAN, Reserva de Celestun, 1 9 Celestun
(P), 30-VI-1997, Malaise trap.

Leucotabanus itzarum (Bequaert), 1831

This is apparently an endemic species for Mex-
ico, specifically PY (Fairchild and Burger 1994).
The species has been reported in all the states of
PY (Bequaert 1931, 1933; Pearse 1945; Fairchild
1971; Fairchild 1985; Ibanez-Bernal 1992; Fair-
child & Burger 1994). It seems to be restricted to
deciduous tropical forests.
Material Examined: YUCATAN, Celestun Re-
serve, 1 9 Rancho Loma Bonita (DTF), 15-VII-
1996; Cuxtal Reserve, 1 9 Tunkas, 19-1-1996, net,
11 9 Xmatkuil (DTF), 8-10-VI-1994, Malaise
trap, Xmatkuil (DTF), 1 9 10-17-III-1997, Mal-
aise trap, 7 9 Ibid. 6-12-V-1997, 11 9 Ibid. 13-20-
V-1997, 5 9 Ibid. 20-28-V-1997, 4 9 Ibid. 18-22-
VI-1996, 7 9 Ibid. 28-V-2-VI-1997, 43 9 Ibid. 4-8-
VI-1996, 6 9 Ibid. 10-VI-1997, 15 9 Ibid. 16-VI-
1996, 46 9 Ibid. 18-28-VI-1996, 1 9 Ibid. 18-VI-5-
VII-1996, 3 9 Ibid. 24-VI-1-VII-1997, 1 9 Ibid.
26-VI-5-VII-1996, 9 9 Ibid. 1-24-VII-1997, 1 9

Ibid. 5-16-VII-1996, 2 9 Ibid. 29-VII-1996, 1 9
Ibid. 29-IX-4-X-1996, 1 9 Ibid. 28-X-4-XI-1996.

Lepiselaga crassipes (Fabricius), 1805

Widely distributed south American species
that occurs from southern Mexico to northern Ar-
gentina and Cuba, Jamaica, Hispaniola, Puerto
Rico (Fairchild 1942b; 1986; Cruz and Garcia
1974; Fairchild and Burger 1994). It was collected
in Campeche, Quintana Roo and Yucatan, the
first specific record for PY, and its three states. All
localities are disturbed areas.
Material Examined: CAMPECHE, Calakmul
Reserve, 1 9 El Refugio, 2-IX-1993, Malaise trap.
QUINTANA ROO, 4 9 Vallehermoso, 24-VII-
1993, Malaise trap. YUCATAN, 1 9 Rancho Ho-
bonil, 10-V-1995, Malaise trap.

Tribe Tabanini
Tabanus campechianus Townsend, 1897

Known Distribution: Mexico to Belize (Fair-
child & Burger 1994). In Mexico it has been re-
ported from Campeche (as between Campeche and
Esperanza, 48 mi. N Puerto Real and Campeche)
(Townsend 1897; Bequaert 1931; Pearse 1945;
Fairchild 1978) and Quintana Roo (Cancun and
Sian Ka'an) (Fairchild 1978; Ibanez-Bernal 1992).
It is reported here for the first time for Yucatan. It
seems to be widely distributed in the coastal areas
of PY, although at very low densities.
Material Examined: YUCATAN, Celestun Re-
serve, 1 9 Ecoparaiso (M), 16-X-1996, Malaise
trap, 1 9 Rancho Loma Bonita (P), 25-III-1996, 1
9 Rancho Loma Bonita (DTF), 23-X-1996; Ria
Lagartos Reserve, 15 9 Ojo de Agua (P), 31-VIII-
1995, Malaise trap, 1 9 Rio Lagartos, Peten
Tucha, 28-XI-1995, Malaise trap, 1 9 El Cuyo
(M), 1 9 Ibid. 5-V-1994, Malaise trap, 11-VII-
1996, Malaise trap, 6 9 Ibid. 30-VIII-1995, 2 9
idem 16-X-1996, 1 9 Ibid. 28-XI-1995, 5 9 San
Felipe, 26-V-1995, net; 1 9 Chuburna Puerto
(CD), 15-V-1995, net, 1 9 Ibid. 15-IV-1995.

Tabanus colombensis Macquart, 1846

Known Distribution: Texas, USA to Brazil and
Trinidad (Fairchild & Burger 1994). In Mexico it
has been reported for Veracruz (Misantla) (Willis-
ton 1901) and for PY only in Quintana Roo (Sian
Ka'an) (Ibanez-Bernal 1992). It is reported here
for the first time for Campeche and Yucatan.
Material Examined: CAMPECHE, 1 Y La Lib-
ertad, 24-VI-1993, net. YUCATAN, Celestun Re-
serve, 1 Y Celestun (CD), 23-1-1997, 6 Y Ibid. 27-
III-1996, 6 Y Ibid. 4-IX-1995, 2 Y Ibid. 4-XII-
1995, 4 Y Dumac (M), 28-III-1996, 1 Y Ibid. 3-
VII-1997, 3 Y Ecoparaiso (CD), 21-1-1997, 2 Y
Ibid. 17-VII-1996, 6 Y Ibid. 26-X-1996, 1 Y Ibid.
7-XII-1995, 1 Y Ecoparaiso (M), 9-IV-1997, 6 Y

September 2001

Manrique-Saide et al.: Horseflies from the Yucatan

Ibid. 18-VII-1996, 1 9 Celestun (P), 20-1-1997, 2
9 Ibid. 30-VI-1997, 16 9 Rancho Loma Bonita
(P), 23-VI-1996, 11 9 Ibid. 7-IX-1995, 9 9 Rancho
Loma Bonita (DTF), 26-III-1996, 1 9 Ibid. 7-IV-
1997, 18 9 Ibid. 16-VII-1996, 4 9 Ibid. 5-IX-1995,
2 Ibid. 23-X-1996, 10 9 Ibid. 5-XII-1995; Cuxtal
Reserve, 1 9 Xmatkuil, 24-28-II-1994, Malaise
trap; Dzilam Reserve, 1 9 Dzilam, 5-V-1993, net;
Ria Lagartos Reserve, 2 9 Ojo de Agua (P), 16-I-
1997, 1 9 Ibid. 21-III-1996, 3 9 Ibid. 7-VII-1997,
20 9 Ibid. 31-VIII-1995, 1 9 Ibid. 4-IX-1995, 2 9
Ibid. 14-X-1996, 1 9 Ibid. 30-XI-1995, 3 9 Ibid. 1-
XII-1995, 3 9 Peten Tucha, 4-IV-1997, 3 9 Ibid. 8-
VII-1996, 2 9 Ibid. 1-IX-1995, 1 9 El Cuyo (M),
14-1-1997, 2 9 Ibid. 20-III-1996, 2 9 Ibid. 11-VII-
1997, 3 9 Ibid. 16-X-1996, 1 9 Ibid. 28-XI-1995, 1
9 Entrada a Sac-Bo (DTF), 19-III-1996, 2 9 Ibid.
9-VII-1997, 1 9 Ibid. 29-VIII-1995, 2 9 Ibid. 15-
X-1996, 2 9 La Darcena (DC), 18-III-1996, 9 1
Ibid. 28-VIII-1995, 3 9 Ibid. 17-X-1996.

Tabanus commixtus Walker, 1860

Known Distribution: southern Mexico to Vene-
zuela, Hispaniola, Trinidad and Martinique
(Fairchild 1983; Fairchild & Burger 1994). This
species has been reported for PY in Yucatan
(Chichen Itza) (as T maya Bequaert 1931)
(Bequaert 1931, 1933; Pearse 1945) and for Quin-
tana Roo (Sian Ka'an) (Ibanez-Bernal 1992). It is
reported here for the first time for Campeche.
Material Examined: CAMPECHE, Calakmul
Reserve, 1 9 archaeologic zone, 3-V-1997, Malaise
trap, 1340 9 Ibid. 3-V-1997, 6 9 Ejido Nuevo Be-
can, El Chorro, 30-IV-1997, Malaise trap, 1 9 El
Refugio, 2-1-1993, Malaise trap; 3 9 La Libertad,
24-VI-1993. QUINTANA ROO, El Eden Reserve, 1
9 El Eden, 18-III-1996, net; 2 9 Vallehermoso, 21-
24-VII-1993, Malaise trap. YUCATAN, Celestun
Reserve, 16 9 Celestun (CD), 27-III-1996, 3 9 Ibid.
4-IX-1995, 59 9 Ibid. 7-XII-1995, 3 9 Dumac (M),
22-1-1997, 24 9 Ibid. 28-III-1996, 1 9 Ibid. 6-IX-
1995, 19 9 Ibid. 6-XII-1995, 2 9 Ecoparaiso (CD),
21-1-1997, 2 9 Ibid. 26-X-1996, 13 9 Rancho Loma
Bonita (P), 25-III-1996, 12 9 Ibid. 7-IX-1995, 4 9
Ibid. 4-XII-1995, 38 9 Rancho Loma Bonita (DTF),
26-III-1996, 12 9 Ibid. 16-VII-1996, 8 9 Ibid. 5-IX-
1995, 1 9 Ibid. 23-X-1996, 102 9 Ibid. 5-XII-1995;
Dzilam Reserve, 2 9 Dzilam, 20-1-1993, 1 9 Ibid.
2-II-1993, 14 9 Ibid. 5-V-1993; Cuxtal Reserve, 1 9
Xmatkuil (DTF), 4-10-III-1997, Malaise trap, 1
Ibid. 10-17-III-1997, 2 9 Ibid. 24-31-III-1997, 6 9
Ibid. 6-12-V-1997, 3 9 Ibid. 13-20-V-1996, 4 9 Ibid.
20-28-V-1997, 1 9 Ibid. 28-V-2-VI-1997, 12 9 Ibid.
4-8-VI-1996, 4 9 Ibid. 8-10-VI-1994, 9 9 Ibid. 18-
22-VI-1996, 1 9 Ibid. 18-28-VI-1996; 6 9 Mococha,
4-11-XI-1993, Malaise trap; 2 9 Muna, 27-IV-1995,
net; Ria Lagartos Reserve, 3 9 El Cuyo (M), 5-V-
1994, Malaise trap, 3 9 Ibid. 5-VI-1994, 1 9 Ibid.
11-VII-1996, 10 9 1 6 Ibid. 29-XI-1995, 8 9 En-
trada a Sac-Bo (DTF), 19-III-1996, 1 9 Ibid. 15-X-

1996, 6 Y Ibid. 28-XI-1995, 1 Y La Darcena (CD),
23-1-1997, 4 Y Ibid. 18-III-1996, 1 Y Ibid. 27-XI-
1995, 1 Y Ojo de Agua (P), 31-VIII-1995, 2 Y Ibid.
30-XI-1995, 3 Y Ibid. 1-XII-1995, 4 Y Peten Tucha
(P), 21-III-1996, 4 Y Ibid. 11-VII-1996, 25 Y 2 2
Rio Lagartos, 6-III-1994, net; 39 Y Tzucacab, 10-V-
1995, Malaise trap.

Tabanus haemagogus Williston, 1901

The type locality of this species is Temax,
Yucatan. Known distribution Mexico (Tabasco) to
Guatemala (Fairchild 1971; Fairchild & Burger
1994). Within PY, it has been reported for Yucatan
(Chichen Itza, Izamal, Merida, Temax and Tohil)
(as T filiolus Williston 1901) (Williston 1901;
Bequaert 1931, 1933; Pearse 1938, 1945). It is re-
ported here for the first time for Quintana Roo. It
probably also occurs in Campeche.
Material Examined: QUINTANA ROO, El
Eden Reserve, 2 Y El Eden, 18-III-1996.
YUCATAN, Celestun Reserve, 7 Y Rancho Loma
Bonita (P), 16-VII-1996, 2 Y Ibid. 7-IX-1995, 4
Rancho Loma Bonita (DTF), 16-VII-1996; Cuxtal
Reserve, 2 Y Xmatkuil (DTF), 10-17-III-1997, Mal-
aise trap, 5 Y Ibid. 4-8-V-1996, 1 Y Ibid. 16-V-
1997, 2 28-V-2-VI-1997, 1 Y Ibid. 1-24-VI-1997, 2
Y Ibid. 4-8-VI-1996, 2 Y Ibid. 5-15-VI-1996, 2 Y
Ibid. 5-16-VI-1996, 4 Y Ibid. 10-VI-1997, 7 Y Ibid.
17-VI-1997, 20 Y 1 S Ibid. 18-22-VI-1996, 2 Y
Ibid. 18-24-VI-1997, 38 Y 4 S Ibid. 18-22-VI-1996,
11 Y 2 S Ibid. 18-28-VI-1996, 1 Y Ibid. 28-VI-5-
VII-1996, 4 Y Ibid. 18-VI-5-VII-1996, 1 Y 1 S Ibid.
1-15-VII-1997, 1 Y Ibid. 5-15-VII-1996, 1 Y Ibid.
14-VII-1993, 2 Y Ibid. 1-24-VII-1997, 1 Y 1 S Ibid.
24-VII-4-VIII-1996, net, 44 Y Ibid. 24-VII-4-VIII-
1996, 3 Y 4 S Ibid. 29-VII-4-VIII-1996, net, 3 Y
Ibid. 29-VII-1996, 30 Y 2 S 1-VIII-29-IX-1996, 1 Y
Ibid. 2-IX-1992, net, 7 Y Ibid. 17-25-IX-1996, 4 Y
Ibid. 29-VIII-4-IX-1996, 3 Y Ibid. 17-25-IX-1996, 2
Y Ibid. 7-14-X-1996, 5 Y Ibid. 9-16-X-1996, 1 Y
Ibid. 10-X-1992, 2 Y Ibid. 11-18-X-1996, 1 Y 14-21-
X-1996, 1 Y Ibid. 21-28-X-1996, 1 Y 26-30-X-1996,
6 Y 3 S Ibid. 28-X-4-XI-1996, 1 Y Ibid. 24-28-XI-
1994, 1 Y Ibid. 26-30-XI-1996; Ria Lagartos Re-
serve, 1 Y Ojo de Agua (P), 1-XII-1995, 3 Y Peten
Tucha, 11-VII-1996, 3 Y Entrada a Sac-Bo (DTF),
9-VII-1996; 15 Y Tzucacab, 23-IX-1994, net.

Tabanus occidentalis var. dorsovittatus Macquart, 1855

Known Distribution: Mexico to Argentina and
Trinidad (Fairchild 1971; Fairchild & Burger
1994). Within PY it has been reported only for
Quintana Roo (Sian Ka'an) (Ibanez-Bernal 1992).
It is here reported for the first time for Campeche
and Yucatan.
Material Examined: CAMPECHE, Calakmul
Reserve, 2 Y archaeologic zone, 3-V-1997, Malaise
trap, 4 9 El Refugio, 2-1-1993, Malaise trap.
YUCATAN, 2 9 Celestun (CD), 23-1-1997, 5 Y

Florida Entomologist 84(3)

Ibid. 27-III-1996, 1 Y Ibid. 4-IX-1995, 6 Y Ibid. 7-
XII-1995, 1 Y Dumac (M), 22-1-1997, 1 Y Ibid. 28-
III-1996, 1 Y Ibid. 3-VII-1997, 1 Y Ibid. 6-IX-
1996, 1 Y Ibid. 6-XII-1995, 35 Y Ecoparaiso (CD),
21-1-1997, 2 Y Ibid. 17-VII-1996, 3 Y Ibid. 26-X-
1996, 2 Y Ecoparaiso (M), 18-VII-1996, 2 Y Peten
2, 20-1-1997, 2 Y Ibid. 8-IV-1997, 123 Y Rancho
Loma Bonita (P), 25-III-1996, 1 Y Ibid. 8-IV-1997,
51 Y Ibid. 7-IX-1995, 3 Y Ibid. 4-XII-1995, 60 Y
Rancho Loma Bonita (DTF), 20-III-1996, 5 Y Ibid.
7-IV-1997, 168 Y Ibid. 16-VII-1996, 33 Y Ibid. 5-
IX-1995, 22 Y Ibid. 23-X-1996, 38 Y Ibid. 5-XII-
1995; Cuxtal Reserve, 1 Y Xmatkuil, 24-VII-4-
VIII-1996, net; Dzilam Reserve, 1 Y Dzilam, 2-I-
1993, 1 Y Ibid. 20-1-1993; Ria Lagartos Reserve, 4
Y Ojo de Agua (P), 16-1-1997, 1 Y Ibid. 21-III-
1996, 10 Y Ibid. 8-IV-1997, 41 Y Ibid. 7-VII-1997,
5 Y Ibid. 31-VIII-1995,23 Y Ibid. 14-X-1996, 13 Y
Ibid. 30-XI-1995, 26 Y Ibid. 1-XII-1995, 1 Y Peten
Tucha, 21-111-1996,4 Y Ibid. 4-IV-1997, 45 Y Ibid.
8-VII-1996, 4 Y Ibid. 1-IX-1995, 1 Ibid. 1 -XII-
1995, 17 El Cuyo (M), 11-VII-1996, 12 Ibid. 16-X-
1996,22 Y Ibid. 29-XI-1995, 3 Y Entrada a Sac-Bo
(DTF), 19-III-1996, 1 Y La Darcena (CD), 18-III-
1996, 2 Y Ibid. 12-VII-1996, 15 Y Ibid. 17-X-1996.

Tabanus oculus Walker, 1848

Known Distribution: Northern Mexico (Tampico,
Tamaulipas) to Panama (Fairchild 1971; Fair-
child & Burger 1994). Within PY, it has been re-
ported from Yucatan (Chichen Itza) (Bequaert
1931, 1933; Pearse 1945) and Quintana Roo (Co-
lonia Santa Maria, Sian Ka'an and southeast of
Peto) (Bequaert 1931; Pearse 1945; Ibaniez-Ber-
nal 1992). Pearse (1945) erroneously cited Colo-
nia Santa Maria, which is near to Puerto Morelos,
as Puerto Morelos. It is reported here for the first
time for Campeche.
Material Examined: CAMPECHE, 17 9 La
Libertad, 24-VI-1993, net; Calakmul Reserve, 5 9
El Refugio, 2-IX-1993, Malaise trap, 16 9 Ibid.
11-IX-1993, 24 9 archaeologic area, 3-V-1997,
Malaise trap. QUINTANA ROO, 13 9 Valleher-
moso, 21-VII-1993. YUCATAN, Celestun Reserve,
2 9 Rancho Loma Bonita (P), 26-III-1996, 2 9
Ibid. 7-IX-1995, 19 Ibid. 4-XII-1995, 2 9 Rancho
Loma Bonita (DTF), 23-III-1996, 1 9 Ibid. 16-VII-
1996, 2 9 Ibid. 5-IX-1995, 5 9 Ibid. 23-X-1996, 4
9 Ibid. 5-XII-1995; Cuxtal Reserve, 1 9 Xmatkuil
(DTF), 8-10-VI-1994, Malaise trap, 1 9 Ibid. 8-28-
VI-1996; Ria Lagartos Reserve, 19 Ojo de Agua
(P), 7-VII-1997, 2 9 Ibid. 16-X-1996, 6 9 Ibid. 1-
XII-1995, 1 9 Peten Tucha (P), 7-VII-1997.

Tabanus pungens Wiedemann, 1828

Known Distribution: Texas to northern Argen-
tina, Trinidad, but not West Indies nor Chile (Cos-
car6n 1979; Fairchild 1986; Fairchild & Burger
1994). This species was erroneously described

from Mexico by Bellardi (as T sallei Bellardi 1859
and T propinquus Bellardi 1859). No specific data
about its locality were given. It is reported here
for the first time for PY (Yucatan).
Material Examined: YUCATAN, Celestun Re-
serve, 2 9 Rancho Loma Bonita (DTF), 1-VII-1997;
Ria Lagartos Reserve, 3 9 Ojo de Agua (P), 31-VIII-
1995, 1 9 Peten Tucha, 21-III-1996, 1 9 Ibid. 8-VII-
1996, 1 9 Ibid. 1-IX-1995, 1 9 Ibid. 31-XI-1995, 1 9
Rio Lagartos (DTF), 13-1-1997, Malaise trap.

Tabanus vittiger ssp. guatemalanus Hine, 1906

Known Distribution: USA (Florida) to Brazil,
Bahamas, Cuba, Cayman Islands, Jamaica, Pu-
erto Rico and Galapagos Islands (Cruz & Garcia
1974; Fairchild 1978; Fairchild & Burger 1994).
In Mexico, its known distribution is apparently
restricted to the southeast. Within PY, it had been
reported for Campeche (Philip 1954b) and Quin-
tana Roo (Sian Ka'an) (Ibianez-Bernal 1992). In
1983, Fairchild wrote that this species occurs at
least to Yucatan and Campeche, without specific
records. It is reported here, probably for the first
time for Yucatan.
Material Examined: YUCATAN, Celestun Re-
serve, 49 9 Celestun (CD), 22-1-1997, 27 9 Ibid.
27-III-1996, 75 9 1 6 Ibid. 4-IX-1995, 57 9 Ibid. 7-
XII-1995, 29 9 Dumac (M), 23-1-1997, 7 9 Ibid. 28-
III-1996, 1 9 Ibid. 19-VII-1996, 11 9 Ibid. 6-IX-
1995, 14 9 Ibid. 6-XII-1995, 2 9 Celestun (P), 30-
VI-1997; 1 9 Ecoparaiso (CD), 21-1-1997, 8 9 Ibid.
17-VII-1996, 22 9 Ibid. 26-X-1996, 5 9 Ecoparaiso
(M), 9-IV-1997, 30 9 Ibid. 18-VII-1996, 10 9 Ibid.
25-X-1996, 8 9 Rancho Loma Bonita (P), 26-III-
1996, 23 9 Ibid. 7-IX-1995, 1 9 Ibid. 24-X-1996, 2
9 Ibid. 4-XII-1995, 2 9 Rancho Loma Bonita
(DTF), 21-1-1997, 49 Ibid. 26-111-1996, 4 9 Ibid. 7-
IV-1997, 37 9 Ibid. 17-VII-1996, 19 9 1 6 Ibid. 4-
5-IX-1995, 1 9 Ibid. 23-X-1996, 49 Ibid. 5-XII-
1995; 3 9 Chuburna Puerto, 15-IV-1995, Malaise
trap; Dzilam Reserve, 2 9 Dzilam, 2-II-1993, net;
Ria Lagartos Reserve, 1 9 Ojo de Agua (P), 16-I-
1997, 4 9 Ibid. 7-VII-1997, 71 9 Ibid. 31-VIII-
1995, 11 9 Ibid. 4-IX-1995, 3 9 Ibid. 14-X-1996, 4
9 Ibid. 30-XI-1995, 7 9 Ibid. 4-IV-1997, 3 9 El
Cuyo (M), 14-1-1997, 9 9 Ibid. 11-VII-1996, 2 9
Ibid. 30-VIII-1995, 18 9 Ibid. 16-X-1996, 1 9 Ibid.
28-XI-1995, 16 9 Ibid. 29-XI-1995, 1 9 Entrada a
Sac-Bo (DTF), 13-1-1997, 3 9 Ibid. 19-111-1996,2 9
Ibid. 9-VII-1997, 1 9 Ibid. 29-VIII-1995, 1 9 Ibid.
15-X-1996, 11 9 La Darcena (CD), 11-IV-1997, 1 9
Ibid. 10-VII-1996, 1 9 Ibid. 28-VIII-1995, 5 9 Ibid.
17-X-1996, 1 9 Ibid. 27-XI-1995, 1 9 Peten Tucha,
21 9 Ibid. 8-VII-1996, 6 9 Ibid. 1-IX-1995, 6 9 Rio
Lagartos, 6-III-1994, Malaise trap and net.

Tabanus yucatanus Townsend, 1897

The type locality is in the Cenote of Xcolak, 10
mi. SE of Izamal, Yucatan (Towsend 1897).

September 2001

Manrique-Saide et al.: Horseflies from the Yucatan

Known Distribution: Southeast Mexico to
Guatemala, Honduras, El Salvador, Costa Rica
and Nicaragua. In Mexico the distribution is con-
fined to Chiapas and PY (Fairchild 1971; Fair-
child & Burger 1994). Within PY it has been
collected in Campeche (Fairchild & Burger 1994),
Quintana Roo (Colonia Santa Maria) and Yucatan
(Chichen Itza and Merida) (Bequaert 1931;
Pearse 1945).
Material Examined: YUCATAN, Celestun Re-
serve, 1 Y Rancho Loma Bonita (DTF), 5-IX-1995.


The tabanid fauna of the biotic province of
Yucatan includes a little more than 14% of the
species known for Mexico. Because collecting has
been scattered, knowledge of the tabanid fauna of
PY and Mexico is incomplete. However, we con-
sider that species diversity recorded to date is
similar to the actual species diversity from this
province. The records of most species show wide
distribution within the biotic province of Yucatan,
probably related to homogeneous physiography
(climate, surface, geology and hydrology) which
characterizes the province (Barrera 1962;Alvarez
& de Lachica 1991).
Thus, according to information provided by
Fairchild (1969) and current known distribution
of species, we can consider that most of the ta-
banid fauna of the biotic province of Yucatan is
mainly composed of Neotropical elements. Some
species found are widely distributed in the Anti-
lles (i.e., D. ferrugatus, L. crassipes and S. jamai-
censis), thus suggesting an important input of the
Antillean component to the tabanid fauna of the
biotic province of Yucatan.
The best represented genera, Chrysops and
Tabanus, have a worldwide distribution, and do
not seem to show affinity to any specific biogeo-
graphic pattern. However, 11 of the 15 species of
Chrysops and Tabanus reported show an exclu-
sively Neotropical distribution (amphitropical).
In contrast, C. flavidus, T colombensis, T pun-
gens and T vittiger have Neotropical distributions
but extend to the southern Nearctic Region.
A comparison of the fauna of both biotic prov-
inces in PY is difficult. The biotic province of Pe-
ten is a larger area and ecologically more diverse
than the biotic province of Yucatan. It includes
southern PY, the Mexican states of Tabasco and
Chiapas (north), Belize and the Peten of Guate-
mala (Alvarez and de Lachica 1991). In contrast
to the biotic province of Yucatan, it includes rivers
and higher lands (600-1500 meters).
The biotic province of Peten has a higher spe-
cies diversity. According to the catalogue of Fair-
child & Burger (1994), 85 species have been
reported from this province. At least 23 of the col-
lected species within the biotic province of
Yucatan also have been recorded in the biotic

province of Peten. In addition, we expect that
C. auroguttatus, C. flavidus, L. itzarum, T co-
lombensis and T commixtus could also be distrib-
uted within this province. Because of the few
records of S. pechumani, we cannot speculate
about its probable distribution. We consider that
the number of species reported for Peten is con-
servative, since the Mexican states of Tabasco,
southern Campeche and northern Chiapas have
been poorly sampled. At present, because of the
general distribution of the predominantly Neotro-
pical elements of the tabanid fauna of both biotic
provinces, it is not possible to use the tabanid
fauna to sustain or reinforce the criteria defining
the biotic provinces of Yucatan and Peten. How-
ever, a complete characterization of tabanid
fauna of the Peten Province, including informa-
tion on the unique northern distribution limits of
the species could help to confirm or redefine the
limits of these biotic provinces.
Thus, to obtain a complete characterization of
PY (including the biotic provinces of Yucatan and
Peten), it is necessary to intensify sampling in
non-collected areas (i.e, those with different hab-
itats such as cenotes, rivers, medium and high
tropical forests), and southern PY included in the
northern part of the biotic province of Peten (Al-
varez & de Lachica 1991).
The tabanid fauna of some protected and non-
protected areas of wetlands, cenotes, medium and
high tropical forests remains unknown. For
Campeche and Quintana Roo, most of the records
are from Calakmul and Sian Ka'an, both under-
sampled. It would be desirable to have more sys-
tematic sampling within the southern parts of
both reserves, especially for Calakmul, which is
adjacent to a very important biogeographic Meso-
american corridor that penetrates southern Mex-
ico. Also special efforts should be made in coastal
areas such as the Area de Protecci6n de Flora y
Fauna de Laguna de Terminos (Campeche), Yum
Balam-El Eden and islands such as Cozumel and
Isla Contoy (Quintana Roo).

We greatly indebted to the late Dr. G. B. Fairchild, to
Dr. J. F. Burger, Dr. L. Foil and Dr. J. T. Goodwin for
some useful reprints of references. We are thankful to
Ramiro Rubio (Head of Ria Lagartos), Jos6 Arellano
(Manager of Ecoparaiso) and David Alonzo (DUMAC
southeast Regional coordinator). This study was par-
tially supported by CONABIO (GRANT GO 11) Dipteros
hemat6fagos y taxa relacionados de dos areas protegi-
das del Estado de Yucatan, M6xico. Our compliments to
Juan Carlos Chab for redrawing figure 1 and Alejandra
Gonzalez for her curatorial help.


ALVAREZ, T., AND F. DE LACHICA. 1991. Zoogeografia de
los vertebrados de M6xico. SITESA. M6xico. 65 p.

BARRERA, A. 1962. La Peninsula de Yucatan como provin-
cia bi6tica. Rev. Soc. Mexicana Hist. Nat. 23: 71-105.
BELLARDI, L. 1859. Saggio di ditterologia messicana.
Parte I. Mem. Real. Accad. Sci. Torino (Ser. 2) 19: 1-80.
BEQUAERT, J. 1931. Tabanidae of the Peninsula of
Yucatan, Mexico, with descriptions of new species. J.
New York Entomol. Soc. 39(4): 533-553.
BEQUAERT, J. 1933. Contribution to the entomology of
Yucatan. Carnegie Inst. Washington Publ. 431: 547-
BEQUAERT, J. 1940. The Tabanidae of the Antilles. Rev.
de Entomologia 11(1-2): 253-369.
BEQUAERT, J. 1944. Further studies of the Tabanidae of
Trinidad, B. W. I. Psyche 51(1-2): 12-21.
G. CALDERON, AND J. CARRANZA. 1995. Conservaci6n
de la Cuenca Hidrol6gica Alta de la Bahia Del Es-
piritu Santo, Quintana Roo, M6xico. Sian Ka'an Se-
rie Documentos 3: 1-37.
MOGUEL. 1995. Water resources of the Yucatan pen-
insula, Mexico: Special concerns and management
priorities. Land and water Nov/Dec: 18-20.
COSCARON, S. 1979. Notas sobre tabanidos argentinos.
XV. El g6nero Tabanus Linnaeus. Obra Centenaria
del Museo de la Plata. VI: 251-278.
CRUZ, J., AND A. I. GARCIA. 1974. Los Tabanos (Diptera:
Tabanidae) de Cuba. Poeyana 125: 1-90.
FAIRCHILD, G. B. 1940. Notes on Tabanidae (Dipt.) from
Panama. II. The genus Dichelacera and related gen-
era. Ann. Entomol. Soc. America 33 (4): 685-700
FAIRCHILD, G. B. 1942a. Notes on Tabanidae (Dipt.)
from Panama. III. The genus Chrysops Meigen. Proc.
Entomol. Soc. Washington 44(1): 1-8.
FAIRCHILD, G. B. 1942b. Notes on Tabanidae (Dipt.)
From Panama. IX. The genera Stenotabanus Lutz,
Lepiselaga Macquart and related genera. Ann. Ento-
mol. Soc. America 35(3): 289-309.
FAIRCHILD, G. B. 1953. Notes on neotropical Tabanidae
(Diptera) with descriptions of new species. Ann. En-
tomol. Soc. America 46(2): 259-280.
FAIRCHILD, G. B. 1969. Notes on Neotropical Tabanidae
XII. Classification and distribution, with keys to
genera and subgenera. Arq. Zool. Sao Paulo 17(4):
FAIRCHILD, G. B. 1971. Family Tabanidae. Fasciculo 28:
1-163. In N. Papavero (eds.). A Catalogue of the
Diptera of the Americas South of the United States.
Museo de Zoologia, Sao Paulo.
FAIRCHILD, G. B. 1978. New and little known Florida
Tabanidae. Florida Entomol. 61(3): 121-137.
FAIRCHILD, G. B. 1983. Notes on Neotropical Tabanidae
(Diptera). XIX. The Tabanus lineola complex. Misc.
Publ. Entomol. Soc. America 57: 1-51.
FAIRCHILD, G. B. 1985. Notes on Neotropical Tabanidae
(Diptera) XVIII. The genus Leucotabanus Lutz.
Myia 3: 299-331.
FAIRCHILD, G. B. 1986.The Tabanidae of Panama. Contr.
American Entomol. Inst. 22(3): 1-139.

September 2001

FAIRCHILD, G. B., AND J. F. BURGER. 1994. A Catalog of
the Tabanidae (Diptera) of the Americas South of the
United States. Mem. American Entomol. Inst. 55: 1-
FLORES, J. S., AND I. ESPEJEL. 1994. Tipos de vegetacion
de la Peninsula de Yucatan. Etnoflora Yucatanense-
UADY 3: 1-135.
IBANEZ-BERNAL, S. 1991 (1992). Una nueva especie de
Stenotabanus (Aegialomyia) Philip del Caribe Mexi-
cano (Diptera: Tabanidae). Folia Entomol. Mexicana
83: 133-141.
IBANEZ-BERNAL, S. 1992. Tabanidae (Diptera) de Quin-
tana Roo, M6xico, pp. 241-285 In D. Navarro and J.
Robinson (eds.). Diversidad biol6gica en la Reserva
de la Biosfera de Sian Ka'an, Q. Roo, M6xico. Vol. 2.
IBANEZ, S., 0. CANUL, 0., AND J. CAAMAL. 1990. Los
dipteros de la Reserva de la Biosfera de Sian Ka'an.
pp. 307-316. In D. Navarro and J. Robinson (eds.).
Diversidad biol6gica en la Reserva de la Biosfera de
Sian Ka'an, Q. Roo, M6xico. Vol. 1.
IBANEZ, S., AND S. COSCARON. 2000. Tabanidae
(Diptera). pp. 593-606. In J. Llorente, E. Gonzalez
and A. N. Garcia Aldrete (eds.). Biodiversidad, tax-
onomia y biogeografia de M6xico: hacia una sintesis
de su conocimiento. Institute de Biologia UNAM,
CONABIO y Facultad de Ciencias UNAM, M6xico,
Vol. 2.
PEARSE, A. S. 1938. Insects from Yucatan caves. Carn-
egie Inst. Washington Publ. 491: 237-249.
PEARSE, A. S. 1945. La Fauna, pp. 109-271. In Gobierno
del Estado de Yucatan (ed.). Enciclopedia Yucatan-
ense. M6xico. Tomo I.
PHILIP, C. B. 1954a. New North American Tabanidae.
VIII. Notes on and keys to the genera and species of
Pangoniinae exclusive of Chrysops. Rev. Brasileira
Entomol. 2: 13-60.
PHILIP, C. B. 1954b. New North American Tabanidae,
VII Descriptions of Tabanidae from Mexico
(Diptera). American Mus. Vov. 1695: 1-26.
PHILIP, C. B. 1958. New records of Tabanidae in the An-
tilles. Supplemental report. American Mus. Nov.
1921: 1-7.
PHILIP, C. B. 1966. New North American Tabanidae.
XVIII. New species and addenda to a nearctic cata-
log. Ann. Entomol. Soc. America. 59(3): 519-527.
PHILIP, C. B. 1978a. New North American Tabanidae
(Diptera). XXV. The genus Hybomitra and some
other new Tabanidae horseflies in Mexico. Pan-
Pacific Entomol. 54: 107-134.
PHILIP, C. B. 1978b. New North American Tabanidae (In-
secta: Diptera). XXIV. Further comments on certain
Pangoniinae in Mexico with special reference to Esen-
beckia. Proc. California Acad. Sci. 41(14): 345-356.
SEMARNAP. 1996. Program de areas naturales prote-
gidas de M6xico 1995-2000. Secretaria de Medio Am-
biente, Recursos Naturales y Pesca, M6xico.
TOWNSEND, C. H. T. 1897. Diptera from Yucatan and
Campeche. I. Canadian Entomol. 29(8): 197-199.
WILLISTON, S. W. 1901. Supplement (part), pp. 249-264.
In F. D. Godman and 0. Salvin (eds.). Biologia Cen-
trali-americana, UK.

Florida Entomologist 84(3)

Teixeira & Polavarapu: Trapping Immature R. mendax Females


Blueberry and Cranberry Research and Extension Center, Rutgers University
125A Lake Oswego Road, Chatsworth, NJ 08019


We compared the performance of 9 cm diameter green spheres and red spheres and Pherocon
AM traps, all baited with the same mixture of ammonium acetate and protein hydrolysate,
in attracting blueberry maggot flies, Rhagoletis mendax Curran, of different reproductive
maturity stages and sex. We also evaluated the effect of trap placement in relation to bush
canopy (within the top 15 cm of the canopy, or near the base of the bush, 25 cm above ground,
within a row) on attraction to these classes of flies. Results of this study showed that cap-
tures of flies on red or green spheres were better than on Pherocon AM traps, irrespective of
maturity status or sex. Captures of flies were similar among traps placed in the top or the
base of the bush, in the case of small bushes (1.2 m high). Traps placed within the top of the
canopy captured more flies than those at the base, in the case of larger bushes (1.5-2.0 m
high). At both positions, capture patterns were also not dependent on reproductive maturity
or sex. Regression analysis between capture ratio of mature females on Pherocon AM/
Spheres and time revealed a significant inverse relationship, which might have been caused
by differential aging of these traps. These data show that sphere traps capture more imma-
ture flies than Pherocon AM traps, and therefore can be deployed early in the season, when
most flies present are immature females. The combination of more effective sphere traps and
correct placement strategy depending on bush characteristics can further optimize blue-
berry maggot monitoring programs.

Key Words: Rhagoletis mendax, ovarian development, traps, highbush blueberry


Comparamos el desempeno de esferas verdes y esferas rojas de 9 cm de diametro y trampas
Pherocon AM, todas cebadas con la misma mezcla de acetato de amonio y protein "hydro-
lysate", en atraer moscas de arandano, Rhagoletis mendaz Curran, de diferentes etapas de
madurez reproductive y sexo. Tambien evaluamos el efecto de colocaci6n de trampas en re-
laci6n al dosel del arbusto (dentro de los 15 cm superiores del dosel, o cerca de la base del ar-
busto, 25 cm sobre el suelo, dentro de la hilera) en atracci6n a estas classes de moscas. Los
resultados de este studio demostraron que las captures de moscas en esferas rojas o verdes
fueron mejores que en las trampas Pherocon AM, sin respect del estado de madurez o sexo.
Capturas de moscas en esferas rojas o verdes fueron similares entire trampas colocadas en
el tope o la base del arbusto, en caso de arbustos pequenos (1.2 m de altura). Trampas colo-
cadas dentro del tope del dosel capturaron mas moscas que aquellas en la base, en casos de
arbustos mas grandes (1.5-2.0 m de altura). En ambas posiciones, patrons de capture tam-
bien fueron independientes de madurez reproductive o sexo. Analisis de regresi6n entire pro-
medio de capture de hembras maduras en Pherocon AM /esferas y tiempo revelaron una
relaci6n inversa significativa, la cual pudo ser causada por la vejez diferencial de estas tram-
pas. Estos datos demuestran que trampas esfericas capturan mas moscas inmaduras que
trampas Pherocon AM, y por lo tanto pueden ser utilizadas temprano en la temporada,
cuando la mayoria de las moscas presents son hembras inmaduras. La combinaci6n de
trampas esfericas mas efectivas y una estrategia de colocamiento correct dependiendo de
las caracteristicas del arbusto pueden optimizar aun mas los programs de control y segui-
miento de gusanos de arandano.

The blueberry maggot fly, Rhagoletis mendax 1978; Guibord et al. 1985; Vincent & Lareau
Curran, (Diptera: Tephritidae), is considered the 1989; Gaul et al. 1995; Teixeira & Polavarapu
most important pest of commercially grown low 2001). Quarantine regulations are in place to pre-
and highbush blueberries, Vaccinium angustifo- vent blueberry maggot introduction from infested
hum Aiton and V corymbosum L., respectively, in areas in the U.S. and Canada east of the Rocky
the eastern and midwestern United States and Mountains. The Blueberry Maggot Certification
Atlantic provinces of Canada (Prokopy & Coli Program requires growers who want to export

Florida Entomologist 84(3)

fruit from infested areas to non-infested areas in
Canada to choose either an IPM-based blueberry
maggot management program or a calendar-
based spray program (Canadian Food Inspection
Agency 1999). Growers involved in an IPM pro-
gram are required to monitor the presence of
adults using baited Pherocon AM traps, deployed
at least 2 weeks before the earliest expected
emergence. Growers following a calendar spray
program are required to start insecticide applica-
tions beginning 10 d after the first adult capture
in the Pherocon AM traps in the area.
The choice of appropriate trap type is an im-
portant factor that can affect the reliability of
first detection of adults. The model for most of the
research on the roles of visual and olfactory stim-
uli affecting attraction of Rhagoletis flies has
been the sibling species apple maggot fly, R.
pomonella (Walsh). Flies are attracted to yellow
panels because of their hue and reflectance, and
to dark ~7.5 cm diameter spheres because of their
shape and intensity contrast with background
light (Prokopy 1968; 1972). Fly visual preferences
were attributed to the similarity of the reflectance
spectrum of yellow panels with that of foliage,
and of the shape and contrast of spheres to those
of host fruit (Prokopy 1968). Although the origi-
nal host of the apple maggot is the hawthorn
(Crataegus L. spp.), flies show a preference for 7.5
cm diameter spheres, as do other Rhagoletis spe-
cies with smaller host fruits (Prokopy 1977).
Blueberry maggot flies found in the field early
in the season are predominantly immature fe-
males (Lathrop & Nickels 1932). Therefore, the
trap deployed for detecting the onset of adult
emergence is required to be very attractive to im-
mature females. Immature females spend more
time on foliage, looking for food sources, and ma-
ture females on fruit, the site for mating and ovi-
position (Smith & Prokopy 1981; 1982). To the
extent that trap captures reflect fly behavior, it
has been suggested that foliage type traps (Phero-
con AM) should be more attractive to immature
females, and fruit-type traps (spheres) to mature
females (Prokopy 1968; Neilson et al. 1984). Re-
cently, Liburd et al. (1998) showed that sphere
traps, baited with ammonium acetate and protein
hydrolysate, were more attractive to the blue-
berry maggot than Pherocon AM traps, with an
equal amount of bait. However, there are no stud-
ies that compare the attraction of baited spheres
and Pherocon AM traps to different reproductive
maturity stages and sex of blueberry maggot flies.
Captures of Rhagoletis flies are greatly in-
creased if traps are placed in locations where the
range of visual and odor components includes areas
preferred by the insects (Reissig 1975; Drummond
et al. 1984; Liburd et al. 2000). Site characteristics,
like wind exposure and distance to berries, were
found to influence blueberry maggot fly captures in
lowbush blueberries (Gaul et al. 1995). In highbush

blueberries, Liburd et al. (2000) showed that plac-
ing traps 15 cm within the top of the canopy in-
creased captures over the usual placement of 15
cm above the canopy. Immature females, depend-
ing on food or fruit abundance, may be present for
longer periods or be more active in different loca-
tions of the plant canopy than mature females.
These factors, together with trap characteristics,
should be taken into account when deploying
traps to capture immature R. mendax females, by
placing traps in locations where flies spend more
time or are more active.
The objective of this study was to determine
the trap type and placement more attractive to
immature blueberry maggot females. We evalu-
ated the reproductive maturity of female flies cap-
tured on baited Pherocon AM and 9 cm diameter
green spheres and red spheres. These traps were
placed either 15 cm within the top of the canopy,
or near the base of the bush, 25 cm above ground,
on bushes of different sizes.


Research was conducted during 1999 at 2
highbush blueberry fields, Whitesbog and Chats-
worth, located in Burlington County, N.J. The
Whitesbog site consisted of a 0.5-ha planting of
the cultivar 'Elizabeth'. The Chatsworth site was
a 1-ha field planted with the cultivar 'Bluecrop'.
Neither field was sprayed or pruned for several
years. In Chatsworth, bush height was uniformly
about 1.2 m. Bush height in Whitesbog was more
variable, ranging from 1.5 to 2.0 m. Bushes in
Whitesbog were also larger, with dead wood in the
lower portion of the bush.
Traps were hung from metal poles. Two trap
placement strategies relative to bush canopy were
evaluated. Traps were positioned either within
the top 15 cm of the canopy, at a height dependent
on bush height, or near the base of the bush, 25 cm
above ground (from here on, high and low traps,
respectively). In the case of high traps, branches
were pruned so that they did not contact the trap.
We used Pherocon AM yellow boards (Trece, Sali-
nas, CA), 9 cm diameter plastic green spheres and
red spheres (Great Lakes IPM, Vestaburg, MI).
Pherocon AM traps were already baited with am-
monium acetate and protein hydrolysate, pre-
mixed in the adhesive. Spheres were manually
coated with the same average amount of baited
adhesive (13 g) as in Pherocon AM traps (Liburd
et al. 1998), obtained from the same source. Phero-
con AM traps (23 x 28 cm) were hung folded in an
~65o angle, with a V shaped cross-section and the
sticky yellow surface facing out.
At both locations, all 3 types of traps were ar-
ranged in a randomized complete block design.
Main factors were trap type (Pherocon AM, red
spheres, green spheres) and placement (high,
low). Each of the 6 treatments (3 trap types x 2

September 2001

Teixeira & Polavarapu: Trapping Immature R. mendax Females

placements) was replicated 4 times, yielding 24
traps per experiment. Rows were used for blocks,
and traps were 6 m apart. In Whitesbog, blocks
were separated by 10 to 30 m to take advantage of
more homogenous areas. In Chatsworth, blocks
were 8 m apart.
Traps were checked twice weekly, and re-ran-
domized after each inspection. Sampling took
place throughout the blueberry maggot fly active
season, starting when the first flies appeared in
early June, until late July, when the population
began to decline. Traps were changed once, 3
weeks after the beginning of the experiment. Cap-
tured flies were cleaned with Histoclear (National
Diagnostics, Atlanta, GA) and stored in glass vi-
als, in 70% ethanol. Flies were sexed, and females
were dissected to determine the degree of ovarian
maturation. We categorized females as immature
if there was no visible development of ooblasts in
the ovarioles. In the apple maggot fly, this was
shown to correspond to females of less than 7-8
days of age (Dean 1935; Duan & Prokopy 1994).
Previously, females have been classified as ma-
ture or immature based on the presence of at least
1 fully developed egg (Neilson et al. 1984; Rey-
nolds & Prokopy 1997). We used a more conserva-
tive criterion because we wanted to document
differences in trap attraction among recently
emerged and older females.
Capture data were square-root transformed (x
+ 0.5) and analyzed by 2-way analysis of variance
using PROC ANOVA (SAS Institute 1989) with
trap type and placement as main factors. Total
trap captures, immature and mature female, and
male captures were analyzed separately. Means
were separated using Tukey Studentized Range
test (SAS Institute 1989). Differences were con-
sidered significant at the P = 0.05 level.
We calculated the ratios of captures of imma-
ture and mature females, on Pherocon AM/Sphere
traps (mean captures on Pherocon AM/mean cap-
tures on both types of spheres), and on High/Low
traps (pooled mean captures on high traps/low

traps), for each location. Ratios of captures of im-
mature and mature females at each location were
plotted to compare relative trends over time. The
relationship between capture ratios within each
maturity class and days after first sampling was
determined using PROC REG (SAS Institute 1989).


Trap type was a significant factor in captures
of all classes of flies in both locations (P < 0.05),
with the single exception of immature flies in
Whitesbog (Table 1). Trap placement was a signif-
icant factor for all classes of flies in Whitesbog (P
< 0.05), but not significant in Chatsworth (Table
1). There was one significant interaction between
trap type and placement (Whitesbog) where ma-
ture females were captured in relatively lower
numbers in low placed Pherocon AM traps than in
high placed ones (P = 0.016). However, because
captures in different trap types and placement
follow the same general pattern, we did not per-
form a separate 1-way analysis of variance.
Therefore, for all fly categories we present cap-
ture data pooled by either trap type (Table 2) or
placement (Table 3). Throughout the sampling
period, a total of 17,603 flies were captured in
Chatsworth (2,154 immature; 8,765 mature; 6,684
male), and in Whitesbog, 7,169 flies were cap-
tured (457 immature; 4,048 mature; 2,664 male).
Green and red spheres attracted 2-3 fold more
flies than Pherocon AM traps, in Chatsworth (Ta-
ble 2), irrespective of sex or female maturity. In
Whitesbog, both types of spheres captured more
flies in each class, but differences in captures
were not significant (P = 0.077) for immature fe-
males. There were no differences in captures of
flies of any class between green and red spheres
at either location (Table 2).
With respect to trap placement, in Chat-
sworth, flies did not show a preference for traps at
different positions, irrespective of sex or female
maturity (Table 3). In Whitesbog, flies were cap-


Immature Mature Male Total
Effect df F P F P F P F P

Type 2,15 9.85 0.0019 55.59 0.0001 91.39 0.0001 66.00 0.0001
Placement 1,15 1.34 0.26 3.32 0.089 3.86 0.068 0.11 0.74
Type x Placement 2,15 0.95 0.41 1.13 0.35 2.91 0.086 0.73 0.50
Type 2,15 3.06 0.077 52.35 0.0001 36.74 0.0001 69.00 0.0001
Placement 1,15 9.68 0.0071 61.48 0.0001 147.39 0.0001 146.36 0.0001
Type x Placement 2,15 1.53 0.25 6.10 0.016 0.44 0.65 1.65 0.23

Florida Entomologist 84(3)


Mean + SEM no. flies per trap
Type Immature Mature Male Total

Red Sphere 111.6 + 8.6 a 453.3 + 47.3 a 395.0 + 59.5 a 959.4 + 119.6 a
Green Sphere 100.4 + 14.9 a 448.0 + 45.9 a 330.0 + 33.4 a 878.9 + 88.9 a
Pherocon AM 57.3 + 23.3 b 194.4 + 19.0 b 110.5 + 10.8 b 362.1 32.0 b
Red Sphere 19.6 + 3.9 a 206.5 + 20.1 a 136.5 + 29.5 a 362.6 + 49.9 a
Green Sphere 21.5 + 4.8 a 203.3 + 25.0 a 138.0 + 32.5 a 362.8 + 59.5 a
Pherocon AM 16.0 + 5.1 a 96.3 + 22.4 b 58.5 + 15.2 b 170.8 + 39.8 b

Means in the same column followed by the same letter (at each location) are not significantly different (P = 0.05, Tukey Studentized Range Test).

tured in larger numbers in traps placed in the top
of the canopy (Table 3). At this location, high
traps captured 44% more immature females, 2-
fold more mature females, and 3-fold more males.
The ratios of captures of immature females be-
tween Pherocon AM/Spheres and High/Low traps
generally follow a similar trend to those of mature
flies, in the sampling dates where both classes
were captured (Fig. 1). Ratios of captures in
Whitesbog are much more variable than in Chat-
sworth, possibly reflecting increased heterogene-
ity in bush size. However, ratios of captures of
immature and mature females still follow a simi-
lar pattern. Regression analysis revealed that, in
Chatsworth, for mature flies, there was a signifi-
cant inverse relationship (Table 4) between
Pherocon AM/Sphere trap capture ratio and time
(P = 0.031). There was also a significant inverse
relationship between the High/Low trap capture
ratio and time in the case of immature females in
Whitesbog (P = 0.047). No significant relation-
ships (P < 0.05) were found in Whitesbog.


Results of this study show that baited red or
green spheres were better than Pherocon AM
traps in attracting blueberry maggot adults, irre-
spective of sex or female maturity. These results
support the use of baited sphere traps early in the
season, when most flies found in the field are im-
mature females. Trap placement in relation to
bush canopy was also found to influence fly cap-
tures. In case of small bushes (1.2 m), traps placed
15 cm within the top of the canopy were as attrac-
tive as those placed near the base of the bush, 25
cm above ground. Where bushes are larger (1.5 to
2.0 m), with fruit and leaves mostly on the top of
the canopy, flies were captured in larger numbers
on high traps. Liburd et al. (2000) have shown
that total captures of flies on traps placed within
the top 15 cm of the canopy were equal to or better
than traps placed 45 cm above ground for bushes
of size comparable to the large ones in this study
(1.5 to 1.8 m). These data provide information on


Mean + SEM no. flies per trap
Height Immature Mature Male Total

High 81.3 + 9.1 a 341.0 + 41.5 a 301.3 + 53.6 a 723.5 + 100.5 a
Low 98.3 + 19.0 a 389.4 + 53.2 a 255.8 + 41.9 a 743.4 + 111.6 a
High 22.5 + 3.9 a 208.2 + 17.1 a 166.3 + 22.9 a 396.9 + 42.3 a
Low 15.6 + 3.3 b 129.2 + 23.4 b 55.8 + 9.4 b 200.5 + 34.3 b

Means in the same column followed by the same letter (at each location) are not significantly different (P = 0.05, Tukey Studentized Range Test).

September 2001

Teixeira & Polavarapu: Trapping Immature R. mendax Females


J maure

. Q

~ I-


0n .my ...


6 11 4f 17
o a

Sampling Date

Fig. 1. Ratios of captures of immature and mature
blueberry maggot fly females between Pherocon AM/
Sphere traps, and High/Low traps, throughout the sam-
pling period.

the best placement of traps with respect to bush
physical characteristics, for optimizing captures
of immature flies early in the season.
Contrary to previous expectations (Prokopy
1968; Neilson et al. 1984), baited spheres cap-
tured significantly more immature females than
Pherocon AM traps. The different behavioral pref-
erences exhibited by the blueberry maggot fly for
foliage and fruit (Smith & Prokopy 1981; 1982),
may not translate into higher trap captures of im-
mature females on baited Pherocon AM traps
than on sphere traps. Our study suggests that im-
mature females, like mature females, are recep-
tive to the ammonia stimulus emanating from the
bait, and the shape and intensity contrast stimu-
lus from the spheres. There are few studies focus-
ing on relative trap attraction to Rhagoletis flies
regarding sex and reproductive maturity.
Prokopy (1968) evaluated the maturity of female

apple maggot flies captured on unbaited yellow
boards and red spheres. The percentage of imma-
ture females was larger in yellow boards, but the
absolute numbers captured were very similar.
Neilson et al. (1984) captured slightly more blue-
berry maggot immature females on baited Phero-
con AM traps than on 8.5 cm diameter unbaited
red spheres. This is the first study that compares
the attraction of baited sphere and Pherocon AM
traps to different reproductive maturity stages
and sex of blueberry maggot flies.
In this study we found no differences between
the performance of red and green spheres for each
fly category. Liburd et al. (1998) also had found
little difference between red, green, yellow, and
blue spheres. These results seem to indicate that
the blueberry maggot fly has a response to colored
spheres similar to the apple maggot fly, using
shape and intensity contrast to detect traps
(Owens & Prokopy 1986). Differences in back-
ground in relation to red and green spheres were
shown to have a complex influence on apple mag-
got fly captures (Prokopy 1986). Here, differences
in background (sky and foliage for traps placed in
the canopy, foliage and ground for those near the
base of the bush) do not seem to be the major in-
fluence on trap captures.
Physical characteristics of the bush were an
important factor in captures of all classes of flies.
In the case of large bushes, the results are clearly
in favor of placement of traps 15 cm within the top
of the canopy. Low traps placed near tall bushes
were close to dead or non-bearing branches. These
areas are probably not very attractive for flies
looking for either food sources or fruit for oviposi-
tion. Shade from large bushes might have also
contributed to the poor performance of low traps.
Low Pherocon AM performed poorly when com-
pared to high Pherocon AM traps, possibly be-
cause they require flies to go lower to come into
visual contact with the yellow surface (facing
down), in contrast to spheres, which are visible
from all angles. Even though traps placed 25 cm

JULY TO 29 JULY 1999.

Immature Mature
Ratio a b R2 P a b R2 P

Pherocon AM/Sphere 0.75 -0.017 0.60 0.07 0.57 -0.006 0.51 0.03
High/Low 1.59 -0.061 0.67 0.05 0.73 0.008 0.15 0.30
Pherocon AM/Sphere 1.02 -0.024 0.11 0.52 0.47 0.000 0.00 0.98
High/Low 1.82 -0.008 0.00 0.90 1.61 0.013 0.02 0.73

Florida Entomologist 84(3)

above the ground were somewhat hidden, a large
number of flies were still captured in these traps,
probably because of a strong effect of the ammo-
nia bait. Aluja & Prokopy (1993) found that if the
visual stimulus is strong, odor did not increase
the probability of flies finding a fruit, but flies do
increasingly rely on odor to find fruit when the vi-
sual stimulus becomes weaker.
In smaller bushes (1.2 m), capture patterns
were very similar at both 15 cm within the top of
the canopy, or 25 cm above ground. Flies of all cat-
egories were captured in similar numbers at either
trap placement. Gaul et al. (1995) found that plac-
ing traps close to fruit increased captures in low-
bush blueberries. Behavioral observations have
shown that both male and mature females tend to
visit and spend more time on fruit, the site for mat-
ing and oviposition (Smith & Prokopy 1981; 1982).
Lack of preference for high or low traps seems to
indicate that if fruit and foliage is more evenly dis-
tributed, as on smaller bushes, flies do not restrict
their search for food sources or mating and ovipo-
sition sites to any particular area of the canopy.
Sheltered Pherocon AM traps captured more flies
than those exposed to wind (Gaul et al. 1995). This
suggests that factors like wind speed at different
heights of the canopy, light intensity, temperature,
and humidity might also control fly movement and
influence captures on low traps.
The significant inverse relationship between
captures of mature females on Pherocon AM/
Sphere traps and time might be caused by several
factors. First, we classified flies as mature if there
was any visible developing ooblasts in the ovari-
oles. It is possible that flies might experience a
shift in preference towards spheres after reaching
the maturity threshold we established in this
study. Given the synchronized pattern of emer-
gence and maturity of the population in the field,
the slight trend we observed is not what would be
expected if preference were dependent on fly ma-
turity. Second, Pherocon AM and sphere traps
may have different ammonia release rates, be-
cause of differences in surface area causing slower
bait depletion on spheres compared to Pherocon
AM traps. In a previous study, Liburd et al. (2000)
showed that Pherocon AM and sphere traps expe-
rienced a drop in captures at 11 and 40 d, respec-
tively. Finally, attraction to Pherocon AM traps is
strongly dependent on yellow hue and reflectance
(Prokopy 1972). It is possible that decreased
transparence of the adhesive, accumulation of de-
tritus, or pigment decay may affect captures.
Spheres, on the other hand, are attractive because
of shape and light intensity contrast, which does
not change with time. There was also a significant
inverse relationship between captures of imma-
ture females on High/Low with time, in Chat-
sworth. Given the few data points we could use in
the regression, the reasons for the week relation-
ship (P = 0.047) are not clear. In Whitesbog, vari-

ability on trap captures between sampling dates,
caused by increased heterogeneity in bush size,
might have obscured any patterns.
Results of this study indicate that the use of
red or green 9 cm diameter plastic sphere traps,
baited with ammonium acetate and protein hy-
drolysate, can significantly increase captures of
adult blueberry maggot flies, including immature
females, over Pherocon AM traps. In the case of
small bushes, captures of flies on traps placed 15
cm within the top of the canopy are equivalent to
captures on traps placed 25 cm above ground, for
all categories of flies. Smaller bushes are the
norm in commercial blueberry plantings. One ad-
vantage of low placement is that it does not re-
quire pruning of the bush. In case of larger
bushes, placement within the top of the canopy is
better than near the base of the bush. Blueberry
maggot certification programs, either IPM- or cal-
endar-based, that require monitoring of adult
presence can benefit from the increased accuracy
of spray recommendations as a result of using
more effective trap type and placement appropri-
ate to given bush size.


The Portuguese Fulbright Commission, the
"Fundagao para a Ciencia e Tecnologia" (grant # BD/
5757/95 Praxis XXI Program), and Rutgers University
are thanked for providing support. This manuscript is
part of a Ph.D. thesis submitted to the Graduate School
of Rutgers University.


ALUJA, M., AND R. J. PROKOPY. 1993. Host odor and vi-
sual stimulus interaction during intratree host find-
ing behavior of Rhagoletis pomonella flies. J. Chem.
Ecol. 19: 2671-2696.
ments for the import and domestic movement of
fresh blueberry fruits moving from infested areas in
North America to non infested areas in Canada.
CFIA, Nepean, Ontario.
DEAN, R.W. 1935. Anatomy and postpupal development
of the female reproductive system in the apple mag-
got fly, Rhagoletis pomonella Walsh. New York St.
Ag. Exp. Sta. Tech. Bull. 229.
DUAN, J. J., AND R. J. PROKOPY. 1994. Apple maggot fly
response to red sphere traps in relation to fly age
and experience. Ent. Exp. Appl. 73: 279-287.
Comparative efficacy and optimal positioning of
traps for monitoring apple maggot flies (Diptera: Te-
phritidae). Environ. Entomol. 13: 232-235.
CROZIER, AND M. FULLER. 1995. Deployment and
utility of traps for management ofRhagoletis mendax
(Diptera: Tephritidae). J. Econ. Entomol. 88: 134-139.
Note sur 1' aire de distribution de la mouche du
bluet, Rhagoletis mendax (Diptera: Tephritidae), au
Canada. Phytoprotection 66: 63-67.

September 2001

Teixeira & Polavarapu: Trapping Immature R. mendax Females

LATHROP, F. H., AND C. B. NICKELS. 1932. The biology
and control of the blueberry maggot in Washington
County, ME. U.S. Dep. Agric. Tech. Bull. 275.
LAVARAPU. 1998. Effect of trap color, bait, shape, and
orientation in attraction of blueberry maggot (Diptera:
Tephritidae) flies. J. Econ. Entomol. 91: 243-249.
CASAGRANDE. 2000. Effect of trap size, placement, and
age on captures of blueberry maggot flies (Diptera: Te-
phritidae). J. Econ. Entomol. 93: 1452-1458.
1984. Captures of blueberry maggot adults, Rhagoletis
mendax (Diptera: Tephritidae), on Pherocon AM traps
and on tartar red dark sticky spheres in lowbush blue-
berry fields. Canadian Entomol. 116: 113-118.
OWENS, E. D. AND R. J. PROKOPY. 1986. Relationship be-
tween reflectance spectra of host plant surfaces and
visual detection of host fruit by Rhagoletis
pomonella flies. Physiol. Entomol. 11: 297-307
PROKOPY, R. J. 1968. Visual responses of apple maggot
flies, Rhagoletis pomonella (Diptera: Tephritidae):
orchard studies. Entomol. Exp. Appl. 11: 403-422.
PROKOPY, R. J. 1972. Responses of apple maggot flies to
rectangles of different colors and shades. Environ.
Entomol. 1: 720-726.
PROKOPY, R. J. 1977. Attraction of Rhagoletis flies
(Diptera: Tephritidae) to red spheres of different
sizes. Canadian Entomol. 109: 593-596.

PROKOPY, R. J. 1986. Alightment of apple maggot flies in
fruit mimics in relation to contrast against back-
ground. Florida Entomol. 69: 716-721.
PROKOPY, R. J. AND W. M. COLI. 1978. Selective traps
for monitoring Rhagoletis mendax flies. Prot. Ecol. 1:
REISSIG, W. H. 1975. Performance of apple maggot traps
in various apple tree canopy positions. J. Econ. En-
tomol. 68: 534-538.
REYNOLDS, A. H., AND R. J. PROKOPY. 1997. Evaluation
of odor lures for use with red sticky spheres to trap
apple maggot (Diptera: Tephritidae). J. Econ. Ento-
mol. 90: 1655-1660.
SAS INSTITUTE. 1989. SAS/STAT user's guide, version 6,
4th ed., vol. 1. SAS Institute, Cary, NC.
SMITH, D. C. AND R. J. PROKOPY. 1981. Seasonal and di-
urnal activity of Rhagoletis mendax flies in nature.
Ann. Entomol. Soc. America 74: 462-466.
SMITH, D. C. AND R. J. PROKOPY. 1982. Mating behavior
ofRhagoletis mendax (Diptera: Tephritidae) flies in
nature. Ann. Entomol. Soc. America 75: 388-392.
rence of late emerging populations of the blueberry
maggot fly (Diptera: Tephritidae). Canadian Ento-
mol. 133: 239-250.
VINCENT, C., AND M. J. LAREAU. 1989. Update on the
distribution of the blueberry maggot, Rhagoletis
mendax (Diptera: Tephritidae), in Canada. Acta
Hortic. 241: 333-337

Florida Entomologist 84(3)

September 2001


'10923 SW 78th Ave., Miami, FL 33156

2USDA-ARS CQFIR, 2301 S. International Blvd., Weslaco TX 78596


Field trapping ofAnastrepha suspense (Loew) in groves of mamey sapote, Pouteria sapota
(Jacq.), showed that fly populations were present in high numbers in all of the groves used
for the experiments. Fly populations were highest at the beginning and end of the sampling
period. More than 646 fruit of mamey sapote weighing a total of 459.9 kg were exposed to
Caribbean fruit flies either in the laboratory or under natural conditions. In one test in the
laboratory, 9 Caribbean fruit fly larvae were recovered from mamey sapote fruit. All of the
control guava fruit had infestations, some as high as 70 larvae per fruit. In the field tests, no
mamey sapote had infestations of Caribbean fruit flies, either naturally occurring or from
caged infestation tests. Pressure measurements showed that mamey sapotes averaged -80 to
-130 Newtons which is much harder than guavas which averaged -30 Newtons. Magana and
Pantin mamey sapote collected in the field in Florida were not found to be hosts to the Car-
ibbean fruit fly, but laboratory infestation was found to occur.

Key Words: Pouteria sapota, Anastrepha suspense, host status, quarantine


La capture en el campo deAnastrepha suspense (Loew) en arboledas de mamey zapote, Poute-
ria sapota (Jacq.), demostraron que populaciones estaban presents el altas cantidades en to-
das las arboledas usadas para los experiments. Populaciones de mosca fueron mas altas al
comienzo y al terminar del period de muestreo. Mas de 646 frutas de mamey zapote pesando
459.9 Kg. fueron expuestas a la mosca de fruta del Caribe ya sea en el laboratorio o bajo con-
diciones naturales. En una prueba en el laboratorio, pocas cantidades de larvas de mosca del
Caribe fueron recuperadas de frutas de mamey zapote. Todas las frutas control de guayaba
tuvieron infestaciones, algunas tan altas como 70 larvas por fruta. En pruebas de campo, no
hubo infestaciones de mosca del Caribe en mamey zapote, ya sea en condiciones naturales o
en pruebas de infestaci6n en jaula. Medidas de presi6n demostraron que mamey sapotes son
much mas duros que guayabas. Las variedades de mamey zapote Magana y Pantin, criadas
comercialmente en la Florida, no son hospedantes de la mosca de fruta del Caribe.

The mamey or mamey sapote, Pouteria sapota
(Jacq.), is a fruit tree in the family Sapotaceae na-
tive to Central and South America (Morton 1987).
The fruit are large, up to several kg, with a salmon
pink, orange, to deep red flesh and a large central
seed. The dark brown skin or rind is very tough and
resistant to puncture or damage, particularly in
unripe fruit. The mamey sapote is grown commer-
cially on a small acreage (108 ha) in South Florida
with an estimated annual value of $1.5 million
(Balerdi et al. 1996; Lamberts & Crane 1990).
The Caribbean fruit fly, Anastrepha suspense
(Loew), has a wide host range of over 80 species of
fruits (Swanson & Baranowski 1972), but has not
been reported to attack mamey sapote. There are
records of Anastrepha ludens (Loew),Anastrepha
obliqua (Macquart) and Anastrepha serpentina
(Wiedemann) attacking mamey sapote in Central
America (Emmart 1933; Norrbom & Kim 1988).
While there are no records ofA. suspense attack-
ing P. sapota, A. suspense does attack Pouteria

campechiana Baehni (the eggfruit or canistel),
however these fruit have very thin rinds, and are
soft, compared with P. sapota. Some of the data on
these host lists are from laboratory studies, and
the hosts are rarely attacked in the field (Norr-
bom & Foote 1989).
In summary, there is evidence that the Carib-
bean fruit fly attacks close relatives of the mamey
sapote and close relatives of the Caribbean fruit
fly attack the mamey sapote. The purpose of this
research was to determine if the Caribbean fruit
fly will attack the mamey sapote and infest it un-
der field and laboratory conditions, and the rela-
tive severity of any infestations.


Experiment 1, laboratory cage trials

In 1997 laboratory studies were conducted
with several mamey sapote cultivars, Mangana,

Gould: Non-Host Status of Mamey Sapote

Pantin (Key West), Pace and Maya (Mayapan).
Fruit were purchased or collected from groves of
each cultivar in Dade County, Florida from April
10 through July 17, 1997. The fruit from each
grove on each date were divided randomly into 3
groups with equal numbers of fruit in each group.
One group of fruit was held without treatment to
determine if any natural infestations were
present. The other 2 groups of fruit were placed in
cages (1 x 1 x 1 m) with 10 female and 10 male 10-
day-old Caribbean fruit flies. The fruit in 1 of the
treatments were punctured (25 pinholes 2-3 cm
into the fruit) before placement into the fly cage to
allow easier access for ovipositing fruit flies. In
addition to these treatments, for each date that
fruit were collected, 1 cage was prepared with 5
heat-disinfested guavas (35 minutes immersion
in 46C water) exposed to 10 female and 10 male
10-day-old fruit flies as a positive control.
After exposure to fruit flies for 24 h (under a
photoperiod of 14:10 L:D) the fruit were removed
from the cages and held 3 to 4 weeks at about 25C.
Any emerging larvae or pupae were collected and
counted. At the end of the holding period, each
fruit was opened and the pulp inspected for pres-
ence of larvae or pupae before disposal.

Experiment 2, field cage trials

In 1998 field tests were conducted with mamey
sapote cultivars Magana and Pantin. Three coop-
erators were selected for each cultivar and groves
were visited every 2 weeks from April to Septem-
ber. Five mamey sapote fruit were individually
bagged on the tree with 5 mated female fruit flies
for 24 h. A control group of 4 guavas was individ-
ually bagged on the mamey sapote tree with 5 fe-
male fruit flies for 24 h to ensure that the flies
were capable of laying eggs.
The fruit were enclosed in a 45 x 45 cm plastic
bag with many small air holes (Delnet pollination
bag, Applied Extrusion Technologies, Inc., Middle-
town, DE) supplied with water-soaked cotton and a
sugar cube. The bag was secured to the tree and
fruit with wire twist ties and rubber bands. A 23 cm
diameter opaque plastic plate was placed above
each bag to shield the fruit from rain and direct sun.
One group of 5 mamey sapote was collected
and held without treatment to determine if a field
infestation existed. Samples of fruit lying on the
ground were also collected if available (some
groves did not have any fallen fruit) and held to
determine if they were infested. All fruit were
then taken to the laboratory where size, weight,
inner peel color and firmness were recorded.
Four glass McPhail traps were placed in each
grove (1 on each side of the grove) at 3/4 tree height
in the exterior part of the tree canopy and baited
with 10 g of torula yeast plus 300 ml of water. The
traps were monitored for the presence of adult flies
each week that the grove was sampled for fruit.


Experiment 1, laboratory cage trials

A total of 396 mamey weighing a total of 237.6
kg were used in this experiment. Mean cultivar
weights were Magaha 1,019.9 36.3 g, Pantin
718.8 + 28.0 g, Maya 603.4 39.1 g, and Pace
468.7 + 13.3 g. Mamey sapote fruit are very firm;
pressure tests showed that while fruit firmness
declined as the season progressed, it remained
higher than -80 Newtons (the force required to
push a 12 mm diameter cylinder into the fruit 3
mm; larger negative number = harder fruit) (Fig.
1). Guavas, which are primary hosts for Carib-
bean fruit flies, average about -30 Newtons when
mature. Mamey sapotes, even at the end of the
test period, were twice as firm as guavas.
No insects were recovered from any of the fruit
held without treatment, therefore there was no
natural infestation. The control guavas in every
replicate had fruit fly larvae present. Only 1 of
the treated replicates had fruit fly larvae in ma-
mey sapote fruit. In the replication from 29 May
1997, 8 larvae were found from unpunctured Ma-
gana mamey sapotes (3 fruit) exposed to female
fruit flies, and 1 larva was recovered from punc-
tured Magaha mamey sapotes (3 fruit) exposed to
female fruit flies. The guava control (5 fruit) for
that replication had 223 larvae, which was much
higher than the larvae from control fruit in other
replications (Table 1). The fruit from the 29 May
1997 test date showed no physical differences
from fruit used on any of the other test dates.

Experiment 2, field cage trials

A total of 250 mamey sapotes weighing a total
of 173 kg was used in this experiment. Magana
and Pantin were the two cultivars tested (weigh-
ing 882 210 g and 617 93 g, respectively). The
fruit were firm, averaging harder than -100 New-
tons throughout the season (Magaha -119 17,
Pantin -120 + 11) (Fig. 2). The fruit collected from
the ground (3.6 kg Magaha, 46.3 kg Pantin) were
usually too soft to measure firmness, and were of-
ten split open with the pulp exposed. The number
of fallen fruit varied greatly because some grow-
ers harvested all fruit or cleaned up fallen fruit.
Other growers allowed fallen fruit to rot on the
ground under trees.
No larvae were recovered from any of the ma-
mey sapotes tested. No field infestation was found
in either fruit on the trees or fruit recovered from
the ground. No larvae were recovered from any
mamey sapote bagged with female fruit flies. Al-
most all of the guava control fruit became in-
fested, with larval numbers ranging up to 325
larvae recovered from 4 guavas (Table 2). Carib-
bean fruit fly adults were present in all of the
groves used for the experiments. Fly populations

Florida Entomologist 84(3)


U -60

' -80

. -100

C -120
M -140

-160 'I
10 April 17 April 1 May 19 May 29 May 5 June 14 July 17 July
Fig. 1. Firmness of mamey sapote used in laboratory infestation tests in 1997 (the force in Newtons required to
push a 12 mm diameter cylinder into the fruit 3 mm).

Number of larvae found
Fruit exposed to flies in lab
Date (number exposed) Guavas Intact mamey Pinholed mamey Control mamey
10 April Maya (9) 1 0 0 0
Pantin (9) 1 0 0 0
17 April' Magana (2) 0
Pace (2) 0
Pantin (3) 0
17 April Magana (9) 16 0 0 0
Pace (15) 16 0 0 0
1May Magana (9) 5 0 0 0
Pace (15) 5 0 0 0
19 May Magana (9) 5 0 0 0
Pace (15) 5 0 0 0
29 May Magana (9) 223 8 1 0
Pace (15) 223 0 0 0
5 June Pantin (90) 11 0 0 0
18 June Pantin (12) 6 0 0 0
14 July Pantin (12) 14 0 0 0
Maya (12) 14 0 0 0
17 July1 Maya (150) 0
'Fruit held to determine if a natural infestation exists, not exposed to caged flies.



September 2001

i ;

Gould: Non-Host Status of Mamey Sapote


a, -90

o -100

0 -110

.E -120


E -140

u. -150

21 April 16 June 30 June 14 July 28 July 17 Aug. 2 Sept.
Fig. 2. Firmness of mamey sapote used in field infestation tests in 1998 (the force in Newtons required to push
a 12 mm diameter cylinder into the fruit 3 mm).

were highest at the beginning and end of the sam-
pling period (Fig. 3).

The collection of fruit for the laboratory tests
(1997) covered the commercial season for Magana
and the first half of the season for Pantin which
make up 95-98% of the commercial acreage
(Balerdi et al. 1996). Most of these fruit ripen dur-

ing the period April through August. Often a few
fruit are available at other times, however Carib-
bean fruit fly populations are also highest in the
early summer (Hennessey 1994). In the laboratory
study 9 larvae were recovered from mamey sa-
potes in 1 of the replications. The control infesta-
tions were highest at this time in the experiment
(223 larvae). There were no physical differences
found between the fruit that were infested and
those that were not infested. The high oviposition

Number of larvae found (Mean + SEM)
Bagged mamey sapote Untreated control
Cultivar Replication and dates Guava (4 fruit) (5 fruit) mamey sapote (5 fruit)
Magania R1, 21 Apr.-1 May 30.0+ 15.3 0 0
R2, 7 May-13 May 49.3 + 32.3 0 0
R3, 19 May-1 June 9.0 + 9.0 0 0
Pantin R1, 16 June 19.7 + 4.8 0 0
R2, 30 June 279.0 + 46.0 0 0
R3, 14-15 July 218.7 + 45.9 0 0
R4, 28-30 July 170.7 + 11.7 0 0
R5, 17 Aug. 66.3+ 15.9 0 0
R6, 2 Sept. 49.5 + 8.5 0 0

Florida Entomologist 84(3)





0 -

21 April 1 May 13 May 28 May 30 June 29 July 2 Sept.
Fig. 3. Total numbers of adult Caribbean fruit flies trapped in field tests, 1998. Four McPhail traps per grove col-
lected weekly.

pressure in this case may have overwhelmed the
natural barriers to oviposition. The low number of
larvae resulting indicates that the mamey sapote
is a poor host under forced infestation.
The collection of fruit and tests done in the
field (1998) covered the main harvest season for
both varieties tested (Magana and Pantin). Fruit
fly populations were present in the fields, and the
flies in the field cage tests produced thousands of
larvae in the control guavas. No larvae were
found in any of the field-collected fruit samples in
either 1997 or 1998, and the cage tests did not
produce any infestations in mamey sapotes. In
addition none of the fruit which were collected
from the ground, and which were often overripe
and broken open, produced any fruit fly larvae.
Several fruit have been shown to be non-hosts
for the Caribbean fruit fly. Based on Hennessey et
al. (1992) and Nguyen & Fraser (1989), limes are
not hosts to the Caribbean fruit fly. Limes and
other citrus have biochemical defenses which
cause mortality of eggs and small larvae (Greany
et al. 1983). In this study we found that unripe
mamey sapote had a distinct chemical odor and
white latex juice present when cut. This suggests
that chemical defenses to fruit fly infestation

could be present in mamey sapote. Eggs and lar-
vae placed on unripe mamey sapote slices had a
high mortality rate (author unpublished data).
Using very high numbers of ovipositing fruit
flies increases the likelihood of declaring a non-
host a host, since many species of fruit may be in-
fested under the pressure of hundreds of oviposit-
ing females in a confined setting. Under field
conditions it is likely that there are fewer than 5
flies visiting any given fruit.
'Forcing' larvae into ripe fruit under artificial
conditions does not fairly represent field condi-
tions. Therefore, under the protocol of Cowley
et al. (1992), mamey sapote were tested to see
whether they would be infested under caged con-
ditions in the field, as well as collecting fruit for
evidence of a natural infestation.
In the laboratory and field forced infestation
tests, 5 female flies were used. This is probably
much higher than the fly population any given
fruit is naturally exposed to in the field, and it is
more realistic than using hundreds of flies in a
small cage. The only infestation that occurred was
in a single replication in the laboratory and at a
very low rate compared to the control fruit (9 lar-
vae vs. 223 control larvae).

September 2001

Gould: Non-Host Status of Mamey Sapote

In this two year study, a balanced approach
was used which included field and laboratory
tests with as large a sample size as was realistic.
The protocol of finding host status as proposed by
Cowley et al. (1992) was used as a guide. Gould et
al. (1999) found that lychees and longans were
not hosts to Caribbean fruit flies using similar
field and laboratory studies.
A total of more than 646 fruit of Magana, Pan-
tin, Pace, and Maya mamey sapote weighing in
total more than 410 kg were exposed to Carib-
bean fruit flies either in the laboratory or under
forced or natural field conditions. No Caribbean
fruit fly larvae were recovered from any fruit col-
lected in the field or exposed to caged flies in the
field. In addition, no larvae were recovered from
50 kg of fruit (approximately 100 fruit) which was
collected from the ground in the field, and which
would be the most likely to have larvae if an infes-
tation were present in mamey sapote.
Mamey sapote are very firm, normally being
harvested when the fruit is harder than -80 New-
tons. Guavas, one of the best hosts for Caribbean
fruit fly, are much softer averaging -41.2 + 7.1
Newtons (n = 10) when mature green. Guavas
have a very strong odor while unripe mamey sa-
potes are almost odorless. Some fruit odors have
been found to be powerful attractants for fruit
flies (Robacker et al. 1990; Nigg et al. 1994). In
addition there was evidence of chemical defenses
to fruit fly infestation in unripe mamey sapotes.
Based on the protocol put forth by Cowley et al.
(1992) and extensive laboratory and field tests,
Magana and Pantin mamey sapotes are not hosts
to the Caribbean fruit fly in commercial mamey
sapote groves and present no risk of transporting
A. suspense from Florida to other locations.


We thank P. Mendez, W. Montgomery, and P.
Shorb of the USDA-ARS for their assistance. We
thank the Tropical Fruit Growers of South Flor-
ida, Inc. for supporting this study. We would like
to thank the following for providing fruit and ac-
cess to orchards: M. Ellenby, E. Grenet, Dr. A. Mo-
rales, C. Morojon, R. Perez, S. Sentelli. We also
thank Dr. R. Schnell of USDA-ARS and Dr. J.
Crane, Tropical Research and Education Center,
for sharing mamey sapote fruit from their genetic
study for this work.


1996. The mamey sapote. FC-30. Horticultural Sci-

ences Dept., Florida Cooperative Extension Service,
Institute of Food and Agricultural Sciences, Univer-
sity of Florida. 8 pp.
COWLEY, J. M., R. T. BAKER, AND D. S. HARTE. 1992.
Definition and determination of host status for mul-
tivoltine fruit fly (Diptera: Tephritidae) species. J.
Econ. Entomol. 85: 312-317.
EMMART, E. W. 1933. The eggs of four species of fruit
flies of the genus Anastrepha. Proc. Entomol. Soc.
Washington. 35: 184-191.
RAS, R. NGUYEN, AND J. CRANE. 1999. Nonhost sta-
tus of lychees and longans to Caribbean fruit fly
(Diptera: Tephritidae). J. Econ. Entomol. 92: 1212-
AND D. L. CHAMBERS. 1983. Biochemical resistance
of citrus to fruit flies. Demonstration and elucidation
of resistance to the Caribbean fruit fly, Anastrepha
suspense. Ent. Exp. & appl. 34: 40-50.
1992. Absence of natural infestation of Caribbean
fruit fly (Diptera: Tephritidae) from commercial
Florida 'Tahiti' lime fruits. J. Econ. Entomol. 85:
HENNESSEY, M. K. 1994. Analysis of Caribbean fruit fly
(Diptera: Tephritidae) trapping data, Dade County,
Florida, 1987-1991. Florida Entomol. 77: 126-135.
LAMBERTS, M., AND J. H. CRANE. 1990. Tropical Fruits,
pp. 337- 355. In J. Janick and J. E. Simon [eds.] Ad-
vances in new crops. Timber Press, Portland, OR.
MORTON, J. F. 1987. Fruits of warm climates. J. F. Mor-
ton, publisher, Miami, FL. 505 pp.
NGUYEN, R., AND S. FRASER. 1989. Lack of suitability of
commercial limes and lemons as hosts ofAnastrepha
suspense (Diptera: Tephritidae). Florida Entomol.
72: 718-720.
NATION, AND J. A. ATTAWAY. 1994. Caribbean fruit
fly, Anastrepha suspense (Loew), attraction to host
fruit and host kairomones. J. Chem. Ecology 20: 727-
NORRBOM, A. L., AND K. C. KIM. 1988. A list of the re-
ported host plants of the species of Anastrepha
(Diptera: Tephritidae). APHIS 81-52, Washington,
NORRBOM, A. L., AND R. H. FOOTE. 1989. The taxonomy
and zoogeography of the genus Anastrepha (Diptera:
Tephritidae), pp. 15-26. In A. S. Robinson and G.
Hooper (eds.), Fruit flies. Their biology, natural ene-
mies and control. Elsevier, Amsterdam.
Attraction of a laboratory strain of Anastrepha
ludens (Diptera: Tephritidae) to the odor of fer-
mented chapote fruit and to pheromones in labora-
tory experiments. Environ. Entomol. 19: 403-408.
range and infestation by the Caribbean fruit fly,
Anastrepha suspense (Diptera: Tephritidae), in
south Florida. Proc. Florida State Hort. Soc. 85: 271-

Florida Entomologist 84(3)

September 2001


1Florida Medical Entomology Laboratory, IFAS/Univ. of Florida, 200 9th Street, S.E., Vero Beach, FL 32962

2Division of Vector-Borne Infectious Diseases, National Center for Infectious Diseases
Centers for Disease Control and Prevention, Public Health Service
U. S. Department of Health and Human Services, P.O. Box 2087, Fort Collins, CO 80522


Adult mosquitoes were collected for virus isolation from two sites in Indian River County,
FL, from May 5 through August 13, 1998 using dry ice-baited CDC light traps (81 trap-
nights) and CDC gravid traps (254 trap-nights). A total of 46,150 female mosquitoes (923
mosquito-pools, 50 females/pool) were processed for virus isolation. These females repre-
sented 18 species of mosquitoes, with Culex nigripalpus comprising 77.4% of all mosquitoes
collected, followed byAedes infirmatus (4.9%),Ae. vexans (4.0%) and Cx. erraticus (2.4%). No
St. Louis encephalitis (SLE) and eastern equine encephalitis (EEE) virus isolates were ob-
tained. Keystone (KEY) and Tensaw (TEN) viruses were isolated fromAe. albopictus (one
isolate of KEY); Anopheles crucians (two isolates of TEN); Cx. nigripalpus (one isolate of
TEN and 2 isolates of KEY); Coquilletidia perturbans (two isolates of TEN); and Wyeomyia
vanduzeei (one isolate of TEN). All isolates were obtained from mosquitoes collected in CDC
light traps, except for the KEY virus isolate from Ae. albopictus, which was collected in a
CDC gravid trap. The isolation of TEN virus from Wy. vanduzeei is a first record for Florida.

Key Words: Arbovirues, Tensaw, Keystone, mosquitoes, Culex nigripalpus,Aedes albopictus,
Coquilletidia perturbans, Anopheles crucians, Wyeomyia vanduzeei

Mosquitos adults fueron colectados para aislamiento de virus de dos sitios en el Condado de
Indian River, FL, desde mayo 5 hasta agosto 13, 1998, usando trampas de luz CDC cebadas
con hielo seco (81 noches de trampa) y trampas CDC gravidas (254 noches de trampas). Un
total de 46,150 mosquitas (923 grupos de mosquitos, 50 hembras/grupo) fueron procesadas
para aislamiento de virus. Estas hembras representaron 18 species de mosquitos, con Culex
nigripalpus componiendo 77.4% del total, seguido por Aedes infirmatus (4.9%), Ae. vexann
(4.0%). y Cx. erraticus (2.4%). No se obtuvieron virus aislados de encefalitis St. Louis (SLE)
o encefalitis oriental equina (EEE). Los virus Keystone (KEY) y Tensaw (TEN), fueron ais-
lados deAe. albopictus, (un aislado de virus KEY); de Anopheles crucians (dos aislados de vi-
rus TEN); de Cx. nigripalpus (un aislado de TEN y dos aislados de KEY); de Coquillletidia
perturbans (dos aislados de TEN); y de Wyeomyia vanduzeei (un aislado de TEN). Todos los
aislados fueron obtenidos de mosquitos colectados en trampas de luz CDC, except por el ais-
lado del virus KEY de Ae. albopictus, el cual fue colectado en una trampa CDC gravida. El
aislamiento de virus TEN de Wy. vanduzeei es una primera constancia para Florida.

Mosquitoes are vectors and/or hosts of several
arboviruses in Florida. These arboviruses include,
St. Louis encephalitis (SLE), eastern equine en-
cephalitis (EEE), trivittatus (TVT), Flanders
(FLA), Sawgrass (SAW), Tamiami (TAM), Ever-
glades (EVE), Shark River (SR), Jamestown Can-
yon (JC), Highlands (HJ), Tensaw (TEN), and
Keystone (KEY) (Chamberlain et al. 1964; Dow et
al. 1964; Wellings et al. 1972; Shroyer 1991; Mitch-
ell et al. 1996). Most of these arboviruses, have been
isolated from counties in Florida other than Indian
River. Only SLE virus was isolated from mosqui-
toes in Indian River County in 1990 (Shroyer,
1991). Recently, Mitchell et al. (1996) isolated EEE,
EVE, KEY, TEN, TVT, SR, and FLA from mosqui-

toes associated with waste-tire piles in counties in
central and north Florida. In 1997, an above aver-
age amount of SLE virus activity in sentinel chick-
ens occurred in Indian River County, Florida,
although no human cases were documented there
(Day & Stark 2000) with continued SLE virus activ-
ity in sentinel chickens in February 1998. There
were no virus isolates from the Culex nigripalpus
Theobald (a proven vector of SLE virus) collected
and tested for SLE virus during 1997 in Indian
River County (Day & Stark 2000). The purpose of
the present study was to isolate and identify arthro-
pod-borne viruses from mosquitoes collected in two
different locations in Indian River County during
the late spring and early summer of 1998.

Nayar et al.: Mosquito arboviruses from Florida


Mosquito collection sites: Mosquitoes were col-
lected from 2 sites in Indian River County. Site 1
was pine woods with scattered cabbage palm, pal-
metto, and oak, 25 Km north of the Florida Medi-
cal Entomology Laboratory (FMEL), Vero Beach;
at this site a sentinel chicken flock was main-
tained by the Indian River Mosquito Control Dis-
trict as part of its SLE virus surveillance
program. The traps were set more than 50 m from
the sentinel chicken flock. Site 2 was an oak ham-
mock on the FMEL grounds.
Mosquito collection and handling methods:
Mosquitoes were collected twice a week from May
5 through August 13, 1998, in 5 CDC gravid mos-
quito traps (Reiter 1983) and 1 or 2 dry ice-baited
CDC miniature light traps. Mosquitoes from the
traps were transported in ice-coolers to the
FMEL, where they were separated on a chill table
and identified to species; 50 females were then
pooled in 2-ml screw-cap cryovials. The cryovials
were labeled and frozen at -80%C until they were
shipped overnight on dry ice to the CDC labora-
tory in Fort Collins, Colorado, for virus isolation
and identification.

Virus Isolations and Virus Identification:

Mosquito pools were triturated in 2 ml of BA-1
diluent by using cold mortars and pestles. BA-1
diluent containing 1x M199 medium with Hanks
balanced salt solution (HBSS), 0.05 M Tris pH
7.6, 1% bovine serum albumin, 0.35 gm/L sodium
bicarbonate, 100 units/ml penicillin, 100 g/ml
streptomycin, 1 g/ml fungizone, and 10 mg/liter
phenol red. Suspensions were centrifuged in Ep-
pendorf tubes at 14,000 rpm for 2 min. Superna-
tants were poured into 1-dram screw-cap vials
and stored at -70C until tested.
Specimens were tested for virus in Vero cell
culture grown in 6-well plates. Specimens were
inoculated in 0.1-ml quantities in 2 wells each
and adsorbed for 1 hr at 37C; the cells were then
overlayed with the first of two nutrient -0.5% aga-
rose overlays. Cell cultures were incubated at
37C and 3 days later a second agarose overlay
containing 1:50,000 neutral red was applied. Cell
cultures were returned to the incubator and ex-
amined daily thereafter for plaques through day
10 postinoculation.
Virus-positive cell cultures were harvested in
2 ml of BA-1 and frozen at -70C until they were
passed into fluid cultures of Vero cells in 25-cm2
flasks. When early cytopathic effects (CPE) were
noted, infected cells were scraped or trypsinized
from the surface of the flask and resuspended in
phosphate-buffered saline (PBS), pH 7.4, contain-
ing 5% Fetal Bovine Serum (FBS). Twelve-well
spot slides were prepared, air-dried, and fixed in
cold acetone. These were tested in an indirect flu-

crescent antibody (IFA) assay (Wulff and Lange
1975) against a battery of hyperimmune grouping
ascitic fluids obtained from NIH and CDC. Usu-
ally, viral type-specific monoclonal antibodies
against common or suspected viruses also were
included in the test to definitively identify iso-
lates at this stage.
Virus-positive cell cultures were also charac-
terized by the reverse transcriptase-PCR (RT-PCR)
by using flavivirus-consensus, SLE-specific, and
bunyavirus serogroup-specific primers (Chang et
al. 1994; Kuno et al. 1996). RT-PCR-positive spec-
imens generated by BCS82C and BCS332V prim-
ers were genetically sequenced by using an ABI
Prism 377 DNA Sequencer (Perkin-Elmer/Applied
Biosystems, Foster City, CA) (Kuno et al. 1996).
Both strands of the cDNA, located between nucle-
otide position of 77 to 273 in the small (S) RNA
segment, were sequenced and compared with the
GenBank data base by using the BLAST search
program (http://www.ncbi.nlm.nih.gov/BLAST/).
Otherwise, antigenically grouped viral isolates
were typed by neutralization (N) assay in Vero
cell cultures against reference polyclonal immune
reagents prepared against specific members of
the antigenic group. Homologous N titers were
predetermined for reference reagents used in the
identifying N tests.


Totals of 54 light trap and 52 gravid trap were
used to collect 81 light and 254 gravid, trap
nights, respectively, at the two sites. A total of
94.9% of the female mosquitoes that were tested
for viruses were captured in light traps while only
5.1% of the female mosquitoes were collected in
gravid traps (Table 1). Mosquitoes in light trap
collections were 80.1% Cx. nigripalpus with other
species comprising less than 5% of the total (Table
1). The main mosquito species in gravid traps
were Aedes albopictus (Skuse), Cx. nigripalpus,
Wyeomyia vanduzeei Dyar and Knab, and Wy.
mitchellii (Theobald). All of the Cx. quinquefas-
ciatus were collected in gravid traps. (Table 1).
No SLE and EEE virus isolates were obtained
from mosquito collected from May 5 through Au-
gust 13,1998, at the 2 sites. Five species of mos-
quitoes yielded 9 isolates of arboviruses belonging
to 2 antigenic groups, Keystone (KEY) virus of the
California antigenic group and Tensaw (TEN) vi-
rus of the Bunyamwera antigenic group (Table 2).
Four species of mosquitoes yielded 6 isolates of
TEN virus: An. crucians (two isolates); Cx. nigri-
palpus (one isolate); Cq. perturbans (two iso-
lates); and Wy. vanduzeei (one isolate). Two
species of mosquitoes yielded 3 KEY virus iso-
lates,Ae. albopictus (one isolate) and Cx. nigripal-
pus (two isolates). All isolates were obtained from
mosquitoes collected in CDC light traps, except
for 1 KEY virus isolate, which was recovered from

Florida Entomologist 84(3)


Site 1 Site 2
Species Total tested Light trap Gravid trap Light trap Gravid trap by species

Ae. albopictus 1,000 50 250 700 2.1
Ae. atlanticus 100 100 0.2
Ae. infirmatus 2,250 100 2,150 4.9
Ae. taeniorhynchus 650 150 500 1.4
Ae. uexans 1,850 1,800 50 4.0
An. crucians 750 500 250 1.6
An. quadrimaculatus 50 50 0.1
Cq. perturbans 550 450 100 1.2
Cx. erraticus 1,100 1,100 2.4
Cx. iolambdis 100 100 0.2
Cx. nigripalpus 35,700 26,450 350 8,650 250 77.4
Cx. quinquefasciatus 900 450 450 2.0
Cx. salinarius 400 350 50 0.9
De. cancer 200 200 0.4
Ma. titillans 50 50 0.1
Ps. ferox 50 50 0.1
Wy. mitchellii 100 50 50 0.2
Wy. vanduzeei 350 300 50 0.8
Total 46,150 31,000 850 12,800 1,500 100.0
Percentage by trap 67.2 1.8 27.7 3.3 100.0
and site

Ae. albopictus specimens collected in a CDC
gravid trap.
The RT-PCR independently confirmed the se-
rological identification that 9 isolates of arbovi-
ruses were not SLE or EEE viruses. Nucleotide
sequences were obtained from 6 TEN virus iso-
lates and 1 KEY virus isolate. All TEN virus iso-
lates had an identical nucleotide sequence in the
S-gene region sequenced. The BLAST search by
using GenBank data base indicated that TEN vi-
rus isolates shared 95.4%, 95.4% and 93.4% nu-
cleotide identity with Northway, Cache Valley,
and Bunyawera viruses, respectively (Table 3).

Keystone virus isolate FL-1290 shared 99.5%,
93.4%, and 92.3% nucleotide sequence identity
with 1 other KEY virus isolate, Jamestown Can-
yon and Jerry Slough viruses, respectively.


St. Louis encephalitis virus was not isolated
from any species of mosquito collected in this
study. Similarly, Mitchell, et al. (1996) failed to
isolate SLE virus from mosquitoes collected from
36 sites in central and north Florida from April to
September, 1993. In 1990, SLE virus was isolated


Species Virus Site Isolate number MIR'

Ae. albopictus KEY 2 FL98-1290 1.0
An. crucians TEN 2 FL98-1199 2.6
TEN 2 FL98-1198 2.6
Cx. nigripalpus TEN 1 FL98-760 0.03
KEY 1 FL98-5240 0.06
KEY 1 FL98-5241 0.06
Cq. perturbans TEN 1 FL98-874 3.6
TEN 1 FL98-875 3.6
Wy. vanduzeei TEN 2 FL98-1207 2.9

'MIR = Minimum infection rate per 1,000 females tested.

September 2001

Nayar et al.: Mosquito arboviruses from Florida


Top three scores by the BLAST search (Accession No.; species)

Strain (FL98-) X73470; Northway X73465; Cache Valley D00353; Bunyamwera

760,874,875, 1198, 1199, 1207 95.4 95.4 93.4
U12801; Keystone U12796; Jamestown Canyon U12798; Jerry Slough
1290 99.5 93.4 92.3

from Cx. nigripalpus in Indian River County dur-
ing late summer and fall (Shroyer 1991). Interest-
ingly, in the almost 50-year history of the
presence of SLE virus in Florida, the virus has
only been recovered from Cx. nigripalpus during
epidemics years (Chamberlain et al. 1964; Dow et
al. 1964; Monath and Tsai 1987; Shroyer 1991;
Wellings et al. 1972). This could be because mos-
quitoes were generally collected for virus isola-
tion only during years when epidemics occurred.
Furthermore, the minimum infection rates of
SLE virus in Cx. nigripalpus were low even dur-
ing epidemics, and ranged from 0.6 to 1.1 per
1000 (Chamberlain et al. 1964; Dow et al. 1964;
Shroyer 1991).
Two viruses, KEY and TEN, were isolated from
five species of mosquitoes. KEY virus was recov-
ered fromAe. albopictus and Cx. nigripalpus, and
TEN virus fromAn. crucians, Cx. nigripalpus, Cq.
perturbans, and Wy. vanduzeei. These two viruses
occur in abundance in Florida from April to Sep-
tember each year and have been isolated from
several species of mosquitoes, including those col-
lected in our study (Taylor et al. 1971; Wellings et
al. 1972; Calisher et al. 1986; Mitchell et al.
1996). Only exception being a (first time) TEN vi-
rus isolate from Wy. vanduzeei. KEY and TEN vi-
ruses also have been isolated from small
mammals in Florida. KEY virus was isolated
from cotton rats and TEN virus from cotton rats,
marsh rabbits, swamp rabbits, and various other
small mammals (Taylor et al. 1971; Wellings et al.
1972; Calisher et al. 1986). KEY and TEN viruses
are of no public health or veterinary importance.
RT-PCR, followed by nucleotide sequencing of
positive specimens, and a GenBank similarity
search by using the BLAST search program
proved to be effective methods of identifying virus
species. By these methods none of the tissue cul-
ture- positive specimens contained a detectable
level of SLE viral RNA. These results also corre-
lated with serological identification and con-
firmed that the specimens contained KEY and
TEN nucleic acids.

This article is a Florida Agricultural Experimental
Station Journal Series No. R-07497.


AND R. G. MCLEAN. 1986. Distribution of Bunyam-
wera serogroup viruses in North America, 1956-
1984. Am. J. Trop. Med. Hyg. 35: 429-443.
L. D. BEADLE. 1964. Vector studies in the St. Louis
encephalitis epidemic, Tampa bay area, Florida,
1962. Am. J. Trop. Med. Hyg. 13: 456-461.
R. M. KINNEY, AND C. J. MITCHELL. 1994. An inte-
grated target sequence and signal amplification
assay, reverse transcriptase-PCR-enzyme-linked
immunosorbent assay, to detect and characterize
flaviviruses. J Clin Microbiol 32(2): 477-83.
DAY, J. F., AND L. M. STARK. 2000. Frequency of Saint
Louis encephalitis virus in humans from Florida,
USA: 1990-1999. J. Med. Entomol. 37: 626-633.
WORK. 1964. Isolation of St. Louis encephalitis virus
from mosquitoes in the Tampa bay area of Florida
during the epidemic of 1962. Am. J. Trop. Med. Hyg.
13: 462-468.
SMITH. 1996. Detecting bunyaviruses of the Bun-
yamwera and California serogroups by a PCR tech-
nique. J. Clin. Microbiol. 34(5): 1184-1188.
Arboviruses associated with mosquitoes from nine
Florida counties during 1993. J. Am. Mosq. Control
Assoc. 12: 255-262.
MONATH, T. P., AND T. F. TSAI. 1987. St. Louis enceph-
alitis: Lessons from the last decade. Am. J. Trop.
Med. Hyg. 37 Suppl. 40S-59S.
REITER, P. 1983. A portable battery-powered trap for col-
lecting gravid Culex mosquitoes. Mosq. News 43:
SHROYER, D. A. 1991. The 1990 Florida epidemic of St.
Louis encephalitis: virus infection rates in Culex ni-
gripalpus. J. Fla. Mosq. Control Assoc. 62: 69-71.
NINGS. 1971. California group arboviruses in Flor-
ida, host-vector relations. Am. J. Trop. Med. Hyg. 20:
Agents encountered during arboviral ecological
studies: Tampa Bay Area, Florida, 1963 to 1970. Am.
J. Trop. Med. Hyg. 21: 201-213.
WULFF, H., AND J. V. LANGE. 1975. Indirect immunoflu-
orescence for the diagnosis of Lassa fever infection.
Bull. W.H.O. 52: 429-436.

Florida Entomologist 84(3)

September 2001


Texas A&M University Research & Extension Center, 17360 Coit Road, Dallas, TX 75252-6599, USA


Melanoplus differentialis (Thomas) (Orthoptera: Acrididae) and several other species of
grasshoppers invade urban/suburban landscapes and retail/wholesale nurseries during the
hot, dry summers in the southern United States to consume the foliage of many species of
landscape plants and turfgrass. Two experiments were conducted to determine which insec-
ticides could be used to safely provide residual control for the continual daily migration of
grasshoppers in urban landscapes and nurseries. Leaves from treated Hibiscus moscheutos
were harvested sequentially in time at 1-, 5-, and 11-days posttreatment and adult differen-
tial grasshoppers were confined on them for 24-, 48- and 72-hr exposures. Treatments with
two synthetic pyrethroids, bifenthrin 0.66F (0.782 ml/liter) and lambda-cyhalothrin 9.52 WP
(0.748 g/liter), provided 94 and 83%, mortality respectively, with 24-hr exposure to the 1-
day-old treated leaves. Both chemicals provided 100% control of the grasshoppers during 72-
hr exposure. The half rate (0.391 ml/liter) of bifenthrin also provided 89% control within the
72-hr evaluation. Treatments with diazinon AG600 (4.25 ml/liter) also provided 80-85% con-
trol with 72-hr exposure on the 1-day-old treated leaves. Acephate 75% S (0.803 g/liter) pro-
vided 33-39% control on the 1-day-old residues. Lambda-cyhalothrin provided 84% control
with 72-hr exposure to the 5-day-old treated leaves. Residual control was also provided at 5
days by bifenthrin and acephate (53% and 46-50%, respectively). Most materials evaluated
failed to provide any protection at all and none of the treatments provided residual control
when grasshoppers were exposed to 11-day-old residues. No phytotoxicity to hibiscus was
observed due to any of the treatments.

Key Words: differential grasshopper, bioassays, bifenthrin, lambda-cyhalothrin, diazinon,
acephate, landscape pest, nursery pest


Melanoplus differentialis (Thomas) (Orthoptera: Acrididae) y varias otras species de salta-
montes invaden paisajes urbanos / suburbanos y viveros de venta al por menor / por mayor
durante los veranos calientes y secos al sur de los Estados Unidos para consumer el follaje
de muchas species de plants paisajistas y grama de c6sped. Dos experiments fueron lle-
vados a cabo para determinar cuales insecticides pudieran ser usados para proveer control
residual con seguridad para la continue migraci6n diaria de saltamontes en viveros y paisa-
jes urbanos. Hojas de Hibiscus moscheutos tratadas fueron cosechadas en secuencia de
tiempo a 1, 5, y 11 dias despues de tratamiento y saltamontes adults diferenciales fueron
confinados con ellas por exposici6n de 24, 46 y 72 horas. Tratamientos con dos piretroides
sinteticos, bifenthrin 0.55F (0.782 ml/litro) y lambda-cyhalothrin 9.52 WP (0.78 g/litro), pro-
veyeron mortalidad de 94 y 83%, respectivamente, con exposici6n de 24 hr. a las hojas trata-
das de 1 dia. Ambos quimicos proveyeron 100% de control de los saltamontes durante
exposici6n por 72 hr. La media dosis (0.391 ml/litro) de bifenthrin tambien provey6 89% de
control dentro de la evaluaci6n de 72 hr. Tratamientos con diazinon AG600 (4.25 ml/litro)
tambien provey6 80-85% de control con exposici6n de 72 hr. en las hojas tratadas de 1 dia.
Acephate 75% S (0.803 g/litro) provey6 33-39% de control en los residues de 1 dia. Lambda-
cyhalothrin provey6 84% de control con exposici6n por 72 hr. a las hojas tratadas por 5 dias.
Control de residues fue tambien proveido a los 5 dias por bifenthrin y acephate (53% y 46-
50%, respectivamente). La mayoria de los materials evaluados fracasaron en proveer al-
guna protecci6n del todo y ninguno de los tratamientos proveyeron control residual cuando
los saltamontes fueron expuestos a los residues de 11 dias. No fitotoxicidad a los hibiscos fue
observada debido a alguno de los tratamientos.

Several species of grasshoppers invade urban/ grasshoppers has been studied (Feaver 1985,
suburban landscapes and retail/wholesale nurs- Fielding and Brusven 1992, Harvey and Thomp-
eries during the hot, dry summers in the southern son 1993, Hinks et al. 1990). Based upon limited
United States to consume the foliage of many spe- surveys during the summer and fall of 1998 and
cies of landscape plants and turfgrass. The feed- 1999 and the summer of 2000, the differential
ing behavior of several species of Melanoplus grasshopper, Melanoplus differentialis (Thomas)

Reinert et al.: Chemical control for Melanoplus differentialis

(Orthoptera: Acrididae), is the species most fre-
quently encountered in damaging numbers in the
Texas landscape. Additionally, the two-striped
grasshopper, M. bivitattus (Say), and migratory
grasshopper, M. sanguinipes (Fabricius), migrate
into the urban environs to cause significant dam-
age to the landscape.
Cooperative extension reports from Kansas
(Bauernfeind 1992) and Texas (Patrick 1998) also
report these species as the primary grasshopper
pests of gardens and urban landscapes. Nymphs
are usually not a problem in urban plantings as
they normally develop in pasture and field-crop
settings. Only after they molt to the adult stage
do we see the migration into urban landscapes.
However, in rural landscapes, severe damage may
result from both nymphs and adults that readily
move by walking from adjacent fields and road-
sides. A mature grasshopper feeding on a small
shrub or bedding plant can soon disfigure it and
several feeding adults can ruin the aesthetic
value of plants around a home within a short
time. The economic impact to a retail or wholesale
nursery can be very high. Plants are sold for their
aesthetic value and even limited grasshopper
feeding can soon render the plants unsaleable.
Outbreaks are usually preceded by several
years with hot, dry summers and warm autumns
(Patrick 1998). Also, the dry weather increases
survival of both nymphs and adults. The ex-
tremely hot and dry summer of 1998 created ideal
conditions for extensive outbreaks of grasshop-
pers across much of the southern United States.
Populations the following years (1999 and 2000)
were also high and caused extensive damage. As
pastures, field crops and uncultivated areas were
either harvested or desiccated from the drought
conditions, mature grasshoppers readily dis-
persed into plant nurseries and the urban land-
scape in search of food. As a result, extensive
damage was common, especially in Texas, on
lawns and many species of landscape plants. Con-
trol strategies were needed to manage the inva-
sion within the urban scape and in plant
nurseries. The purpose of these experiments was
to determine which insecticides could be used to
safely control grasshoppers on landscape plants
and in nursery culture and also, to determine if
any of the treatments could provide residual con-
trol for the continual daily migration of the adults.


Two experiments were conducted to evaluate
insecticides for residual control. Chemicals and
rates evaluated are given in Tables 1 and 2. For
each experiment, 'Disco Rose Red' Hibiscus, Hi-
biscus moscheutos, plants [ca. 30 to 40 cm high
grown in 15-cm diam. (1 gal) pots] were obtained
from a local nursery. Plants were sprayed to run-
off with the respective treatments. Silwet, an or-

ganosilicon wetting agent, was added to each
treatment at a rate of 1 ml/liter of water. Two
plants were treated with each insecticide in each
replicate to ensure adequate treated foliage
would be available for sequential residual evalu-
ations. Plants were maintained in full sun to al-
low maximum bio-degradation of the treatment
chemicals. Leaves were clipped at 1- and 5-days
posttreatment (DAT) in Experiment 1 and at 1-,
5-, and 11-DAT in the second experiment, bagged,
placed in a cooled ice chest and taken to the labo-
ratory. Two to three treated leaves were caged
with each individual adult grasshopper in 9-cm
diam. x 20 mm plastic growth chambers and the
individuals were observed every 24 hr for up to
five days. Each feeding chamber was first pro-
vided with two 7-cm filter paper discs, saturated
with distilled water, to maintain foliage turgidity.
For both experiments, 5 reps each with 4 adults
were evaluated for the respective days after treat-
ment. For each evaluation period, mortality rat-
ings were made at 24-, 48-, 72-, 96- and 120-hr
after grasshoppers were caged on the clipped,
treated plant material.
For these studies, field populations of adult dif-
ferential grasshoppers were collected from stands
of Johnsongrass, Sorghum halepense (L.) Pers.,
growing in railroad or highway rights-of-way at
sites in either Denton or Collin Co., TX. Adults
were individually collected with a sweep net,
transferred to stems and leaves of Johnsongrass
in plastic shoe boxes that had been modified with
screen lids, and stored in cooled ice chests for im-
mediate transport to the laboratory. Only grass-
hoppers that appeared healthy the next day were
used to initiate the residue studies. Either males
only or females only were used within each repli-
cate, to ensure that any differences in susceptibil-
ity due to sex would be accounted for statistically
as replication error.
Data were adjusted (Abbott 1925) in Experi-
ment 1 (Table 1) since mortality in the untreated
check approached 10% at both 72-hr evaluations.
No adjustment was needed in Experiment 2, since
no grasshoppers died in the untreated check dur-
ing the study. All data were analyzed using Anal-
ysis of Variance and General Linear Model
Procedures. Treatment means were separated by
Waller-Duncan k-ratio t test (k = 100) (P = 0.05)
(SAS Institute 1990). Percent mortality data was
transformed by arcsines before analysis. Un-
transformed means are presented.


Treatments with two synthetic pyrethroids,
bifenthrin 0.66F (Talstar) (0.782 ml/liter) and
lambda-cyhalothrin 9.52 WP (Simitar) (0.748 g/li-
ter), provided 94 and 83% mortality, respectively,
with 24-hr exposure to the 1-DAT hibiscus leaves
(Table 2). Furthermore, both chemicals provided


Exposed 1-DAT b Exposed 5-DATb
Treatment' (ml or g product/liter) 24hr 48hr 72hr 24hr 48hr 72hr

Diazinon AG600 4.25 ml/1 40 ad 53 a 83 a -
+ Abamectin 0.15 EC + 0.31 ml/1
Diazinon AG600 4.25 ml/1 20 b 58 a 78 a 0 b 0 b 0 b
Acephate 75WP 0.8 g/1 5 bc 21b 33 b 25 a 35 a 50 a
CGA293,343 25 WG 0.16 ml/1 0 c 0 c 22 bc -
+ Emamectin Benzoate 5 SG + 0.23 ml/1
CGA293,343 25 WG 0.32 ml/1 0 c 5 bc 17 bcd -
Pymetrozine 50 WG 0.19 g/1 5 bc 5 bc 5 cd -
Abamectin 0.15 EC 0.31 ml/1 5 bc 5 bc 5 cd -
Emamectin Benzoate 5 SG 0.23 ml/1 0 c 0 bc 0 cd -
Pymetrozine 50 WG 0.19 g/1 0 c 0 c 0 cd -
+ Abamectin 0.15 EC + 0.31 ml/l
Pymetrozine 50 WG 0.19 g/1 0 c 0 c 0 d -
+ Emamectin Benzoate 5 SG + 0.23 ml/l
CGA293,343 25 WG 0.16 ml/1 0 c 5 bc 0 cd -
+ Abamectin 0.15 EC + 0.31 ml/l
Untreated Check 0 0 c 0 bc 0 cd 0 b 0 b 0 b

was assayed after 24-, 48- and 72-hr ex-

'Silwet, an organosilicon wetting agent was added to all treatments at a rate of lml/iter of water.
Leaves were harvested from plants with the respective treatments at 1- and 5-days-after-treatment and caged with individual grasshopper adults. Mortality of the grasshopper
posure and feeding on the treated leaves.
'Analysis was made on arcsine transformation of the percent mortality data: percent mortality is presented.
Means in a column not followed by the same letter are significantly different by Waller-Duncan k-ratio t-test (k = 100) (P = 0.05).


Exposed 1-DAT" Exposed 5-DAT" Exposed 11-DAT"
Treatment' (ml or g product/liter) 24hr 48hr 72hr 24hr 48hr 72hr 24hr 48hr 72hr

Bifenthrin 0.66F 0.782 ml/l 94.4 ad 94.4 ab 100 a 40.0 ab 45.5 b 52.6 b 0 n 0 0
Lambda-cyhalothrin 0.748 g/1 83.3 a 100 a 100 a 55.0 a 68.4 a 84.2 a 0 0 0
Bifenthrin 0.66F 0.391 ml/l 50.0 b 83.3 b 88.9 b 15.0 bc 15.0 c 26.3 cd -
Acephate 75%S 0.803 g/1 33.3 bc 38.9 c 38.9 c 40.0 ab 45.5 b 45.5 bc 0 0 0
Deltamethrin 50SC 1.564 ml/l 16.7 cd 16.7 cd 16.7 d 5.0 c 10.5 c 26.3 cd -
Deltamethrin 50SC 0.391 ml/l 0 d 0 d 0 d 0 c 10.5 c 26.3 cd -
Carbaryl 4 SL 2.50 ml/1 0 d 0 d 0 d 0 c 0 c 5.3 d -
Imadocropid 75WP 0.038 g/1 0 d 0 d 0 d 0 c 0 c 5.3 d -
Untreated Check 0 0 d 0 d 0 d 0 c 0 c 0 d 0 0 0

'Silwet, an organosilicon wetting agent was added to all treatments at a rate of 1 ml/liter of water.
Silwet, an organosilicon wetting agent was added to all treatments at a rate of 1 ml/hliter of water. Leaves were harvested from plants with the respective treatments after 1-, 5-, and 11-days-after-treatment and caged
with individual grasshopper adults. Mortality of the grasshoppers was assayed at 24-, 48- and 72-hr exposure and feeding on the treated leaves.
'Analysis was made on arcsine transformation of the percent mortality data: percent mortality is presented.
Means in a column not followed by the same letter are significantly different by Waller-Duncan k-ratio t-test (k = 100) (P = 0.05).

100% control of the grasshoppers with 72 hr expo-
sure. The half rate of bifenthrin (0.391 ml/liter)
also provided 89% control with 72 hr exposure.
Treatments with either diazinon AG600 (4.25 ml/
liter) or diazinon AG600 (4.25 ml/liter) + abamec-
tin 0.15 EC (Avid) (0.31 ml/liter) also provided 78-
83% control with 72 hr exposure to the 1-DAT
leaves (Table 1). Acephate 75% S (Orthene TTO)
(0.803 g/liter) provided limited initial control (33-
39%) (Table 2). Other treatments evaluated did
not provide more than 22% control for the 1-day
residue evaluation. Mortality at 96- and 120-hr
was not significantly greater than for the 72-hr
To determine residual control, leaves that had
been treated 5 and 11 days earlier were also har-
vested and grasshoppers were caged on them.
Lambda-cyhalothrin provided 84% control within
72 hr on the 5-DAT leaves. Both bifenthrin and
acephate also provided residual control (53 and
46-50%, respectively) at 5 days. The increase in
residual control for acephate, from 33- 39% at 1
day to 46-50% at 5 days could probably be attrib-
uted to its systemic action. None of the treat-
ments provided any residual control when
grasshoppers were exposed to 11-DAT leaves. No
phytotoxicity to hibiscus was observed due to any
of the treatments.
A higher level of control might have been
achieved with these treatments if they were ap-
plied directly to the feeding grasshoppers or if the
grasshoppers were immediately exposed to the
treated foliage. Also, a higher level of control
would be anticipated if the treatments were ap-
plied to the immature stages. It was the main
purpose of these experiments, however, to evalu-
ate the effect of these toxicants on grasshoppers
that were migrating onto the treated plants. Only
a limited percentage of the grasshopper popula-
tion will actually be sprayed when the treatment
is applied.
These experiments provide important manage-
ment information for the nursery and landscape
industries. These experiments show bifenthrin
and lambda-cyhalothrin (both synthetic pyreth-
roids) and diazinon, each provide significant con-

September 2001

trol of grasshoppers, even when they migrate onto
the treated foliage a day after treatments are ap-
plied. This level of control may increase if the
grasshoppers are directly contacted with the spray
treatments. Bifenthrin, lambda-cyhalothrin and
acephate also provided at least 5-day residual
control for the differential grasshopper. By choos-
ing one of the more residual chemicals, repeat
applications should only be necessary every 5
days or even weekly to protect landscape plants.
Each of the effective insecticides is labeled for
grasshoppers and available for homeowner or
commercial treatment of landscape plants.
This article reports the results of research
only. Mention of a proprietary product does not
constitute an endorsement or recommendation by
the Texas A&M University Agriculture Program.


ABBOTT, W. S. 1925. A method of computing the effec-
tiveness of an insecticide. J. Econ. Entomol. 18(2):
BAUERNFEIND, R. J. 1992. Grasshoppers in the lawn and
garden. Kansas State Univ. Coop Ext. Serv. Entomol.
487, Leaflet L-868. 4 p.
Feaver, M. N. 1985. Grasshopper (Orthoptera: Acididae)
damage to pine seedlings at night in a seed orchard.
Florida Entomol. 68(4): 694-696.
FIELDING, D. J., AND M. A. BRUSVEN. 1992. Food and
habitat preference of Melanoplus sanguinipes and
Aulocara elliotti (Orthoptera: Acrididae) on dis-
turbed rangeland in southern Idaho. J. Econ. Ento-
mol. 85(3): 783-788.
HARVEY, T. L., AND C. A. THOMPSON. 1993. Differences
in leaf feeding on corn hybrids by the differential
grasshopper, Melanoplus differentialis (Thomas).
J. Agric. Entomol. 10(1): 31-34.
WORTH, AND W. CRAIG. 1990. Preference and perfor-
mance in grasshopper, Melanoplus sanguinipes
(Orthoptera: Acrididae), feeding on kochia, oats, and
wheat: implication for population dynamics. J. Econ.
Entomol. 83(4): 1338-1343.
PATRICK, C. D. 1998. Grasshoppers and their control.
Texas A&M Univ., Agric. Ext. Serv. Leaflet L-5201.
4 p.
SAS Institute. 1990. SAS/STAT User's Guide, version
6.10, ed. SAS Institute, Cary, NC.

Florida Entomologist 84(3)

Carrel: Red Widow Spider Populations


Division of Biological Sciences, 105 Tucker Hall, University of Missouri-Columbia, Columbia, MO 65211-7400


Populations of the red widow spider, Latrodectus bishop, in native Florida scrub at the Arch-
bold Biological Station were monitored annually on ten ~0.5 ha transects in late winter from
1987 to 2000. Of 398 L. bishop detected in the study, all but three had their silken retreats
built in palmetto leaves. L. bishop at rest in retreats in saw palmetto (Serenoa repens) were
higher above the ground (-0.5 m) than spiders in scrub palmetto (Sabal etonia) (-0.3 m).
From a peak of 31 spiders/ha in 1989, the average L. bishop density declined exponentially
to only 0.3 spiders/ha in 1997, after which L. bishop densities began to recover. Burning of
scrubby transects in spring or summer appeared to have no affect on subsequent L. bishop
populations. There were no significant correlations between L. bishop population density
and local temperature or precipitation data. These results suggest that undescribed biotic
factors may regulate populations of the red widow spider in a density-dependent fashion.
Key Words: Latrodectus, Florida scrub, ecology, populations, dynamics, fire


En diez transectos de -0.5 ha, determinamos cada invierno de 1987-2000 las poblaciones de
la arana Latrodectus bishop en matorral nativo de Florida en la Estaci6n Biol6gica Arch-
bold. De las 398 L. bishop que encontramos, todos menos tres habian construido sus retiros
sedosus entire las ojas de palmitos. Las L. bishop que reposan en el palmito Serenoa repens
estan mas alto (-0.5 m) que ellos que reposan en el palmito Sabal etonia (-0.3 m). El prome-
dio densidad de L. bishop disminuy6 exponencialmente desde 31 arahas/ha en 1989 a 0.3
arahas/ha en 1997, despues de que las densidades de L. bishop empezaban a recuperar.
Quemando el matorral en la primavera o verano no afect6 las poblaciones subsiguientes de
L. bishop. No habia correlaciones significativas entire la densidad de L. bishop y la tempe-
ratura o precipitaci6n local. Los results sugieren que hay factors biol6gicos no describido
que regular poblaciones de L. bishop en una manera densidad-dependiente.

The red widow spider, Latrodectus bishop Kas-
ton 1938, is endemic to xeric, upland ecosystems
found in Central and Southeastern Florida (Levi &
Levi 1990; Edwards 1994). It is restricted to sand
pine scrub and scrubby flatwoods in several coun-
ties that depend on periodic burning to maintain
species diversity (McCrone & Levi 1964; McCrone
& Stone 1965; Kaston 1970; Levi & Levi 1990;
Abrahamson et al. 1984). Little is known about
this rare spider. In large part this stems from the
fact that it is difficult to find, even when it is locally
abundant. Although L. bishop builds a large tan-
gled web on palmetto shrubs (McCrone & Levi
1964; Edwards 1994; Sierwald & Fenzl 1999), the
very fine silk is not highly visible in bright sun-
light. Furthermore, its funnel-shaped, silken re-
treat usually is hidden within a folded palmetto
leaf (McCrone & Levi 1964; Sierwald & Fenzl
1999). Field biologists studying vertebrates at the
Archbold Biological Station in Highlands County,
Florida, have noticed that local populations of the
L. bishop seem to erupt every 10-20 years. Early
reports suggested mild winters and periods of
drought subsequently result in an increased abun-
dance of spiders of the genus Latrodectus in many
regions of the world (Chamberlain & Ivie 1935).

This study was undertaken in order to gain ba-
sic knowledge about the interannual population
dynamics of L. bishop. Specifically, I censused
web-sites of subadult and adult female L. bishop
in ten replicate tracts of native scrub annually for
twelve out of fourteen years in a row at the Arch-
bold Biological Station to ascertain long-term
changes in population density. I identified the
plant species used for a retreat and twice during
the study I measured the height of each spider's
retreat above the ground as an indication of web-
site preference byL. bishop. In addition, I used a
null model to test the short-term affect of fire on
the density of L. bishop. Finally, using weather
data obtained from Archbold records, I tested
whether density of L. bishop is correlated in a
simple way with temperature or precipitation.


Study Area

The Archbold Biological Station is located near
the southern terminus of the Lake Wales Ridge in
Highlands County, Florida (2711'N lat., 8121'W
long.), 12 km south of the town of Lake Placid.

Florida Entomologist 84(3)

The elevation of the study area ranges from ap-
proximately 38 to 46 m above mean sea level. The
predominant vegetative associations in the study
area are scrubby flatwoods, which are dominated
by low shrubby oaks (Quercus inopina, Q. chap-
manii, Q. geminata) and palmettos (Serenoa
repens and Sabal etonia). Interspersed among the
scrubby flatwoods to varying degrees are two
other vegetative associations: sand pine scrub,
with widely scattered stands of sand pine (Pinus
clausa) and an understory of xerophytic shrubs,
and flatwoods, with open stands of south Florida
slash pine (P. elliottii var. densa) and an under-
story and ground cover of mesic grasses, herbs,
saw palmetto (Serenoa repens), and assorted
shrubs (Abrahamson et al. 1984).

Spider Censuses

In 1987 Mary Haskins, Zhaofen Yang, and I
discovered that L. bishop webs are easily seen
and reliably identified from a distance of many
meters at dawn on very foggy mornings because
the dew-laden cobweb is highly reflective. Using
this knowledge, I devised a drive-by method to
census L. bishop in scrub on the side of primitive,
sandy roads that pass through the scrub. I estab-
lished a total of ten permanent, roadside
transects in scrub that had been burned in 1984
or 1985. Each transect extended 10 m from the
road into the scrub and ranged in length from 375
to 730 m. The average area (SE) of each transect
was 0.55 + 0.03 ha. At dawn on foggy mornings in
late winter (February-early March), I drove
slowly (1-3 km/h) in a light truck along the edge of
each transect, looking from a height of ~2 m into
the scrub for L. bishop webs. Upon sighting a
web, I stopped the truck, walked to the web, lo-
cated the spider in its retreat, and marked the
web-site with surveyor tape tied 1-2 m high on
nearby vegetation. After repeating each drive-by
survey three or four times within a 2 week period,
I ceased to find additional webs. Subsequently
during the daytime I revisited each L. bishop
web-site, carefully opened the retreat, noted
whether it was occupied by an adult or an imma-
ture female, and recorded the species of plant
used by a spider for its retreat. In 1989 and 1999
I also measured the height ofL. bishop retreats
in the two species of palmettos.
To verify the efficacy of the drive-by method for
detecting L. bishop webs, in the second year of
the study (1989) I walked through each transect
during daytime (0900-1600 h) looking forL. bish-
opi webs several days before I began the drive-by
censuses. I visually inspected the leaves of every
palmetto within a transect at close range (< 1 m).
If I found a L. bishop web, I marked its location
cryptically by burying a piece of surveyor tape in
the sand near it in such a fashion that the tape
was not visible from the nearby road. I spent a to-

tal of 30 h searching on foot and 20 h driving
slowly looking for webs. After the drive-by survey
was completed, I compared the number ofL. bish-
opi webs detected by the two methods.
Representative specimens of L. bishop were
preserved in the collection of arthropods at Arch-
bold. Statistical analyses were performed using
SYSTAT (Wilkinson 1989).

Fire Affects on L. bishop Populations

A record of the date, area, and location of fires
on the main property at Archbold is kept as part
of the fire management plan (Main & Menges
1997). In addition to burns in 1984 or 1985, eight
of my transects were burned once and two
transects were burned twice during the course of
my study. To test the short-term affects of fire on
local L. bishop populations, I developed a null
model against which to test the observed data.
The null model posited that fire in late spring or
summer would have no significant affect on L.
bishop spider densities determined several
months later in winter. Hence, one would expect
an equal proportion of transects (1/3) to show an
increase, a decrease, or no change in density in
winter after the burn event relative to the winter
before the burn. The Fisher exact test was used to
test the difference between observed and ex-
pected outcomes (Zar 1974).

Weather Affects on L. bishop Populations

I conducted my censuses for twelve years
(1987-2000, except for 1988 and 1991) late in win-
ter when many of the native shrubs and trees be-
gan to flower or produce new foliage. I obtained
weather records starting from the official weather
center at Archbold, which has been in operation
continuously since 1952. As an indicator of long-
term climatic conditions that prevail at Archbold,
I calculated the 30-year mean value (data for
1952-1981) and the 95% confidence interval (95%
C.I.) for four weather parameters: mean daily
temperature in winter (Jan., Feb., & Mar.), mini-
mum winter temperature, total annual precipita-
tion, and mean monthly precipitation in winter.
Subsequently I compared the same four parame-
ters for each year starting 1985 with the 30-year
means to determine whether there were signifi-
cant annual deviations during my study.


Comparison of Sampling Methods for L. bishop

As summarized in Table 1, searching on foot
during the daytime for L. bishop web-sites in the
scrub was very inefficient compared to the drive-
by method conducted at dawn on foggy mornings.
When I searched on foot in 1989, I found a total of

September 2001

Carrel: Red Widow Spider Populations


Number of web-sites detected

On-foot Drive-by By both
Transect search search search
number only only methods Total

1 4 22 28 54
2 1 2 10 13
3 0 0 0 0
4 2 3 5 10
5 1 7 3 11
6 1 8 4 13
7 0 4 0 4
8 1 13 2 16
9 3 14 4 21
10 4 10 12 26

Sum 17 83 68 168
Percent 10.1 49.4 40.5 100.0

85 L. bishop on the ten transects, 68 of which I
subsequently detected in the drive-by survey. On
the other hand, I detected a total of 151 webs us-
ing the drive-by method, 68 of which I had previ-
ously found in my laborious searches on foot of
the many palmettos in the transects. Hence,
searching on foot was only about 50% effective
whereas the drive-by method was about 90% ef-
fective for finding L. bishop. The webs I missed
using the drive-by method often were located low
to the ground and on the side of a palmetto plant
facing away from the road. Because the drive-by
method seemed to be a reasonably accurate way
of censusing L. bishop populations, I adopted it
throughout the remainder of the study.

Plants Used by L. bishop for Web-sites

Of 398 L. bishop detected in this study, 395
(99.2%) had their retreats located in leaves of pal-
mettos. Two L. bishop had retreats hidden be-
neath leaves of staggerbush, Lyonia fruticosa
(Michx.) Torr. [Ericaceae] and a third spider was
resting in a retreat spun under leaves of sand live
oak, Quercus geminata Small [Fagaceae]; both
plants are evergreen shrubs.
L. bishop used saw palmettos (Serenoa
repens) as web-sites much more often than scrub
palmettos (Sabal etonia). As indicated in Table 2,
75-80% of L. bishop webs in 1989 and 1999 were
in saw palmettos and 20-25% were in scrub pal-
mettos. Analysis of variance revealed that the


Saw palmetto Scrub palmetto
Year (Serenoa repens) (Sabal etonia)

1989 0.51 + 0.02' 0.32 + 0.03b
1999 0.47 + 0.03' 0.33 + 0.04b

'Means + standard errors followed by the same letter are not signifi-
cantly different by ANOVA followed by Tukey HSD test (P > 0.05).

height of a L. bishop retreat off the ground was
highly dependant on the palmetto species (F =
21.36, P < 0.0001), but not the year of sampling or
the species*year interaction. On average L. bish-
opi retreats in saw palmettos were about 0.5 m
above ground, whereas those in scrub palmettos
were only 0.3 m above the sandy soil. Because the
distance from a L. bishop retreat to the top of a
typical palmetto leaf was about 0.3-0.4 m (J. Car-
rel, unpublished data), this means that the maxi-
mal height of palmettos harboring L. bishop
generally was below 1.0 m.

Annual Changes in Density of L. bishop

The density of L. bishop declined one hun-
dredfold from 1989 until 1998, but thereafter it
began to increase (Fig. 1). The highest mean den-
sity (SE), achieved in 1989, was 30.7 + 8.1 L.
bishopi/ha and the lowest mean density, achieved
in 1997, was 0.27 + 0.27 L. bishopi/ha. The de-
cade-long decline in mean L. bishop density was
highly exponential. IfY = meanL. bishop density

1 I

1988 1990 1992 1994 1996

1998 2000

Fig. 1. Annual variation in density of red widow spi-
ders (Latrodectus bishop) at the Archbold Biological
Station, Lake Placid, FL. Means + standard errors are
indicated (N = 10 permanent transects).

Florida Entomologist 84(3)

in year X, and X = 1 for the year 1989 and X = 10
in 1998, then the best fit regression equation is: Y
= 30.827 (10 -2332x) and the correlation coefficient
is highly significant (R = 0.977, df= 8, P < 0.0001).
The steady increase in L. bishop densities that
occurred from 1998 to 2000 suggests that local spi-
der populations may erupt in the near future. If this
were to happen, then in 2003-2005 L. bishop popu-
lations would resemble those found in 1987-1989.

Affect of Burning on L. bishop Density

Burning of the scrub in late spring or summer
had no affect on subsequent L. bishop spider pop-
ulations. The observed changes in L. bishop den-
sity on any transect were identical to the expected
values based on the null model. Of the 12 burn
events on transects that happened in the course
of this study, four corresponded with increases in
L. bishop density, 4 with decreases in L. bishop
density, and four with no change in L. bishop
density. These results are not very surprising con-
sidering the long period of time (~6-10 months)
between the occurrence of a fire and my field mea-
surements of spider densities. For example, the
palmettos and some other shrubs had fully regen-
erated many new leaves by the time I conducted
my censuses in winter.

Affect of Weather on L. bishop Density

In 1987 and 1989 when L. bishop densities
were highest, temperatures were unusually cold
(Fig. 2) and precipitation was normal in winter
but relatively low during the remainder of the
year (Fig. 3). However, in 1990 when L. bishop


~o u -

I l o
1980 19 1920



40 -

185 190 1995 2000
Fig. 3. Total annual precipitation (open squares) and
mean monthly precipitation in winter (solid squares)
from 1985 to 2000 at the Archbold Biological Station,
Lake Placid, FL. Each set of three horizontal lines indi-
cates the 30-year mean + the 95% confidence interval
for both types of data.

densities fell greatly compared to the year before,
conditions were unusually warm and somewhat
dry. In fact, the minimum daily temperature in
winter 1990 never dipped below 3.30C, making
this winter the least extreme on record at Arch-
Annual L. bishop density was not correlated (r
< 0.45, P > 0.05, df = 9 or 10) with any of the tem-
perature or precipitation indices I used (see
Methods section). Hence, there was no evidence
that weather limits L. bishop populations in a
straightforward fashion.


S Regulation of Spider Populations



Fig. 2. Mean daily temperat
squares) and minimum tempera
squares) from 1985 to 2000 at th
Station, Lake Placid, FL. Each s
lines indicates the 30-year mean
interval for both types of data.

Interannual variation in the density of spiders,
like those of other animals, can be caused both by
abiotic and by biotic factors (Price 1975;Watson &
Ollason 1982; Askew & Yalden 1985; Gaston &
McArdle 1993; Wise 1993). The lack of a signifi-
cant correlation between L. bishop density and
temperature, precipitation, or fire events during
S, my long-term, highly replicated study suggests
that abiotic factors probably did not determine
__ the pattern of change observed in L. bishop pop-
19s 200 ulations. On the other hand, the exponential de-
cline in spider densities from 1989 to 1998 implies
that a density dependent mechanism might regu-
ure in winter (open late L. bishop populations.
ture in winter (solid
e Archbold Biological Although there seem to be no long-term stud-
et of three horizontal ies of population dynamics for any other Latro-
+ the 95% confidence dectus species, there is good evidence that natural
enemies commonly limit spider population densi-

1 n- 0 aI

September 2001

Carrel: Red Widow Spider Populations

ties (Wise 1993 and references therein). Candi-
date species that effectively prey in a density-
sensitive fashion on immature and adult L. bish-
opi are the sphecid wasps C i'..,! .... californicum
and Sceliphron caementarium and the Florida
scrub-jay Aphelocoma coerulescens. These three
species are known to eat L. bishop at Archbold
(M. Deyrup and G. Woolfenden, unpublished ob-
servations) and, more generally, they are known
to modify their feeding habits in response to
changes in the relative abundance of prey (Coville
1987; Woolfenden & Fitzpatrick 1984). In addi-
tion, theridiid spiders of the genusArgyrodes liv-
ing in L. bishop webs may do more than act as
kleptoparasites stealing the host's prey; these
small spiders may prey on their hosts (Sierwald &
Fenzl 1999). Finally, scelionid wasps of the genus
Idris may contribute significantly to regulation of
L. bishop populations since these tiny animals
could easily exhibit both functional and numeri-
cal responses as their food base changes. Idris is
the largest genus of insects at Archbold, consist-
ing of many undescribed species about which lit-
tle is known except that they specialize in feeding
on insect and spider eggs (M. Deyrup, unpub-
lished results).

Site Selection by Widow Spiders

Results presented here indicate thatL. bishop
strongly prefers saw palmetto (Serenoa repens)
more than all other shrubs for web-sites. The
cause of this preference is not clear. Whether
chance or necessity determines habitat selection
in web-building spiders has been investigated to a
limited extent (Lubin et al. 1993; Wise 1993; Foe-
lix 1996). Considering that palmettos comprise
the dominant shrub type in burned scrub at Arch-
bold and saw palmetto is much more common
than scrub palmetto (Sabal etonia) (Abrahamson
1995), a fully probabilistic model for site selection
might generate the observed L. bishop web-site
Alternatively, a deterministic process of site
selection by L. bishop might involve two aspects
of palmetto architecture. First, there is a major
difference in the patterns of growth between the
two palmetto species. The central axes of saw pal-
metto leaves originate in expanded basal sheaths
extending upward from terminal tufts on horizon-
tal stems just above or at the soil surface,
whereas scrub palmetto leaves arise on axes ex-
tending up from subterranean stems. L. bishop
females occasionally use the tubular spaces in
leaf bases of saw palmettos for their retreats,
which they cannot do with scrub palmettos. Per-
haps young L. bishop, as they disperse from their
natal web and actively search for protected sites
near the soil surface, preferentially select the ter-
minal tufts on saw palmettos and then tend to re-
main there as they grow and mature.

Second, saw palmettos consistently have more
leaves packed densely in narrower crowns than
scrub palmettos. Consequently, saw palmetto
leaves often overlap and self-shade (Abrahamson
1995). This suggests that the arrangement of saw
palmetto leaves may offer L. bishop more protec-
tion from enemies and from thermal extremes
than scrub palmetto leaves. Lubin et al. (1993) re-
ported comparable evidence for the desert widow
spider, L. revivensis: selection of larger shrubs in
the Negev desert improves spider survival,
growth, and reproductive success. But the cost of
moving to new web-sites for desert widow spiders
is high (40%) mortality. Experiments designed to
determine mechanisms and the cost/benefit ratio
of web-site selection byL. bishop in Florida scrub
are in progress.

Implications for Conservation of Rare Spiders

The L. bishop is restricted to scrub habitats
that are remnants of ancient islands in peninsu-
lar Florida, now recognized collectively as a major
site of biotic endemism (Deyrup & Eisner 1993).
But Florida scrub is threatened by loss of habitat
resulting from rapid development and by frag-
mentation that limits gene flow and heightens the
probability of extinction in local populations. Re-
cently Skerl (1999) recommended that spiders
which are naturally rare because they have
highly restricted ranges might be listed nation-
ally as "species of conservation concern," even if
they are locally abundant. In addition, Skerl and
Gillespie (1999) advocated targeting for conserva-
tion action spiders with narrow habitat require-
ments, limited dispersal abilities, restricted
ranges, and immediate threats. L. bishop seems
to meet all of their criteria. Hence, I suggest that
L. bishop be listed as a species of conservation
concern and that its populations be surveyed in
peninsular Florida in order to determine the lim-
its of its distribution.

I thank Mark Deyrup, James Layne, and Glen Wool-
fenden for many informative discussions about the nat-
ural history of the Lake Wales Ridge; Nancy Deyrup for
providing weather data; the staff of the Archbold Biolog-
ical Station for providing research facilities; and Jan
Weaver, Mary Haskins, and Zhaofen Yang for helping in
the field. Funding for this work came in part from a
grant from the Research Council and from the Develop-
ment Fund at the University of Missouri. Steve Latta
kindly provided the Spanish abstract.


ABRAHAMSON, W. G. 1995. Habitat distribution and
competitive neighborhoods of two Florida palmettos.
Bull. Torrey Bot. Club 122: 1-14.
P. A. PERONI. 1984. Vegetation of the Archbold Bio-

logical Station, Florida: an example of the southern
Lake Wales Ridge. Florida Scient. 47: 209-250.
ASKEW, R. R., AND D. W. YALDEN. 1985. The Woodches-
ter Park valley. In L. M. Cook [ed.]. Case Studies in
Population Biology. Manchester University Press,
Manchester, UK. pp. 1-26.
CHAMBERLAIN, R. V., AND W. IVIE. 1935. The black
widow spider and its varieties in the United States.
Bull. Univ. Utah 25: 3-29.
COVILLE, R. E. 1987. Spider-hunting sphecid wasps. In
W. Nentwig [ed.]. Ecophysiology of Spiders.
Springer-Verlag, Berlin. pp. 309-318.
DEYRUP, M., AND T. EISNER. 1993. Last stand in the
sand. Natural History 102 (12): 42-47.
EDWARDS, G. B. 1994. Red widow spider Latrodectus
bishop Kaston. In M. Deyrup and R. Franz [eds.].
Rare and Endangered Biota of Florida. Vol. 5. Inver-
tebrates. pp. 250-251.
FOELIX, R. F. 1996. Biology of Spiders, Second Edition. Ox-
ford University Press, NY and Oxford, UK. 330 pp.
GASTON, K. J., AND B. H. MCARDLE. 1993. All else is not
equal: temporal population variability and insect
conservation. In K. J. Gaston, T. R. New, and M. J.
Samways [eds.]. Perspectives on Insect Conserva-
tion, Intercept Limited, Andover, UK. pp. 171-184.
KASTON, B. J. 1970. Comparative biology of American
black widow spiders. Trans. San Diego Soc. Nat.
Hist. 16: 33-82.
LEVI, H. W., AND L. R. LEVI. 1990. A Guide to Spiders
and Their Kin. Golden Press, NY. Western Publish-
ing Company, Inc., Racine, WI. 160 pp.
LUBIN, Y., S. ELLNER, AND M. KOTZMAN. 1993. Web re-
location and habitat selection in a desert widow spi-
der. Ecology 74: 1915-1928.

September 2001

MAIN, K. N., AND E. S. MENGES. 1997. Station manage-
ment plan. Archbold Biological Station Land Man-
agement Publication 97-1. 104 pp.
MCCRONE, J. D., AND H. W. LEVI. 1964. North American
widow spiders of the Latrodectus curacaviensis
group (Araneae: Theridiidae). Psyche 71: 12-27.
MCCRONE, J. D., AND K. J. STONE. 1965. The widow spi-
ders of Florida. Arthropods of Florida and Neighbor-
ing Land Areas. Vol. 2, 8 pp.
PRICE, P. W. 1975. Insect Ecology. John Wiley & Sons,
NY. 514 pp.
SIERWALD, P., AND T. FENZL. 1999.Argyrodes in webs of
the Floridian red widow spider (Areaneae: Theridi-
idae). Florida Entomol. 82: 359-361.
SKERL, K. L. 1999. Spiders in conservation planning: a
survey of US natural heritage programs. J. Ins. Con-
serv. 3: 341-347.
SKERL, K. L., AND R. G. GILLESPIE. 1999. Spiders in con-
servation-tools, targets and other topics. J. Ins.
Conserv. 3: 249-250.
WATSON, R. M., AND J. OLLASON. 1982. Animal Popula-
tion Dynamics. Chapman and Hall, London and New
York. 80 pp.
WILKINSON, L. 1989. SYSTAT: The System for Statis-
tics. SYSTAT, Incorporated, Evanston, IL. 638 pp.
Florida Scrub Jay: Demography of a Cooperative-
Breeding Bird. Princeton University Press, Prince-
ton, NJ. 406 pp.
WISE, D. H. 1993. Spiders in Ecological Webs. Cam-
bridge University Press, Cambridge, UK and New
York. 328 pp.
ZAP, J. H. 1974. Biostatistical Analysis. Prentice-Hall,
Englewood Cliffs, NJ. 620 pp.

Florida Entomologist 84(3)

Childers et al.: Pesticide Toxicities to Euseius mesembrinus on Florida Citrus 391


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

2Research Institute, Dokki, Egypt

Residual toxicities of registered and selected experimental pesticides used on citrus against Eu-
seius mesembrinus (Dean) (Acari: Phytoseiidae) were compared. A tractor-drawn airblast
sprayer calibrated to deliver 2,338 liters/ha was used to apply pesticides at one or more recom-
mended rates on mature 'Ruby Red' grapefruit trees. Pesticides rated as highly toxic were: azin-
phos-methyl 50WP at 4.48 kg/ha, dicofol 4EC at 7.01 liters/ha, formetanate 92SP at 5.84 kg/ha,
dimethoate 4EC at 5.85 liters/ha, malathion 57EC at 5.85 liters/ha, propargite 6.55EC at 3.51 li-
ters/ha, benomyl 50WP at 1.68 kg/ha + ferbam 76GF at 5.60 kg/ha, ferbam 76GF at 16.81 kg/ha,
carbaryl XLR plus at 18.7 liters/ha + FC435-66 petroleum oil at 46.8 liters/ha, pyridaben 75WP
at 462 g/ha + FC435-66 petroleum oil at 46.8 liters/ha, carbaryl 80S at 11.21 kg/ha, ethion 4EC
at 7.01 liters/ha + FC435-66 petroleum oil at 46.8 liters/ha, benomyl 50WP at 3.36 kg/ha, chlor-
fenapyr 2SC at 1.46 liters/ha, and pyridaben 75WP at 462 g/ha. Pesticides that were moderately
to slightly toxic were: sulfur 80DF at 16.81 kg/ha, abamectin 0.15EC at 731 ml/ha + FC435-66
petroleum oil at 46.8 liters/ha, chlorfenapyr 2SC at 971 ml/ha + FC435-66 petroleum oil at 46.8
liters/ha, FC435-66 petroleum oil at 93.5 liters/ha, and chlorpyrifos 4EC at 5.85 liters/ha. Pesti-
cides that were considered non-toxic were: FC435-66 petroleum oil at 46.8 liters/ha, carbaryl 80S
at 4.48 kg/ha, chlorfenapyr 2SC at 971 ml/ha, chlorpyrifos 4EC at 5.85 liters/ha, fenbuconazole
2F at 292 ml/ha + FC435-66 petroleum oil at 46.8 liters/ha, copper hydroxide 77WP at 4.48 kg
metallic/ha, benomyl 50WP at 3.36 kg/ha, and fenbuconazole 2F at 584 ml/ha. Ferbam 76GF at
16.81 kg/ha, benomyl 50WP + ferbam 76GF, carbaryl 80S at 11.21 kg/ha, carbaryl XLR Plus +
FC435-66 petroleum oil, and benomyl 50WP at 3.36 kg/ha had significantly higher numbers of
missing females from treated leaf surfaces suggesting these products were repellent, irritating,
and/or excitatory to the gravid females.

Key Words: Acaricides, fungicides, insecticides, integrated pest and disease management,
toxicity, non-target arthropods


Toxicidad residual de pesticides registrados y experimentales selectos usados en citricos contra
Euseius mesembrinus (Dean) (Acari: Phytoseiidae) fueron comparados. Una asperjadora de aire
a presi6n y halada por un tractor, calibrada para entregar 2,338 litros/ha fue usado para aplicar
pesticides a una o mas dosis recomendadas en arboles de toronja 'Ruby Red'. Pesticidas clasifi-
cados como altamente t6xicos fueron: azinphos-methyl 50WP a 4.48 kg/ha, dicofol 4EC a 7.01 li-
tros/ha, formetanate 92SP a 5.84 kg/ha, dimethoate 4EC a 5.85 litros/ha, malathion 57EC a 5.85
litros/ha, propargite 6.55EC a 3.51 litros/ha, benomyl sowr a 1.68 kg/ha + ferbam 76GF a 5.60
kg/ha, ferbam 76GF a 16.81 kg/ha, carbaryl XLR plus a 18.7 litros/ha + FC435-66 aceite de pe-
tr61leo a 46.8 litros/ha, pyridaben 75WP a 462 g/ha + FC435-66 aceite de petr61leo a 46.8 litros/ha,
carbaryl 80S a 11.21 kg/ha, ethion 4EC a 7.01 litros/ha + FC435-66 aceite de petr61leo a 46.8 li-
tros/ha, benomyl 50 WP a 3.36 kg/ha, chlorfenapyr 2SC a 1.46 litros/ha, y pyridaben 75WP a 462
g/ha. Pesticidas que fueron moderadamente o levemente t6xicos eran: azufre 80DF a 16.81 kg/
ha, abamectin 0.15EC a 731 ml/ha + FC435-66 aceite de petr61leo a 46.8 litros/ha, chlorfenapyr
2SC a 971 ml/ha + FC435-66 aceite de petr61leo a 46.8 litros/ha, FC435-66 aceite de petr61leo a
93.5 litros/ha, y chlorpyrifos 4EC a 5.85 litros/ha. Pesticidas considerados no t6xicos fueron:
FC435-66 aceite de petr61leo a 46.8 litros/ha, carbaryl 80S a 4.48 kg/ha, chlorfenapyr 2SC a 971
ml/ha, chlorpyrifos 4EC a 5.85 litros/ha, fenbuconazole 2F a 292 ml/ha + FC435-66 aceite de pe-
tr61leo a 46.8 litros/ha, hidr6xido de cobre 77WP a 4.48 kg metalico/ha, benomyl 50WP a 3.36 kg/
ha, y fenbuconazole 2F a 584 ml/ha. Ferbam 76GF a 16.81 kg/ha, benomyl 50WP + ferbam 76GF,
carbaryl 80S a 11.21 kg/ha, carbaryl XLR Plus + FC435-66 aceite de petr61leo, y benomyl 50WP
a 3.36 kg/ha tuvieron numerous significativamente mas altos de hembras ausentes de superficies
de hojas tratadas sugiriendo que estos products fueron repelentes, irritantes, y/o excitatorio a
las hembras gravidas.

Florida Entomologist 84(3)

Citrus is a multi-billion dollar agricultural
business in Florida that annually provides 70-
80% of the total United States production (Anon-
ymous 1996). Of this, 85% of the Florida crop is
used in processing (i.e., juice, sections, pulp) with
the balance for fresh market. The pest mite com-
plex on Florida citrus is diverse and includes spe-
cies in four acarine families: Eriophyidae,
Tetranychidae, Tarsonemidae, and Tenuipalpidae
(Childers 1994). The citrus rust mite Phyllocop-
truta oleivora (Ashmead), the pink citrus rust
mite Aculops pelekassi (Keifer) (Eriophyidae),
and several spider mite species, primarily the
Texas citrus mite, Eutetranychus banksi (McGre-
gor) are important. Approximately 171 million
dollars are spent annually by farmers in Florida
for chemical control of these mites including the
combined costs of chemicals and application
equipment, based on estimates for Central, East
Coast and Southwest Florida (Muraro & Hebb
1997, Muraro et al. 1997a, Muraro et al. 1997b).
Historically, this has been one of the largest com-
modity markets for acaricide usage within the
United States with 2-3 applications per hectare
per year on 346,823 hectares of trees (Childers
1994). Citrus rust mites and the fungal pathogen
greasy spot, Mycosphaerella citri Whiteside, are
recognized as key fruit and foliar pests on Florida
citrus, respectively. Control recommendations
currently rely solely on pesticides for both key
pests (Childers et al. 2000a, Roberts & Timmer
2000). Increasing urbanization, the public's con-
cern over pesticide use on food products, in-
creased application and chemical costs,
sustaining competitive food exports to foreign
markets and compliance with the Food Quality
Protection Act of 1996 dictate the need for alter-
native control strategies.
Because of climatic conditions that exist in
Florida, the fungal pathogens: e. g., greasy spot,
melanose, Diaporthe citri Wolf, citrus scab, Elsi-
noe fawcettii Bitancourt and Jenkins, alternaria
brown spot, Alternaria sp. and postbloom fruit
drop disease, Colletotrichum acutatum J. H. Sim-
monds, are important disease problems for many
citrus growers. Current fungicides recommended
for these diseases are limited to ferbam, benomyl,
ferbam + benomyl, copper formulations applied
alone or in combination with petroleum oil, or pe-
troleum oil applied alone (Roberts & Timmer
2000, Timmer 2000a,b,c, McMillan et al. 2000).
Arthropod predators and parasitoids are the
most important naturally occurring biological
control agents of arthropods in most agroecosys-
tems (Croft 1990). Predacious mites are recog-
nized as highly important in regulating
phytophagous mites on citrus worldwide (Keetch
1972, Ferragut et al. 1987, Papacek & Smith
1992, McMurtry & Croft 1997) and Florida citrus
has a rich fauna of both predacious and other ben-
eficial mites (Muma 1975). On-going research has

identified species within several mite families
that have potential in suppressing citrus rust
mites and spider mites (Childers 1994, Childers
& Abou-Setta 1999; C. C. C., unpublished data).
However, little information on comparative toxic-
ities of registered or experimental pesticides to
predacious mites on Florida citrus has been avail-
able. Such information is essential in developing
an effective integrated pest and disease manage-
ment program.
This study was initiated to assess direct and
indirect residual toxicities of different pesticides
either registered for use on Florida citrus or being
developed for registration as insecticides, acari-
cides or fungicides. The pesticides were applied
and weathered in the field against the predacious
mite, Euseius mesembrinus (Dean), a prevalent
phytoseiid species that feeds on spider mites and
other mite species on Florida and Texas citrus
(Abou-Setta et al. 1991; C. C. C., unpublished
data). Different pesticides were applied and
weathered in the field to assess their impact on
gravid females, oviposition, and survival of eclos-
ing larvae.


Field Application

Four series of field applications were com-
pleted between August and September 1997 in a 4
hectare block of 'Ruby red' grapefruit located at
the Citrus Research and Education Center in
Lake Alfred, Florida. The treatment trees were
healthy, vigorous and measured 3.2 to 3.9 m tall
and 3.9 to 5.2 m in diameter. Trees were spaced
7.47 m within the row and 8.11 m between rows
(= 187 trees/ha). Different trees were selected for
each of the series of pesticide treatments with a
minimum of four trees separating single tree
treatments within the row and with two buffer
rows between treatment rows.
Treatments were applied using a tractor-
drawn FMC 352 airblast sprayer beginning each
morning after the citrus leaves had dried. The
sprayer was calibrated to deliver 2,338 liters of
spray/ha. Each single tree treatment was sprayed
while traveling at 2.4 km/h. Tractor speed was
properly adjusted several tree spacings ahead of
the sample tree before engaging the sprayer on ei-
ther side within the rows. The pesticides, applica-
tion dates, water pH, formulations, and rates per
hectare are shown in Table 1. FC435-66 repre-
sents a medium, narrow-range petroleum oil with
a mid-distillation temperature of 224C (= 435'F)
that meets the designated Florida citrus stan-
dards established by Simanton & Trammel (1966).
T-Mulz (Harcros Chemicals, Inc., Kansas City,
KS), a non-ionic surfactant, was mixed at 10 ml
per liter of petroleum oil as the emulsifier. Pesti-
cides tested are registered for use on Florida

September 2001


Date of application Pesticide Pesticide Rate used in
Series and water pH common name and use" Manufacturer trade name Formulationb 2.34 k liters/ha

I 11 August-pH 7.5 1. Copper hydroxide F Griffin Corp. Kocide 101 77WP 4.48 kg metallic

2. Copper sulfate

3. Fenbuconazole
4. Fenbuconazole (RH-
7592) + petroleum oil
5. Benomyl

6. Ferbam

7. Benomyl
+ ferbam
8. Chlorpyrifos

9. Sulfur

10. Untreated

II 18 August-pH 7.3

III 19 August-pH 7.5

1. Pyridaben
2. Pyridaben
+ petroleum oil
3. Abamectin
+ petroleum oil
4. Chlorfenapyr

5. Chlorfenapyr
+ petroleum oil
6. Petroleum oil
7. Petroleum oil
8. Untreated

1. Carbaryl

Valdosta, GA
F Cuproquim Corp.
Memphis, TN
F Rohm and Haas Co.
Philadelphia, PA

, A, I Exxon Co., Houston, TX
F E. I. DuPont de Nemours
Wilmington, DE
F UCB Chemical
Smyrna, GA
I Dow Elanco
Indianapolis, IN
A BASF Research
Triangle Park, NC


Greensboro, NC
American Cyanamid Co.
Princeton, NJ

F, A, I
A, I
F, A, I

F, A, I
F, A, I
F, A, I

Rhone Poulenc Research
I Triangle Park, NC

Copper 53


Orchex 796


+ Ferbam


+ Orchex 796
+ Orchex 796

+ Orchex 796
Orchex 796
Orchex 796


(50% metallic)
(53% metallic)







4.48 kg metallic

584 ml

292 ml
46.8 liters
3.36 kg

16.81 kg

1.68 kg
5.60 kg
5.85 liters

16.81 kg

462 g
462 g
46.8 liters
731 ml
46.8 liters
971 ml

971 ml
46.8 liters
46.8 liters
93.5 liters

4.48 kg

A = acaricide, F = fungicide, I = insecticide.
'WP = wettable powder, F = flowable, GF = granular flowable, EC = emulsifiable concentrate, SC
dispersible flowable.

= soluble concentrate, S, SP = soluble powder, FC 435-66 = Florida citrus 435 oil (Simanton and Trammel 1966), DF



Date of application Pesticide Pesticide Rate used in
Series and water pH common name and use" Manufacturer trade name Formulationb 2.34 k liters/ha

2. Carbaryl I Sevin 80S 11.21 kg
3. Azinphos-methyl I Bayer Corp. Guthion 50WP 4.48 kg
Kansas City, MO
4. Formetanate A, I Agro Evo, Carzol 92SP 1.12 kg
Wilmington, DE
5. Carbaryl I Rhone Poulenc Sevin XLR Plus 41.2% 18.7 liters
+ petroleum oil F, A, I + Orchex 796 AI 46.8 liters
6. Dimethoate I Platte Chemical Co. Dimethoate 400 5.85 liters
Fremont, NE
7. Malathion I Platte Chemical Co. Malathion 57EC 5.85 liters
Fremont, NE
8. Dicofol A Platte Chemical Co. Dicofol 4EC 7.01 liters
Fremont, NE
9. Propargite A Uniroyal Chemical Comite 6.55EC 3.51 liters
Co., Middlebury, CT
10. Untreated -

IV 8 September pH N/A 1. Benomyl F Benlate 50WP 3.36 kg
2. Ferbam F Ferbam 76GF 16.81 kg
3. Chlorpyrifos I Lorsban 4EC 5.85 liters
4. Ethion A, I FMC Corp. Ethion 4EC 7.01 liters
+ petroleum oil Philadelphia, PA
5. Chlorfenapyr A, I Alert 2SC 1.46 liters
6. Chlorfenapyr A, I Alert 2SC 971 ml
+ petroleum oil F, A, I + Orchex 796 FC435-66 46.8 liters
7. Formetanate A, I Carzol 92SP 1.12 kg
8. Dicofol A Platte Chemical Co. Dicofol 4EC 7.01 liters
Freemont, NE
9. Untreated -

A = acaricide, F = fungicide, I = insecticide.
'WP = wettable powder, F = flowable, GF = granular flowable, EC = emulsifiable concentrate, SC
dispersible flowable.

= soluble concentrate, S, SP = soluble powder, FC 435-66 = Florida citrus 435 oil (Simanton and Trammel 1966), DF

Childers et al.: Pesticide Toxicities to Euseius mesembrinus on Florida Citrus

citrus except chlorfenapyr (4-Bromo-2-(4-chlor-
phenyl)-l(ethoxymethyl)-5-(trifluoromethyl) pyr-
role-3-carbonitrile), an experimental acaricide of
American Cyanamid Co., Princeton, NJ and fen-
buconazole (RH-7592), alpha [2-(4-chlorophenyl)
trile, an experimental fungicide of Rohm and
Haas Co., Philadelphia, PA. The acaricides, insec-
ticides, or fungicides are used to control one or
more arthropod or fungal disease pests included
in the Florida Pest Management Guide (Childers
et al. 2000a, b; Browning et al. 2000; Roberts &
Timmer 2000). Daily maximum-minimum air
temperatures and rainfall data during the test in-
tervals between application and completion of
leaf sampling for each series of pesticides are
listed in Table 2.

Field Sampling

Sixteen or more hardened spring flush leaves
were collected at random from the outer exposed
canopy around each sample tree at waist to chest
height (1.2 to 1.6 m) one and 4 days after each se-
ries of pesticide treatments was applied during
1997. Leaves were always collected first from the
unsprayed check trees for each treatment series.
Disposable rubber gloves were changed between
treatments both in the field and laboratory to


Temperature C
Series Date Max Min (mm)

I 11 August 34 22 35
12 34 22
13 36 24
14 37 23 6
15 34 23
16 35 22 25
17 34 22
II 18 36 22
III 19 36 23
20 36 26
21 35 24
22 36 22 6
23 34 19
24 34 21
25 36 20
IV 8 September 33 17
9 33 17
10 34 20
11 34 20 3
12 33 21
13 34 19
14 36 20

avoid potential contamination. Individual dry
leaves were collected into a paper bag by picking
the leaf at the base of the petiole without touching
the leaf surface. Each paper bag was then placed
on the floor of the air conditioned vehicle out of di-
rect sunlight. Sampling in the field and travel
time required less than 1 h before returning to
the laboratory for processing.

Laboratory Preparation

Untreated check leaves were always processed
first. Leaves from each treatment were removed
from a paper bag and briefly placed on a clean pa-
per surface. Leaves for that treatment were then
individually selected and prepared for single
whole leaf arenas. Contact with treated leaf sur-
faces was avoided as much as possible. Six leaves
(= 6 treatment replicates) were selected per treat-
ment based on size, vigor and minimal leaf blem-
ishes. Individual leaf arenas were prepared with
the lower leaf surface facing up based on estab-
lished methods for rearing phytoseiid mites
(Abou-Setta & Childers 1987a). An aluminum foil
sheet was placed beneath the foam padding to
prevent contact of water with the plastic con-
tainer and to avoid potential contamination. A fil-
ter paper was placed over the foam pad and the
leaf arena was then placed over the filter paper.
The sides and petiole areas of each whole leaf
were covered with absorbent cotton stripping to
keep the leaf in position and provide a non-toxic
wet barrier to minimize escape of the gravid fe-
males. Treated leaf arenas were maintained in
the laboratory between 26 and 28C and held in
open boxes to assure adequate ventilation.

Mite Cultures

Euseius mesembrinus was collected from Cen-
tral Florida citrus groves in the Winter-Spring of
1997, maintained in culture on leaf arenas and
fed ice plant, Malephora crocea (Jacquin) pollen
(Abou-Setta & Childers 1987b).
Large numbers of both E. mesembrinus males
and females were transferred to several new un-
treated rearing leaf arenas and held for 24 h to
deposit a cohort of eggs + 12 h old. The following
day, all motile mites were removed from the leaf
arenas and the eggs were allowed to develop.
Gravid E. mesembrinus females were then col-
lected from the arenas after 9-10 days (Abou-
Setta & Childers 1987b). Cotton fibers were
placed on the surface of each leaf for use as an ovi-
position substrate along with a piece of black con-
struction paper about 4 mm wide by 8-10 mm long
that served as a refuge for aggregation. A total of
10 gravid females were transferred individually
using a 5-0 sable brush directly onto the black
construction paper of each replicated arena to
avoid contaminating the brush with pesticides.

Florida Entomologist 84(3)

The mites were checked to ensure that they were
not injured during transfer and any that ap-
peared injured were immediately replaced.

Laboratory Assessment

Each arena was examined 72 h after infestation
to determine mortality of gravid females. Three
categories were recorded for adults: dead, alive, or
missing. A mite was considered dead if it was un-
able to move forward. In addition, number of live
eggs per arena was recorded after 72 h. Eggs were
recorded as live when their size, opaque color and
oblong-oval shape were consistent. Live larvae or
protonymphs per arena were recorded after 72 h.

Statistical Analysis

All data on surviving or missing adults, ovipo-
sition and surviving immatures were analyzed in
each experiment by analysis of variance (ANOVA)
and means were separated using LSD (GLM pro-
cedures, SAS Institute 1991). Means and stan-
dard errors reported here were calculated using
non-transformed data.


Maximum air temperatures in the field during
the 1997 experiment ranged from 33 to 37C and
two significant rains of 35 and 25 mm occurred on
11 and 16 August, respectively, during evaluation
of series I pesticides (Table 2). The 11 August rain
occurred about 2 h after the last treatment appli-
cations (treatments 3 and 4 in series I).

Gravid Females

Pesticides that resulted in less than 30% sur-
vival of gravid female E. mesembrinus included:
azinphos-methyl, dicofol, malathion, propargite,
formetanate, dimethoate, propargite, benomyl +
ferbam, ferbam, carbaryl XLR + petroleum oil,
pyridaben + petroleum oil, cabaryl 80S at 11.21
kg/ha, ethion + petroleum oil, benomyl, chlor-
fenapyr and pyridaben (Table 3). Pesticides that
remained highly toxic to gravid females when ex-
posed to 4-day-old field treated leaves were: azin-
phos-methyl, dicofol, formetanate, ferbam and
pyridaben + petroleum oil. Use of formetanate in
three different treatment regimes in a field exper-
iment on 'Tahiti' limes significantly reduced pop-
ulations of Typhlodromalus peregrinus (Muma)
(Acari: Phytoseiidae) for over 2 months following
the last applications (Childers & Abou-Setta
1999). Subsequent feeding injury from elevated
populations of citrus red mite, Panonychus citri
(McGregor) in this experiment resulted in eco-
nomic loss to the grower. Most of the same pesti-
cides were shown to be toxic to Euseius hibisci
(Chant) in California by Jeppson et al. (1975).

Pesticides listed in Table 4 had significantly
higher numbers of missing E. mesembrinus fe-
males from treated leaf surfaces compared to the
untreated checks and other treatments tested.
Mites could not be found on the treated leaf are-
nas, cotton pad strips surrounding each treated
leaf surface or on the water saturated sponge be-
neath the leaf. The gravid females were presumed
to have escaped the arena since no visible cadav-
ers or injured mites remained. We suspect that
these products were repellent, irritating, and/or
excitatory to gravid females.


Pesticides that were not toxic to gravid females
but resulted in a 50% or greater reduction in egg
production after one day post treatment included:
sulfur and benomyl (Table 3). The untreated
check from series I had low numbers of eggs pro-
duced in the 4 day post spray evaluation. Egg pro-
duction was significantly lower compared to
several other treatments in that series except
copper hydroxide, chlorpyrifos, or copper sulfate.
None of these treatments had significantly lower
egg production in the 1 day postspray series (Ta-
ble 3).

Larval Survival

Pesticides that were not toxic to gravid females
but resulted in a 50% or less reduction in larvae
included: abamectin + FC435-66 petroleum oil af-
ter one day and benomyl or chlorfenapyr +
FC435-66 petroleum oil after 4 days compared
with the untreated checks.

Comparative Toxicities

Our data often showed differential toxicity to
various life stages of E. mesembrinus. Therefore,
a simple method to compare toxicities between
pesticides was used where (x number of surviv-
ing gravid females) x (x number of live eggs pro-
duced/arena) x (x number of live larvae) were
multiplied. The lower the value obtained the more
toxic was the pesticide. Based on these criteria,
pesticides evaluated in this study were deter-
mined to be highly toxic with values below 200,
moderately toxic with values between 200 and
400 and slightly to non-toxic with values greater
than 400 (Table 5).
Phytoseiid mites in the genus Euseius are rec-
ognized as facultative pollen feeders (McMurtry
1992). However, E. hibisci (Chant) was shown to
feed on avocado leaf sap but not on lemon using
radioactive phosphoric acid (Porres et al. 1975).
Subsequent studies failed to indicate whether
this feeding was restricted to the larval stage or
all motile stages of the mite. Abou-Setta &
Childers (1987b) reported that E. mesembrinus

September 2001


1 day 4 days

Rate Surviving Live Live Surviving Live Live
Series Pesticide formulation per ha 2 2 eggs immatures 2 2 eggs immatures

I 1. Copper hydroxide 77WP

2. Copper sulfate 98%

3. Fenbuconazole 2F
4. Fenbuconazole 2F
+ petroleum oil FC435-66
5. Benomyl 50WP
6. Ferbam 76GF
7. Benomyl 50WP
+ ferbam 76GF
8. Chlorpyrifos 4EC
9. Sulfur 80DF
10. Untreated

II 1. Pyridaben 75WP
2. Pyridaben 75WP
+ petroleum oil FC435-66
3. Abamectin 0.15EC
+ petroleum oil FC435-66
4. Chlorfenapyr 2SC
5. Chlorfenapyr 2SC
+ petroleum oil FC435-66
6. Petroleum oil FC435-66
7. Petroleum oil FC435-66
8. Untreated

III 1. Carbaryl 80S
2. Carbaryl 80S
3. Azinphos-methyl 50WP
4. Formetanate 92SP
5. Carbaryl-XLR plus 41.2% ai
+ petroleum oil FC435-66

4.48 kg
4.48 kg
584 ml
292 ml
46.8 liters
3.36 kg
16.81 kg
1.68 kg
5.60 kg
5.85 liters
16.81 kg

462 g
462 g
46.8 liters
731 ml
46.8 liters
971 ml
971 ml
46.8 liters
46.8 liters
93.5 liters

4.48 kg
11.21 kg
4.48 kg
1.12 kg
18.7 liters
46.8 liters

9.6 + 0.2 a 12.3 + 1.9 bc 6.3 + 0.9 abcd

9.1 + 0.4 a 13.5 + 0.9 bc 6.9 + 0.7 ab

8.8+ 0.4 a 21.3 +2.4 a 7.1+ 0.5 a
8.9 + 0.4 a 14.8 + 1.0 abc 5.5 + 1.2 abcde

9.6 + 0.2 a
2.6+ 1.3 b
2.9 + 1.2 b

9.4 + 0.3 a
8.5 0.6 a
9.0 + 0.5 a

10.6 + 2.3 c 7.6 1.0 a
2.8 + 1.7 e 3.8 + 0.9 e
1.9 1.0 e 3.9 0.9de

15.4 + 2.5 abc 4.4 + 0.7 bcde
5.9 2.0 d 4.1 + 0.8 cde
15.8 + 1.1 ab 6.9 + 1.2 abc

7.1+ 0.9 ab 6.4 + 1.8 bc 3.6 + 0.7 ab
5.4 + 1.5 b 4.8 + 1.7 c 1.3 + 0.6 c

8.8 + 0.3 a 11.1 + 1.8 ab 2.5 + 0.6 bc

8.3 + 0.8 a 14.6 + 2.5 a 4.9 + 1.1 ab
7.0 + 1.4 ab 11.0 + 2.3 ab 4.0 + 1.3 b

8.5 + 0.9 a
8.9 + 0.4 a
9.1 + 0.2 a

9.3 + 0.4 a
4.5 0.8 b
0 0 Od
0 0 Od

17.1 +2.9 a 3.4 0.8 b
12.4 2.2 a 3.4 + 1.0 b
12.6 +1.1 a 6.3 +0.6 a

10.0 1.7 a 6.0+ 0.8 a
2.1 l.lb 4.9+ 0.7 a
0 + 0 c 0 + 0 c
0 c 0 0Oc

3.8 + 0.8 b 2.9 0.9b 2.8 0.8b

7.5 + 0.6 ab 2.3 + 1.0 d 4.0 + 0.3 bc

7.9 + 1.7 a 3.4 + 1.6 cd 3.5 + 0.6 bc

9.0+ 0.5 a 13.8 + 2.9 a 7.4 + 1.0 a
8.6 + 0.7 a 13.8 + 1.9 a 3.4 + 0.9 bc

9.8 + 0.2 a
2.4 + 1.1 c
5.9+ 1.4 b

8.1+ 0.5 a
9.5 + 0.3 a
8.1 + 0.4 a

2.0 + 1.1 d
1.5 + 0.8 d
4.3 + 2.0 cd

7.4 + 1.9 bc
11.4 + 1.3 ab
3.6 + 1.3 cd

4.8 + 0.8 b
2.9 + 0.9 c
4.4 + 0.4 bc

3.8 + 0.8 bc
5.0 + 0.7 ab
3.9 + 0.7 bc

9.3+ 0.3 a 3.0+ 0.7 bc 6.4 1.1 d
8.4+ 0.5 a 0.4 + 0.4 d 7.6 + 0.8 cd

9.5 +0.2 a 3.8 0.6ab 6.4 0.8 d

9.1 +0.5 a 6.0 +0.7 a 12.3 + 1.0 ab
9.0 + 0.2 a 3.0 + 0.7 bc 10.1 + 0.7 abc

8.9 + 0.3 a
9.5 + 0.2 a
9.6 + 0.2 a

8.8 + 0.3 a
8.0 + 0.6 ab

4.6 + 0.8 ab
4.9 + 1.0 ab
4.4 + 0.7 ab

13.0 + 1.9 a
7.0 + 1.5 a

7.9 + 0.6 cd
9.5 + 1.0 bc
3.9 + 0.8 e

4.9 + 0.4 bc
4.1 + 0.6 c
0 + Oe
0 + Oe

9.3 0.3 a 13.8 1.6 a 4.9 + 0.4 bc

Means within each column within each series followed by the same letter are not significantly different (P > 0.05).


1 day 4 days

Rate Surviving Live Live Surviving Live Live
Series Pesticide formulation per ha 9 9 eggs immatures 2 9 eggs immatures

6. Dimethoate 400 5.85 liters 0 + 0 d 0 + 0.1 c 0.6 + 0.6 c 9.0 + 0.3 a 12.5 + 0.6 a 7.1+ 1.0 ab
7. Malathion 57EC 5.85 liters 2.3+ 1.1 c 0 + 0 c 0 + 0 c 9.8 + 0.2 a 17.8 + 1.0 a 7.5 + 0.7 a
8. Dicofol 4EC 7.01 liters 1.0+ 0.7 cd 0.1 + 0.1 c 0.5 + 0.5 c 2.6 + 1.5 c 2.9 + 1.9 cd 2.1+ 1.1 d
9. Propargite 6.55EC 3.51 liters 0 + 0 d 0 + 0 c 0.5 + 0.4 c 6.3 + 1.3 b 6.6 + 1.9 b 6.0 + 0.6 abc
10. Untreated 9.5 + 0.3 a 12.5 + 2.0 a 4.9 + 0.9 a 9.5 + 0.3 a 17.3 + 0.8 a 5.6 + 0.5 abc

IV 1. Benomyl 50WP 3.36 kg 6.3 + 1.0 bcd 5.9 + 2.9 e 3.9 + 1.2 ab 9.4 + 0.3 ab 11.8 + 2.6 cd 4.3 + 0.6 bc
2. Ferbam 76GF 16.81 kg 5.8 + 0.9 cd 7.3 + 1.9 cde 4.3 + 0.4 a 3.6 + 0.9 d 6.3 + 2.3 e 5.4 + 0.9 b
3. Chlorpyrifos 4EC 5.85 liters 7.8 + 0.4 abc 12.5 + 1.5 abc 4.1 + 0.8 ab 8.9 + 0.2 ab 16.3 + 1.1 bc 4.8 + 0.8 b
4. Ethion 4EC 7.01 liters 5.4 + 1.3 cd 6.8 + 3.0 de 2.4 + 0.7 b 7.1+ 1.1 bc 7.9 + 2.2 de 2.3 + 0.8 d
+ petroleum oil FC435-66 46.8 liters
5. Chlorfenapyr 2SC 1.46 liters 5.3+ 1.5 d 9.3 + 3.4 cde 3.1+ 0.9 ab 6.0+ 1.6 cd 7.5 + 1.9 de 4.1+ 1.0 bcd
6. Chlorfenapyr 2SC 971 ml 7.8 + 1.2 abc 11.1 + 2.4 bcd 3.3 + 0.8 ab 8.9 + 0.4 ab 11.0 + 1.2 cd 2.8 + 1.2 cd
+ petroleum oil FC435-66 46.8 liters
7. Formetanate 92SP 1.12kg 0+ 0e 00f 00c 0+ 0e 0+ 0f 0+ 0e
8. Dicofol4EC 7.01 liters 0+0e 0+0f 0+0c 0+0e 0+0f 0+0e
9. Untreated 9.0 + 0.3 ab 15.9 + 1.7 ab 4.8 + 0.6 a 9.9+ 0.1 a 23.8 + 2.5 a 10.8 + 0.9 a

Means within each column within each series followed by the same letter are not significantly different (P > 0.05).

Childers et al.: Pesticide Toxicities to Euseius mesembrinus on Florida Citrus 399


Missing females

Series Pesticide Formulation Rate per ha 1 day post spray 4 days post spray

I Ferbam 76WP 16.81 kg 6.3 + 1.2 a 4.5 + 1.1 a
Benomyl 50WP 1.68 kg 3.3 + 1.2 a
+ ferbam 76WP 5.60 kg 5.1 + 1.2 a
Untreated 0.9 + 0.5 b 0.5 + 0.2 b

III Carbaryl 80S 11.21 kg 1.4 + 0.6 a 0.4 + 0.2 a
Carbaryl XLR Plus 18.7 liters
+ petroleum oil FC435-66 46.8 liters 2.3 + 0.9 a 0 + 0 a
Untreated 0.3 + 0.3 b 0.3 + 0.2 a

IV Benomyl 50WP 3.36 kg 1.4 + 0.4 a 0.5 + 0.3 bc
Ferbam 76WP 16.81 kg 1.3 + 0.6 ab 4.4 + 0.9 a
Chlorpyrifos 4EC 5.85 liters 0.8 + 0.4 abc 1.0 + 0.3 b
Untreated 0.4 + 0.2 bc 0.1 + 0.1 c

Means within each column within each series followed by the same letter are not significantly different (P > 0.05).

was able to develop to the protonymph stage
when no food sources were available except the
citrus leaf substrate. In this study, chlorfenapyr +
FC435-66 oil, abamectin + FC435-66 oil, or
FC435-66 oil applied alone at 46.8 or 93.5 liters/
ha had lower larval survival rates compared with
the other treatments (Table 3). These data sug-
gest that reduced larval survival was caused by
either interference with larval feeding attempts
through the petroleum oil film covering the leaf
surface or by direct or indirect toxicity of the pes-
ticides or combinations with acaricides. It is
known that penetration of abamectin into the
wax layers of both citrus leaves and fruit in-
creases considerably when combined with petro-
leum oil under field conditions (Dybas 1990). The
use of petroleum oil provided substantial exten-
sion of residual activity of abamectin versus ab-
amectin applied alone in controlling citrus rust
mite (C. C. C., unpublished data).
Understanding the toxic effects of field weath-
ered pesticides against key predacious mite spe-
cies is important for all commodities. Climatic
conditions in Florida are characterized by warm
temperatures, high humidity, and moderate to
high annual rainfall. During the summer (i.e.,
May through October), humidity conditions at
night approach 100% and result in long hours of
leaf and fruit wetness that often extend into early
to mid-morning. In addition, afternoon rain show-
ers frequently occur, especially between May and
October. Because of these conditions, accelerated
degradation of many pesticides has been shown
(Nigg et al. 1983). For example, organophosphate
pesticide residues on citrus degrade much more
quickly on Florida citrus compared with residues
in California (Nigg et al. 1977, Thompson et al.

1979, Nigg & Stamper 1981). Despite such harsh
environmental conditions, several of the pesti-
cides evaluated in this study maintained rela-
tively long-termed residual toxicities to E.
mesembrinus either directly by reducing female
survival or indirectly by impacting fecundity
rates or larval survival as shown in this study.
This study showed the potentially disruptive ef-
fects of using ferbam and benomyl on Florida cit-
rus by adversely affecting the predacious mite
Euseius mesembrinus.
The results of this study provide a comparison
of direct and indirect toxic effects by various pes-
ticides to E. mesembrinus under field conditions.
This information is step one in a process of iden-
tifying the possible disruptive impact of specific
pesticides used on Florida Citrus. Previous stud-
ies have shown that toxicity of certain pesticides
to populations of predacious mites and conse-
quent reductions in their effectiveness against
phytophagous mite pests (Childers & Enns 1975;
Childers & Abou-Setta 1999). Longer term field
studies are in progress to identify possible subtle
or delayed negative effects of one or more pesti-
cides used in citrus for arthropod and fungal dis-
ease control.


The authors thank S. J. Johnson, Dept. of Entomol-
ogy, Louisiana State University, Baton Rouge; J. E.
Pena, TREC, University of Florida, Homestead; H. N.
Nigg, and J. P. Michaud, CREC, University of Florida,
Lake Alfred for reviewing this manuscript. This re-
search was supported by the Florida Agricultural Ex-
periment Station, and approved for publication as
Journal Series No. R-08031.

Florida Entomologist 84(3)


Calculated Toxicity
Treatment Series ratings rating

Benomyl + ferbam
Carbaryl XLR
+ petroleum oil
+ petroleum oil
Carbaryl (11.21 kg)
+ petroleum oil

+ petroleum oil

+ petroleum oil

+ petroleum oil

Petroleum oil


Petroleum oil
Carbaryl (4.48 kg)
Untreated (III)
Untreated (IV)
Untreated (II)
+ petroleum oil
Copper hydroxide
Copper sulfate
Untreated (I)

III 0 Highly toxic
IV 0 Highly toxic
III 0 Highly toxic
IV Highly toxic
III 0 Highly toxic
III 0 Highly toxic
III 0 Highly toxic
III 0.05 Highly toxic
I 21 Highly toxic
I 28 Highly toxic

III 31 Highly toxic

II 34 Highly toxic
III 46 Highly toxic
IV 88 Highly toxic

IV 145 Highly toxic
IV 153 Highly toxic
II 164 Highly toxic
IV 182 Highly toxic
I 206 Moderately
to slightly
II 244 to slightly
IV 286 to slightly
II 308 to slightly
II 375 Moderately
to slightly
IV 400 Moderately
to slightly
II 494 Non-toxic



September 2001


ABOU-SETTA, M. M., AND C. C. CHILDERS. 1987a. A mod-
ified leaf arena technique for phytoseiid or tetrany-
chid mite rearing for biological studies. Florida
Entomol. 70: 245-248.
ABOU-SETTA, M. M., AND C. C. CHILDERS. 1987b. Biol-
ogy of Euseius mesembrinus (Acari: Phytoseiidae):
Life tables on ice plant pollen at different tempera-
tures with notes on its behavior and food range. Exp.
Appl. Acarol. 3: 123-130.
AND H. W. BROWNING. 1991. Comparative morphol-
ogy and reproductive compatibility between popula-
tions of Euseius mesembrinus (Acari: Phytoseiidae)
from Florida and Texas. Exp. Appl. Acarol. 10: 213-220.
ANONYMOUS. 1996. Commercial Citrus Inventory 1996.
Florida Agric. Statistics Serv., Florida Dept. Agric.
Cons. Serv., Orlando, FL.
C. W. MCCOY. 2000. Other insect pests in 2000 Flor-
ida Citrus Pest Management Guide. Fact Sheet
ENY-605. SP 43. Univ. Florida Coop. Ext. Serv.,
IFAS, Gainesville, FL.
CHILDERS, C. C. 1994. Biological control of phytopha-
gous mites on Florida citrus utilizing predatory ar-
thropods, pp. 255-288 in D. Rosen, F. Bennett and J.
Capinera [eds] Intercept. Andover, United Kingdom.
CHILDERS, C. C., AND M. M. ABOU-SETTA. 1999. Yield re-
duction in'Tahiti' lime from Panonychus citri feeding
injury following different pesticide treatment re-
gimes and impact on the associated predacious
mites. Exp. Appl. Acarol. 23: 1-13.
CHILDERS, C. C., AND W. R. ENNS. 1975. Field evalua-
tion of early season fungicide substitutions on tet-
ranychid mites and the predators Neoseiulus fallacis
and Agistemus fleschneri in two Missouri apple or-
chards. J. Econ. Entomol. 68: 719-724.
AND P. A. STANSLY. 2000a. Citrus rust mites in 2000
Florida Citrus Pest Management Guide. Fact Sheet
ENY-603, SP 43. Univ. Florida Coop. Ext. Serv.,
IFAS. Gainesville, FL.
AND P. A. STANSLY. 2000b. Spider mites in 2000 Flor-
ida Citrus Pest Management Guide. Fact Sheet
ENY-602, SP 43. Univ. Florida Coop. Ext. Serv.,
IFAS. Gainesville, FL.
CROFT, B. A. 1990. Arthropod biological control agents
and pesticides. John Wiley & Sons. New York, NY.
DYBAS, R. A. 1990. Abamectin use in crop protection, pp.
287-310. In W. C. Campbell [ed] Ivermectin and Ab-
amectin. Springer-Verlag. New York, NY.
AND R. LABORDA. 1987. Influence of food and temper-
ature on development and oviposition of Euseius
stipulatus and Typhlodromus phialatus (Acari: Phy-
toseiidae). Exp. Appl. Acarol. 3: 317-329.
JESSER, AND H. G. JOHNSON. 1975. Toxicity of citrus
pesticides to some predaceous phytoseiid mites. J.
Econ. Entomol. 68: 707-710.
KEETCH, D. P. 1972. Ecology of the citrus red mite, Panony-
chus citri (McGregor), (Acari: Tetranychidae) in South
Africa. 3. The influence of the predaceous mite,Ambly-
seius l '. .....i - ..........' addoensis van der Merwe &
Ryke. J. Entomol. Soc. South Africa. 35: 69-79.

Childers et al.: Pesticide Toxicities to Euseius mesembrinus on Florida Citrus

L. W. TIMMER. 2000. Postbloom fruit drop in 2000
Florida Citrus Pest Management Guide. Univ. Flor-
ida Coop. Ext. Serv., IFAS Fact Sheet.
McMURTRY, J. A. 1992. Dynamics and potential impact
of 'generalist' phytoseiids in agroecosystems and
possibilities for establishment of exotic species. Exp.
Appl. Acarol. 14: 371-382.
MCMURTRY, J. A., AND B. A. CROFT. 1997. Life-styles of
phytoseiid mites and their roles in biological control.
Annu. Rev. Entomol. 42: 291-321.
MUMA, M. H. 1975. Mites associated with citrus in Flor-
ida. Agric. Exp. Sta. Bull. 640A. IFAS. Univ. Florida.
MURARO, R. P., AND J. W. HEBB. 1997. Budgeting costs
and returns for Indian River citrus production 1996-
97. Economic Inform. Rep. 97-7. Food Res. Econ.
Dept., Univ. Florida. 34 pp.
MURARO, R. P., T. W. OSWALT, AND H. M. STILL. 1997a.
Budgeting costs and returns for central Florida cit-
rus production 1996-97. Economic Inform. Rep. 97-5.
Food Res. Econ. Dept., Univ. Florida. 32 pp.
MURARO, R. P., R. E. ROUSE, AND F. M. ROKA. 1997b.
Budgeting costs and returns for southwest Florida
citrus production 1996-97. Economic Inform. Rep.
97-6. Food Res. Econ. Dept., Univ. Florida. 41 pp.
NIGG, H. N., J. A. HENRY, AND J. H. STAMPER 1983. Re-
gional behavior of pesticides in the United States.
Residue Reviews 85: 257-276.
NIGG, H. N., AND J. H. STAMPER. 1981. Comparative dis-
appearance of dioxathion, malathion, oxydemetonme-
thyl and dialifor from Florida citrus leaf and fruit
surfaces. Arch. Environ. Contam. Toxicol. 10: 497-504.
BROOKS. 1977. Worker reentry and residues of

ethion, parathion, and carbophenothion (Trithion)
on Florida citrus. Proc. Florida State Hort. Soc. 90:
PAPACEK, D., AND D. SMITH. 1992. Integrated pest man-
agement of citrus in Queensland, Australia recent
developments and current status. Proc. Intl. Soc. Ci-
tricult. 3: 973-977.
1975. Investigations of leaf sap feeding by three spe-
cies of phytoseiid mites by labeling with a radioac-
tive phosphoric acid (H232PO4). Ann. Entomol. Soc.
America 68: 871-872.
ROBERTS, P. D., AND L. W. TIMMER. 2000. Greasy Spot in
2000 Florida Citrus Pest Management Guide. Univ.
Florida Coop. Ext. Serv., IFAS. Fact Sheet PP144,
SAS INSTITUTE. 1991. SAS Language and procedures
Usage 2, Version 6, First edition. SAS Institute.
Cary, NC.
SIMANTON, W. A., AND K. TRAMMEL. 1966. Recom-
mended specifications for citrus spray oils in Florida.
Proc. Florida State Hort. Soc. 79: 26-30.
Dislodgable residue of Supracide on citrus leaves.
Agric. Food Chem. 27: 589-592.
TIMMER, L. W. 2000a. 2000 Florida Citrus Pest Manage-
ment Guide: Melanose. Univ. Florida Coop. Ext.
Serv., IFAS. Fact Sheet PP145.
TIMMER, L. W. 2000b. 2000 Florida Citrus Pest Manage-
ment Guide: Citrus Scab. Univ. Florida Coop. Ext.
Serv., IFAS. Fact Sheet PP146.
TIMMER, L. W. 2000c. 2000 Florida Citrus Pest Manage-
ment Guide: Alternaria Brown Spot. Univ. Florida
Coop. Ext. Serv., IFAS. Fact Sheet PP147.

Florida Entomologist 84(3)

September 2001


Centro Internacional de Agricultura Tropical (CIAT), Apartado Aereo 6713, Cali, Colombia


The genus Prosapia Fennah andP. simulans (Walker) are reported for the first time in South
America, based on recent field collections in Colombia and museum specimens from Venezu-
ela. Prosapia simulans was found on Axonopus micay Garcia-Barr., Brachiaria decumbens
Stapf, B. dictyoneura (Fig. & De Not.) Stapf, Cynodon plectostachyus (K. Schum.) Pilger, Hy-
parrhenia rufa (Nees) Stapf and Saccharum officinarum L. (Poaceae). Persistent field popu-
lations were detected from 1060-1621 m elevation, principally associated with B. decumbens,
reaching economic levels in one of the observed sites. On two occasions P. simulans was
found on sugar cane. Evidence suggests that this Central American sugar cane and forage
grass pest is a well-established new arrival, thereby representing a new threat to pasture
production and potential threat to cane production in Colombia's Cauca Valley. The distribu-
tion, bionomics, and pest status of P. simulans are summarized, and its mode of introduction
and potential pest status are discussed.

Key Words: Brachiaria decumbens, Colombia, forage pest, new detection, Prosapia simu-
lans, Saccharum officinarum, spittlebug

El g6nero Prosapia Fennah y P. simulans (Walker) son reportados por primera vez en Sura-
merica, basado en recolecciones recientes del campo en Colombia y especimenes de museo de
Venezuela. Prosapia simulans fue encontrado enAxonopus micay Garcia-Barr., Brachiaria
decumbens Stapf, B. dictyoneura (Fig. & De Not.) Stapf, Cynodon plectostachyus (K. Schum.)
Pilger, Hyparrhenia rufa (Nees) Stapf y Saccharum officinarum L. (Poaceae). Poblaciones
persistentes en campo fueron detectadas desde 1060-1621 msnm, principalmente asociada
con B. decumbens, alcanzando niveles econ6micos en uno de los sitios observados. En dos
ocasiones P. simulans fue encontrado sobre cana de azucar. La evidencia sugiere que esta
plaga centroamericana de cana de azucar y granineas forrajeras es una nueva llegada bien
establecida, representando asf una nueva amenaza para la producci6n de pastos y una ame-
naza potential para la producci6n de cana en el Valle del Cauca en Colombia. Se resume la
distribuci6n, las bion6micas, y el estado de plaga de P. simulans, y se discute su modo de in-
troducci6n y estado de plaga potential.

Grassland spittlebugs (Homoptera: Cercopidae)
are native xylem-feeding insects that are broadly
distributed and damaging pests of graminoid crops
in the Neotropics. Major hosts include sugar cane
(Saccharum officinarum L.) and forage grasses,
particularly the widely sown and highly suscepti-
ble Brachiaria decumbens Stapf. This diverse
group of spittlebugs, which includes dozens of spe-
cies from at least 11 genera, poses a major limita-
tion to productivity and persistence of the most
extensive agricultural activity in the Neotropics,
pastures for the production of forage, milk and beef.
The most widely distributed species in this
pest complex is Prosapia simulans (Walker), oc-
curring in the lowland tropics from Mexico to
Panama (Hamilton 1977). To our knowledge, this
species and the genus Prosapia Fennah have
never been reported in South America. Herein we
report the first field detection of P. simulans in
Colombia, and an additional record from museum
specimens collected in Venezuela. Quantitative

measures of field abundance were performed to
make a preliminary assessment of population
density and persistence. We summarize the liter-
ature on its geographic distribution, bionomics
and pest status; provide diagnostic characters to
distinguish it from other grassland spittlebugs in
northern South America; and discuss its possible
mode of introduction and pest status potential.


We discovered populations of P. simulans on
six farms in the Cauca Valley of Colombia in
1999-2000. All specimens were identified using
characteristics of the male genitalia and com-
pared with type specimens at the Natural History
Museum (BMNH), London.
The first report was obtained 2-VI-1999 when
a single female was captured in the course of
weekly surveys to document population fluctua-
tions of the common local spittlebug species Zulia

Peck et al.: Prosapia simulans detected in South America

carbonaria (Lallemand) in Brachiaria dictyo-
neura (Fig. & De Not.) Stapf (population 1, Haci-
enda Las Palmas, 3.050N, 76.498W) (Table 1,
Fig. 1). Additional surveys in surrounding pas-
tures and sugar cane fields, and weekly surveys
in the same site over the following year, did not
recover more individuals.
On 2-VII-1999 a large population was discov-
ered approximately 94 km northeast in a pasture
of B. decumbens at 1155 m elev. (population 2,
Finca El Mirador del Paraiso, 3.650N, 76.240W)
(Table 1, Fig. 1). On that visit P. simulans was the
only spittlebug species present, with abundant
adults and nymphs (unmeasured). A follow-up
visit to the same pasture on 4-IV-2000 verified a
continued presence at greater abundance than
the first visit, measured at 46.8 nymphs/m2 (n =
10 0.25m2 quadrats) and 190 adults/50 sweeps (n
= 4 series of 50 sweeps). In the second and third
visits, Z. carbonaria and Zulia pubescens (F.)
were also detected. Although economic thresholds
based on quantitative yield loss data have never
been established for grassland spittlebugs, these
levels are considered highly damaging in forage
grasses of Mexico. Padilla and Esquiliano (1966)
designate >30 nymphs/m2 and >25 adults/50
sweeps as "severe" while Velasco & Sifuentes
(1970) designate >46 nymphs/m2 and >150 adults/
50 sweeps as "high" infestations.
Subsequent to the detection of this large popu-
lation, four additional populations were detected
between 1100 and 1621 m elevation, ranging from
the valley floor to just over the top of the western
cordillera of the Andes (Table 1, Fig. 1). These
populations were persistent because the insect
was detected over a few to several months.
In population 3 (Finca Canada, 3.8490N,
76.466oW), P. simulans were detected along with
Z. carbonaria and Z. pubescens in five upland pas-
tures of B. decumbens and Cynodon plectostachyus
(K. Schum.) Pilger. In the second visit, various P.
simulans adults were found feeding on sugar cane
(S. officinarum) but nymphs were absent. In pop-
ulations 4 (Finca La Maria, 3.9400N, 76.438oW)
and 5 (Finca La Albania, 3.9500N, 76.453oW) the
insect was detected along with Z. carbonaria and
Z. pubescens in three upland pastures ofB. decum-
bens. Additional hosts in population 5 included
Axonopus micay Garcia-Barr., C. plectostachyus
and Hyparrhenia rufa (Nees) Stapf. In population
6, P. simulans were detected along with Z. carbon-
aria and Z. pubescens in four lowland pastures of
B. decumbens (Hacienda Piedechinche, 3.3730N,
76.142oW). On a separate occasion, a single
nymph was observed in a spittle mass at the base
of the leaf whorl on sugar cane; at the time onlyP.
simulans adults were found on nearby weeds and
grass (L.A. Lastra, CENICANA, pers. comm.).
Voucher specimens of P. simulans from all six
populations were deposited in the Cornell Univer-
sity Insect Collection under Lot #1227.


A series of P. simulans was identified in the in-
sect collection of the Centro Internacional de Ag-
ricultural Tropical (CIAT), Cali, Colombia (2
specimens) and BMNH (7 specimens). This mate-
rial was all collected on 30-V-1980 by Gerardo
Perez Nieto from Venezuela, Bolivar State, La
Vergarena, in pasture, calculated to be near
6.783N, 63.559W (Fig. 1). No other South Amer-
ican specimens were found in the collections at
BMNH, Cornell University, the Universidad Na-
cional in Palmira or the Universidad del Valle in
Cali (Dept. Valle del Cauca).


The genus Prosapia comprises 14 species and
ranges from Ontario, Canada to Panama (Hamil-
ton 1977). Two new species from Costa Rica are
being described (V. Thompson, Roosevelt Univer-
sity, pers. comm.).Prosapia simulans is widespread
in the lowland tropics (Fig. 1). It is reported in
nine Mexican states (Chiapas, Guerrero, Nuevo
Le6n, Oaxaca, Quer6taro, San Luis Potosi,
Tabasco, Tamaulipas, Vera Cruz), within 70 miles
of the United States border (Fennah 1953; Clark
et al. 1976; Agostini et al. 1981), and throughout
Central America (Belize, Costa Rica, Guatemala,
Honduras, Nicaragua, Panama) (Metcalf 1961;
D.P., personal observation of specimens at
BMNH). Although Guagliumi (1955) lists P sim-
ulans in Colombia, the report is unsubstantiated
and is probably in error.
Prosapia simulans attacks many of the major
forage grass species in this geographic range, in-
cluding Cenchrus ciliaris L. (Enkerlin & Morales
1979), Cynodon nlemfuensis Vanderyst (Peck
1999), Digitaria decumbens Stent (Oomen 1975),
Paspalum notatum Flugg6 (V. Thompson, pers.
comm.), S. officinarum (Box 1953, Oomen 1975),
and Zea mays L. (Ballou 1936). Four additional
host species in greenhouse studies are Avena sa-
tiva L., Pennisetum glaucum (L.) R. Br., Setaria
italica (L.) P. Beauv. and Sorghum sp. (Enkerlin
& Schwartz 1979). The only non-graminoid host
(adults only) reported is the tree Ilex haberi
(Aquifoliaceae) (Peck 1998a).
Enkerlin & Morales (1979) also add the follow-
ing hosts (citing Flores et al. 1965 and Velasco
1968) but do not specify if they are particular toP.
simulans or the sympatricAeneolamia albofasci-
ata (Lallemand): Chloris gayana Kunth, Cynodon
plectostachyus (K. Schum.) Pilger, Cynodon dacty-
lon (L.) Pers., Echinochloa polystachya (Kunth)
Hitchc., H. rufa, Panicum purpurascens Raddi,
Panicum maximum Jacquin, and Pennisetum
purpureum Schumach.
The current pest status ofP. simulans in sugar
cane of Central America is poorly documented.
Although reported as an injurious cane pest in


First detection and subsequent visits
New Location Elev.
population (vereda, municipality, department) (m) Date Estimate of population size

1 Santander de Quilichao, Santander de Quilichao, Cauca 1060 2-VI-1999 1 female
2 Santa Helena (a), El Cerrito, Valle del Cauca 1155 2-VII-1999 Unmeasured
4-IV-2000 46.8 nymphs/m2, 190 adults/50 sweeps
5-V-2000 112 adults/50 sweeps
3 Cordobitas, Yotoco, Valle del Cauca 1535 1-II-2000 Unmeasured
6-IV-2000 2.3 adults/50 sweeps
23-V-2000 Unmeasured
4 Diamante la Gaviota, Calima el Darien, Valle del Cauca 1575 12-VI-2000 9.7 adults/50 sweeps
27-VI-2000 13.8 adults/50 sweeps
5 La Primavera, Calima el Darien, Valle del Cauca 1621 12-VI-2000 4.0 adults/50 sweeps
27-VI-2000 6.5 adults/50 sweeps
6 Santa Helena (b), El Cerrito, Valle del Cauca 1100 6-VII-2000 23 adults/50 sweeps
7-X-2000 26.8 adults/50 sweeps
26-X-2000 Unmeasured

Peck et al.: Prosapia simulans detected in South America

Fig. 1. Known geographic distribution of Prosapia simulans in Mexico (by state) and Central America (by coun-
try) with new locations (by number) in South America where it was collected in Colombia and Venezuela: 1)
Santander de Quilichao 2) El Cerrito (a), 3) Yotoco, 4) Calima el Darien (Diamante la Gaviota), 5) Calima el Darien
(La Primavera), 6) El Cerrito (b), and 7) La Vergarena.

Honduras and Nicaragua (J. Gaviria, consultant,
pers. comm.), its pest status is probably inferior to
the sympatric Aeneolamia postica (Walker).
Information on the specific biology and behavior
of P. simulans is limited because most studies were
conducted on a mixed species complex rather than
P. simulans in particular. Oomen (1975) reported
two generations of P. simulans annually in D. de-
cumbens pastures near the Gulf Coast of Mexico,
with preoviposition plus the egg incubation stage
requiring 33.5 d, nymphal stage 25.5 d, and total
generation time 58 d. Velasco & Sifuentes (1970)
reported a preoviposition period of 4 d, egg incuba-
tion of 18.7 d, nymphal stage of 22-48 d and total
generation time of 58.4 d. Like all other species
studied in pastures, nymphs and adults ofP. simu-
lans occur only during the wet season, while dor-
mant eggs survive the dry season and hatch under
wet conditions. The genus Prosapia lays eggs in
the soil but like other spittlebug species a propor-
tion are attached to the plant stem and litter (Pass
& Reed 1965, Peck 1998b). Current studies at
CIAT show that P. simulans in Colombia primarily
oviposits on the surface of the plant stem in prefer-
ence to the soil or plant litter (CIAT 2000).
In all of the above studies,P. simulans was de-
scribed as sharing pastures with another spittle-
bug species, such as A. albofasciata in northern

Mexico, or A. postica in Honduras. There are no
known published reports of it achieving outbreak
status individually. The pest status of grassland
cercopids in general has increased in forages of
South America over the last decade. Based on re-
ports received by CIAT (D.P., personal observa-
tion), increasingly affected areas include
northern Argentina, Ecuador, and the Andean
hillsides, Caribbean coast, and Amazonian forest
margins of Colombia.


Prosapia simulans can be separated from the
other 13 described species of Prosapia by the fol-
lowing male genitalic and color pattern characters
(Hamilton 1977): two short blunt teeth at tip of
aedeagal shaft, preapical gonopore, subgenital
plates appressed on inner margins, crown of head
lighter than anterior margin of pronotum, one
light-colored transverse band across pronotum
and two across tegmen, head and tegminal bases
red to brown, and mesopleura blotched with black.
Male P. simulans can be distinguished from
the other 17 species associated with wild and cul-
tivated graminoids in Colombia and Ecuador
(Peck 2001) by dorsal color pattern: dark brown to
black with one transverse band across the center

Florida Entomologist 84(3)

of the pronotum and two bands (complete to inter-
rupted) across the tegmen (Fig. 2). These trans-
verse bands are greatly reduced to absent in
females. As the only known member of the genus
in South America, P. simulans is also distin-
guished by the genus definition of Fennah (1949,
1953) and supporting male genitalia characters
discussed by Hamilton (1977).
In Mexico and Central America, there is signif-
icant variation in the color and form of the trans-
verse bands of P. simulans, ranging from yellow to
pink/red to orange, broad to narrow, and distinct
to completely obscured, particularly in females
(Clark et al. 1976, D.P., personal observation).
There is additional variation among Central
American populations in color of subgenital
plates and patches on ventral edge of abdominal
tergites (same color as venter to black). The Co-
lombian populations displayed a particular sub-
set of this color and pattern variation. Of 140
males examined from across the six populations,
all had narrow pale yellow tegminal bands with
some reduction of the posterior band. The color of
the venter was predominantly pink/red (64%), but
some individuals were yellowish brown (16%) or
intermediate (20%). With very few exceptions,
background tegmen color was brown (versus
black), subgenital plates were black, and black
patches on the abdominal tergites were pro-
nounced. Of 43 females examined, all had both
tegminal bands and pronotal band greatly re-
duced to barely evident on a black background.
Female venters were black with red (58%) to yel-
lowish brown (30%) to intermediate (12%) mark-
ings. Background tegmen color was usually
brown (74%) but sometimes black (26%).
For males (population 1, n = 40), mean (mm)
+SE (range) head capsule width, forewing length
and body length (including wings) was 2.04
0.009 (1.93-2.14), 6.84 + 0.045 (5.93-7.43) and
8.52 0.049 (7.36-9.29), respectively. For females
(population 1, n = 40) these measures were 2.31 +
0.009 (2.21-2.43), 6.80 + 0.035 (6.36-7.21) and
8.71 0.052 (7.29-9.29), respectively.


Although the occurrence of P. simulans in Co-
lombia and Venezuela could be attributed to low
endemic populations only recently detected, we
believe this is unlikely because P. simulans is a
conspicuous insect that has warning coloration
and is present in an agricultural activity vital to
the region. Furthermore, the Cauca Valley and
Venezuela have been under relatively high sur-
veillance. The CIAT forages program has been ac-
tive in the Cauca Valley for the last 20 years and
extensive fieldwork on grassland spittlebugs was
conducted in the 1950's across Venezuela (Guag-
liumi 1955, 1956a, b, 1957).

Alternatively, P. simulans may have arrived
through natural or human-mediated dispersal.
Natural dispersal is a possibility for the Cauca
Valley populations. Given the detection of one
high elevation population on the Pacific side of
the western cordillera of the Andes, P. simulans
could have spread from Panama through the
Darien Gap, the Choc6 region of Colombia, and
down the Pacific coast. Testing this mode of intro-
duction will depend on collections from those key
intermediate regions.
Human mediated dispersal is a more likely
possibility for the Venezuelan population given
how remote the Venezuelan site is to Panama and
the absence of P. simulans in northern Colombia.
Spittlebug monitoring programs conducted by the
Corporaci6n Colombiana de Investigaci6n Agro-
pecuaria (CORPOICA), the Universidad del Sucre
and CIAT over the last four years on the Caribbean
coast have not detected this species. The prefer-
ence of P. simulans for oviposition sites on the
plant stem versus soil and litter increase the like-
lihood of entry with infested vegetative material.
Broader surveys should be carried out to iden-
tify the extent of P. simulans populations and
monitor their spread in pastures and cane planta-
tions of the Cauca Valley. At the spittlebug densi-
ties detected in one location of this study, milk
and beef cattle production will be negatively af-
fected and the persistence of improved B. decum-
bens pastures will be compromised. Spread or
introduction of this species to lowland regions of
Colombia such as the Caribbean coast or the ex-
tensive eastern Llanos would have severe eco-
nomic implications.
Up to now, P. simulans has not been reported
on Colombian sugar cane beyond the two observa-
tions noted above. Nevertheless, because the evi-
dence suggests that P. simulans is a recent
arrival, cane producers should consider this spe-
cies a potential threat. With the notable exception
of the Cauca Valley of Colombia, essentially all
cane-producing regions of Central and South
America have experienced major spittlebug pest
problems (Fewkes 1969). For instance, the spittle-
bug Aeneolamia varia saccharina (F.) devastated
the cane industry in Trinidad and other Carib-
bean regions at the turn of the century (Williams
1921). Brazilian cane fields have a long history
with a diverse complex of other species (Guag-
liumi 1972) and in the last decade spittlebug pest
status has increased in cane plantations of Ecua-
dor and Central America.
The menace may be heightened now as man-
agement shifts from preharvest burning to green
production by the year 2005 in the Cauca Valley
(Cenicana 1998). This change in cultural practice
is known to influence the status of insect pests,
such as the emergence ofMahanarva fimbriolata
(Stal) in cane fields in Sao Paolo State, Brazil
(P. Botelho, CCA/UFSCar, pers. comm.).

September 2001

Peck et al.: Prosapia simulans detected in South America


Fig. 2. Prosapia simulans showing most common color patterns (dorsal and ventral) of males (left) and females
from newly detected populations in Cauca Valley, Colombia. Bar = 2 mm.

Florida Entomologist 84(3)

Finally, these observations highlight the need
for care in transfer of vegetative and soil materials
associated with cercopid host plants. There is
some other anecdotal evidence for regional intro-
ductions of grassland spittlebugs such as Z. car-
bonaria from the Cauca Valley into the Colombian
Amazon, and an isolated report of the Central Bra-
zil species Notozulia entreriana (Berg) in the Co-
lombian Llanos (Peck 2001). One well-documented
case is Lepyronia coleoptrata (L.) (Homoptera:
Aphrophoridae), a Palearctic spittlebug with im-
migrant status in the United States (Hoebeke &
Hamilton 1983). With the increasing movement of
vegetative material throughout the Caribbean Ba-
sin and northward insect range expansion due to
warming trends, sugar cane and forage grass pro-
duction in the southern United States, like the
Cauca Valley, would be threatened by the arrival of
new spittlebug pests. The southeast United States
already suffers from the native Prosapia bicincta
(Say), a damaging pest of forage grass, turf grass
and ornamentals (Fagan & Kuitert 1969; Braman
& pendley 1993; Braman & Ruter 1997).


We thank V. Thompson (Roosevelt University) for
alerting us to specimens of P. simulans from Venezuela
in the CIAT collection, Gilberto C6rdoba (CIAT) for find-
ing the first large population of P. simulans in Colombia,
and Mick Webb for access to the collection at the Natural
History Museum, London. E. R. Hoebeke (Cornell Uni-
versity), V. Thompson and an anonymous reviewer pro-
vided valuable comments for improving the manuscript.


1981. Rendimiento y calidad de dos hibridos de zacate
buffel (Cenchrus ciliaris L.) danados por diferentes
poblaciones del complejo mosca pintaAeneolamia al-
hofasciata (Lallemand) y Prosapia simulans
(Walker), Apodaca, N.L., 1980. Agronomia 200: 42-47.
BALLOU, C. H. 1936. Insectos observados durante el ano
1934. Cent. Nac. de Agr. Bol. 20: 1-60.
Box, H. E. 1953. The history and changing status of
some insect pests of sugar cane. Trans. IX Internat.
Congress Entomol. 2: 254-259.
BRAMAN, K. S., AND A. F. PENDLEY. 1993. Relative and
seasonal abundance of beneficial arthropods in cen-
tipedegrass as influenced by management practices.
Hortic. Entomol. 86: 495-504.
BRAMAN, K. S., AND J. M. RUTER. 1997. Preference of
twolined spittlebug for Ilex species hybrids and cul-
tivars. J. Environ. Hortic. 15: 211-214.
CENICANA. 1998. Informe Annual 1998, Cenicana, Cali,
CIAT. 2000. Annual Report 2000, Project IP-5, Tropical
Grasses and Legumes: Optimizing Genetic Diversity
for Multipurpose Use. Centro Internacional de Agri-
cultura Tropical.
CLEAVE. 1976. Taxonomy and biology of spittlebugs
of the generaAeneolamia Fennah and Prosapia Fen-

nah (Cercopidae) in northeastern Mexico. Folia En-
tomol6gica Mexicana 34: 13-24.
ENKERLIN, D., AND J. A. MORALES. 1979. The grass spit-
tlebug complex Aeneolamia albofasciata and Prosa-
pia simulans in northeastern Mexico and its possible
control by resistant buffelgrass hybrids, pp. 470-494.
In M. K. Harris [ed.] Biology and Breeding for Resis-
tance to Arthropods and Pathogens in Agricultural
Plants: Proceedings of a Short Course Entitled "In-
ternational Short Course in Host Plant Resistance".
Texas A&M University, College Station.
ENKERLIN, D., AND A. J. SCHWARTZ. 1979. Estudio de
gramineas como posibles hospederas de la mosca pinta
Prosapia simulans Walker, bajo condiciones de inver-
nadero. Division de Ciencias Agropecuarias y Mariti-
mas, Instituto Tecnologico de Monterrey 16: 89-90.
FAGAN, B. E., AND L. C. KUITERT. 1969. Biology of the
two-lined spittlebug, Prosapia bicincta, on Florida
pastures (Homoptera: Cercopidae). Florida Entomol.
52: 199-206.
FENNAH, R. G. 1949. Autecological notes on three spe-
cies of Aeneolamia (Homoptera: Cercopidae). Ann.
Mag. Nat. Hist., Series 12 2(21): 703-726
FENNAH, R. G. 1953. Revisionary notes on neotropical
monecphorene Cercopoidea (Homoptera) Ann. Mag.
Nat. Hist., Series 12 6: 337-360.
FEWKES, D. W. 1969. The biology of sugar cane froghop-
pers, pp. 283-307. In J. R. Williams, J. R. Metcalfe, R.
W. Mungomery and R. Mathes [eds.], Pests of Sugar
Cane. Elsevier, Amsterdam.
FLORES, C., S. RAMIREZ, AND C. CORTES. 1965. El sal-
ivazo de la cana de azucar. Inst. Mej. Prod. Azucar.
Bol. Divulgaci6n 5: 14-18.
GUAGLIUMI, P. 1955. Contribuciones al studio de la
candelilla de las gramineas en Venezuela. II. Los cer-
copidos causantes de la candelilla. A) Aeneolamia
(= Tomaspis) varia (F.) y sus subespecies. Agron.
Trop. 5: 135-194.
GUAGLIUMI, P. 1956a. Contribuciones al studio de la
candelilla de las gramineas en Venezuela. II. Los cer-
copidos causantes de la candelilla. B) Aeneolamia
flavilatera (Urich) y sus subespecies. Agron. Trop. 6:
GUAGLIUMI, P. 1956b. Contribuciones al studio de la
candelilla de las gramineas en Venezuela. II. Los cer-
copidos causantes de la candelilla. C) Ae. reducta
montana Fennah D)Ae. lepidior (Fowl.). Agron. Trop.
6: 123-133.
GUAGLIUMI, P. 1957. Contribuciones al studio de la
candelilla de las gramineas en Venezuela. III.
Cuadro de distribuci6n geografica de las species de
Aeneolamia Fennah y de sus plants hospederas
senaladas en Venezuela. Agron. Trop. 6: 165-194.
GUAGLIUMI, P. 1972. Pragas da Cana-de-Auicar, Nor-
deste do Brasil. Divulgagao do M.I.C., Instituto do
Auicar e do Alcool, Divisao Administrativa, Servico
de Documentacao, Rio de Janeiro.
HAMILTON, K. G.A. 1977. Review of the world species of
Prosapia Fennah (Rhynchota: Homoptera: Cercopi-
dae). Canadian Entomol. 109: 621-630.
HOEBEKE, E. R., AND K. G. A. HAMILTON. 1983. Lepyro-
nia coleoptrata (L.), a European spittlebug in east-
ern North America: new locality records and new key
to the North American species of Lepyronia Amyot
and Serville (Homoptera: Cercopidae). Proc. Ento-
mol. Soc. Washington 85: 263-271.
SOTELO, AND F. CORDOBA. 1992. Antibiosis to spittle-

September 2001

Peck et al.: Prosapia simulans detected in South America

bugs (Homoptera: Cercopidae) in accessions of Bra-
chiaria spp. J. Econ. Entomol. 85: 1485-1490.
METCALF, Z. P. 1961. General Catalogue of the Ho-
moptera. Fascicle VII Cercopoidea. Part 2 Cercopi-
dae. North Carolina State College, Raleigh.
OOMEN, P. A. 1975. A population study of the spittle bugs
Aeneolamia occidentalis (Walk.) and Prosapia simu-
lans (Walk.) (Homoptera: Cercopidae) in Mexican
pangola pastures. Z. Fur Angew. Entomol. 79: 225-238.
PADILLA, C. C., AND E. D. ESQUILIANO. 1966. Campaina
contra la mosca pinta y la escama algodonosa de los
pastos. Fitrofilo 50: 5-52.
PASS, B. C., AND J. K. REED. 1965. Biology and control of
the spittlebug Prosapia bicincta in coastal Bermuda
grass. J. Econ. Entomol. 58: 275-278.
PECK, D. C. 1998a. Use of alternative host plants exclu-
sively by adult male froghoppers (Homoptera: Cer-
copidae). Biotropica 30: 639-644.
PECK, D. C. 1998b. Natural history of the spittlebug Pro-
sapia nr. bicincta (Homoptera: Cercopidae in associ-
ation with dairy pastures of Costa Rica. Ann.
Entomol. Soc. America 91: 435-444.

PECK, D. C. 1999. Seasonal fluctuations and phenology
of Prosapia spittlebugs (Homoptera: Cercopidae) in
upland dairy pastures of Costa Rica. Environ. Ento-
mol. 28: 372-386.
PECK, D. C. 2001. Diversidad y distribuci6n geografica
del salivaxo (Homoptera: Cercopidae) asociado con
gramineas en Colombia y Ecuador. Rev. Colombiana
Entomol (in press).
VALERIO, J. R., AND 0. NAKANO. 1988. Danos causados
pelo adulto da cigarrinha Zulia entreriana na
produgao e qualidade de Brachiaria decumbens.
Pesq. Agropec. Brasileira 23: 447-453.
VELASCO, H. 1968. Resultados de 5 ciclos de investi-
gaci6n en la mosca pinta de los pastos en el sureste
de M6xico reporte sin publicar). INIA-SAG.
VELASCO, H., AND J. A. SIFUENTES. 1970. Investiga-
ciones sobre la mosca pinta de los pastos en el sur-
este de M6xico. VI Informe C.F.A.S.E. Inst. Nac. Agri.
S.A.G. Mex.
WILLIAMS, C. B. 1921. Report on the froghopper-blight of
sugarcane in Trinidad. Memoirs of the Department
of Agriculture, Trinidad and Tobago 1-170.

Florida Entomologist 84(3)

September 2001


Department of Zoology, Southern Illinois University, Carbondale, Illinois 62901


The life history ofEmesaya b. brevipennis (Say) was studied in southern Illinois from April
to December 1998. The bug also was reared in the laboratory on Drosophila sp. at 26 3.0C
under a 16:8 (L:D) photoperiod. This bivoltine species apparently overwintered as eggs. First
instars were found primarily from early to late May and mid-July to mid-August, second in-
stars from late May to early June and from late July to mid-August, third instars from late
May to mid-June and during August, fourth instars primarily from early June to early July
and early to late August, fifth instars primarily from mid-June to mid-September, and adults
from late June to early December. In the laboratory, the incubation period averaged 33.91 d.
The five nymphal stadia averaged 11.27, 7.84, 8.85, 11.14, and 16.75 d, respectively. Total de-
velopmental time averaged 89.76 d.

Key Words: Bivoltine, copulation, spider webs

El historical de vida de Emesaya b. brevipennis fue estudiado en el sur de Illinois desde Abril
a Diciembre del 1998. El insecto tambi6n fue criado en el laboratorio en Drosophila esp. a 26
3.0 bajo un fotoperiodo de 16:8 (L:O). Esta especie bivoltina aparentemente sobrevivi6 el
invierno como huevos. Los primeros instares fueron encontrados principalmente del co-
mienzo al final de mayo y mitad dejulio a mitad de agosto, los segundos instares de tarde en
mayo a temprano en junio y de finales de julio a medio agosto, terceros instares de finales de
mayo a mitad de junior y durante agosto, cuarto instares principalmente desde el comienzo
de junio al comienzo de julio y todo el mes de agosto, quinto instares principalmente de me-
diados de junio a mediados de septiembre, y adults desde el fin de junio al comienzo de di-
ciembre. En el laboratorio, el period de incubaci6n fue un promedio de 33.91 d. El tiempo
de desarrollo total fue 89.76 d.

Emesaya brevipennis (Say) is one of the
emesine reduviids, a cosmopolitan group of bugs
characterized by markedly slender bodies and ap-
pendages (Wygodzinsky 1966). This New World
species is divided into three subspecies, E. b. aus-
tralis McAtee & Malloch, E. b. occidentalis McA-
tee & Malloch, and E. b. brevipennis (Say)
(Froeschner 1988), all of which occur in America
north of Mexico. E. b. brevipennis, which is the
most widely distributed of the three subspecies,
occurs from New York and Massachusetts south
to Florida and west to Iowa, Kansas, Texas, and
California (Froeschner 1988). It occurs through-
out Illinois (JEM, unpublished data).
Emesaya b. brevipennis has received much at-
tention over the years, probably due, in part, to its
large size (33.0-37.0 mm [Blatchley 1926]) and
wide distribution. Published information on its
biology has consisted mainly of scattered notes. It
has been collected under bridges (e.g., Gates &
Peters 1962); in sheds, barns, and outbuildings
(e.g., Banks 1909; Blatchley 1926; Froeschner
1944; Gates & Peters 1962; Howes 1919; Readio
1926, 1927; Torre-Bueno 1923, 1925; Uhler 1884;
Wickham 1910); from vegetation (e.g., Banks

1909; Blatchley 1926; Froeschner 1944; Gates &
Peters 1962; Torre-Bueno 1923, 1925; Uhler
1884), flood debris, Spanish moss (Elkins 1951),
screens (Brown & Lollis 1963); and in association
with spider webs (Banks 1909; Brown & Lollis
1963; Howes 1919; Readio 1926, 1927; Usinger
1941; Wickham 1910).
Based on the literature, this subspecies appar-
ently is univoltine (Banks 1909; Readio 1927) and
overwinters as eggs (Howes 1919; Readio 1927).
Nymphs occur in the spring and much of the sum-
mer (Brown & Lollis 1963; Readio 1927; Uhler
1884; Wickham 1910), and adults can be found
during the summer and early fall (Brown & Lollis
1963; McAtee & Malloch 1925; McPherson 1992;
Readio 1927; Uhler 1884; Wickham 1910). The
eggs are oviposited in the summer and early fall
(Brown & Lollis 1963; Howes 1919; Readio 1927)
and are attached to spider webs (Brown & Lollis
1963; Readio 1926, 1927), rafters of wooden struc-
tures (Howes 1919; Readio 1926, 1927), and twigs
of bushes and trees (Uhler 1884).
This paper presents information on the field
life history and biology of E. b. brevipennis in
southern Illinois.

Hagerty et al.:Life History ofEmesaya b. brevipennis


Field Life History

During summer 1997, a population of E. b.
brevipennis was discovered near Bluff Lake,
Union Co., IL. The numbers observed and accessi-
bility of the site suggested a life history study was
possible. Therefore, a study was conducted from
April to December 1998, before and after the ac-
tive season, respectively.
The study site is located in the Jonesboro
quadrangle 7.5' topographic (T13S, R2W, NW1/4,
NE1/4, Sec. 20), 4 miles east of state highway 3. It
consists of a Bailey limestone rock face (Nelson &
Devera 1995) covered by vines of Campsis radi-
cans (L.) and Rhus radicans L. The rock face is
approximately 18.5 m high and parallels the east
side of township road 235N for 160.9 m. The site
is surrounded by a forested area containingAcer
barbatum Michaux, Carpinus caroliniana Walter,
Carya glabra (Miller), Carya ovalis (Wangen-
heim), Celtis occidentalis L., Fagus grandifolia
Ehrhart, Quercus rubra L., and Ulmus rubra Mu-
hlenberg. The bugs were observed on webbing of
the araneid spider Anelosimus studiosus (Hentz),
which enclosed the vines on the rock face.

o- (N=22)

-- (N=38)


0. (N=69)

Samples of up to 20 adults and nymphs, and
notes on the bugs' activities, were taken weekly
from early May to early December. Sampling was
by hand picking and confined to an approximately
11.0 m long and 6.0 m high section of the rock
face. Nymphs large enough to be identified to in-
star and adults were not collected. Younger
nymphs were preserved in 70% ethanol and taken
to the laboratory for closer examination. Plant
material, webbing, and the rock face were exam-
ined in the field for eggs.

Laboratory Rearing

Eggs were collected at Bluff Lake on 13 Febru-
ary (n = 163) and 6 March (n = 75) 1999, brought
to the laboratory, placed on moistened filter paper
in the bottoms of petri dishes (approximately 9 cm
diam., 2.0 cm deep) and covered with the lids. Ap-
proximately 4-6 drops of distilled water were
added every 1-2 d to keep the filter paper moist.
The resulting nymphs were placed in 1-pt (ap-
proximately 0. 47 liter) Mason jars with a disc of
moistened filter paper on the bottom. Each jar
was closed with a disc of paper toweling and wire
screening and secured with the band of the 2-
piece Mason jar lid. A strip of paper toweling,

Figure 1. Percent of individuals in each stage per sample of Emesaya b. brevipennis collected at Bluff Lake,
Union Co., IL, during 1998.

Florida Entomologist 84(3)

September 2001

U 50-
S- (N38)


Figure 2. Percent in each sample of total individuals of same stage of Emesaya b. b reipennis collected at Bluff
Lake, Union Co., IL, during 1998.

folded longitudinally (to prevent the strip from
rolling up when damp), was suspended inside the
jar from the lid and secured at the upper end be-
tween the discs of wire screen and paper toweling.
This strip increased surface area for walking and
absorption of excrement. The filter paper was
moistened with 8-10 drops of distilled water per
day. The bugs were fed five Drosophila sp. adults
per day for this and subsequent instars.
To determine if eggs could develop without an
intervening cold period (i.e., without passing
through cold winter temperatures), three adult
females were collected at Bluff Lake on 22 Octo-
ber 1998, brought to the laboratory, and placed in
a 1-qt (approximately 0.95 liter) Mason jar pre-
pared similarly to the 1-pt jars for nymphs. The
strip served a third function in addition to those
for nymphs, that of an ovipositional site. As with
nymphs, the filter paper was moistened with 8-10
drops of distilled water daily, and the bugs were
fed five Drosophila sp. adults per day. The result-
ing eggs also provided incubation data, not possi-
ble with the 1999 field-collected eggs, which were
laid at unknown times.
Eggs were removed, placed on moistened filter
paper in the bottoms of petri dishes, and treated

similarly to the field-collected eggs discussed
The bugs were kept in incubators maintained
at 26 + 3.0C and a photoperiod of 16:8 (L:D) (ap-
proximately 2,800 lux).


Field Life History

Emesaya b. brevipennis is bivoltine in south-
ern Illinois (Figs. 1 and 2) based on peaks in abun-
dance of adults and nymphs.
This species apparently overwintered as eggs,
which were glued lengthwise to the vines, webs,
and rock surface. We could not distinguish fertile
eggs in the field because the eggs are dark and du-
rable, even if not fertile. Of the 238 eggs collected
in February and March for the laboratory-rearing
study, 121 (50.8%) hatched. We assumed all had
been oviposited the previous fall.
First instars were found primarily from early
to late May and mid-July to mid-August, second
instars from late May to early June and from late
July to mid-August, third instars from late May
to mid-June and during August, fourth instars

Hagerty et al.:Life History ofEmesaya b. brevipennis



1961-1990' 19982
Month Max Min Avg Max Min Avg

January 4.8 -5.2 -0.2 7.1 -1.6 2.8
February 7.7 -3.1 4.5 10.7 0.8 5.8
March 13.9 2.4 8.1 12.8 2.7 7.7
April 20.0 7.9 14.0 19.8 7.2 13.5
May 25.0 12.6 18.8 27.0 14.8 20.9
June 29.7 17.2 23.4 29.7 16.7 23.2
July 31.4 19.4 25.4 30.6 18.5 24.6
August 30.5 18.3 24.4 32.8 18.8 25.8
September 26.8 14.6 20.7 30.9 15.7 23.3
October 21.1 8.1 14.6 23.2 8.7 15.9
November 13.9 3.2 8.5 15.7 3.4 9.5
December 7.0 -2.5 2.2 8.9 -0.7 4.1

'At Anna, Illinois (COOPID. 110187, Midwestern Regional Climate Center, Champaign, IL).
'At Anna, Illinois (COOPID. 110187, National Climatic Data Center, Asheville, NC).

primarily from early June to early July and early
to late August, fifth instars primarily from mid-
June to mid-September, and adults from late
June to early December (Figs. 1 and 2).
A few nymphs were found outside the primary
times of occurrence in the field of their respective
instars (Figs. 1 and 2), including four firsts (3 De-
cember), four fourths (6 September, n = 1; 5-13
November, n = 3), and seven fifths (20 May, n = 1;
11 October, n = 3; 13 November, n = 3). We believe
that this was atypical and the result of unusually
mild temperatures during the spring and fall (Ta-
ble 1; note average temperatures for 1961-1990
and 1998).
Copulation was observed in late September (6
pairs) and early October (8 pairs). Interestingly,
on 5 November 1998, a male was seen in copulo
with an apparently dead female.

Laboratory Rearing

Eggs were glued singly and lengthwise to the pa-
per toweling, screening, filter paper, and sides of the
jar. Each egg was dark brown to black with longitu-
dinal rows of thin, toothlike projections and capped
by a cephalic operculum with a central tubercle, as
described by McAtee & Malloch (1925). The incuba-
tion period averaged 33.91 days (Table 2).
The first instar emerged through the cephalic
end of the egg, pushing aside the operculum. It
was whitish, almost transparent, but became
more visible after feeding.
The first through fifth stadia averaged 11.27,
7.84, 8.85, 11.14, and 16.75 d, respectively
(Table 2). The total developmental period aver-
aged 89.76 d. Most nymphs died during the fifth
stadium, which resulted from incomplete ecdysis.


No. Completing
Stage Stadium Range Mean + SE mean age (d)

Egg' 34 30-38 33.91 + 0.39 33.91
1st instar 52 7-19 11.27 0.30 45.18
2nd instar 49 6-14 7.84 + 0.20 53.02
3rd instar 47 6-16 8.85 + 0.30 61.87
4th instar 43 7-20 11.14 + 0.41 73.01
5th instar 20 14-22 16.75 0.53 89.76

35 eggs, all of which laid in laboratory, used for incubation determination.
Nymphs hatched from field-collected eggs.

Brown & Lollis (1962) suggested that females
have a sixth instar; however, none was found.
In conclusion, E. b. brevipennis is bivoltine, at
least in southern Illinois, and overwinters as
eggs. It will feed on Drosophila adults in captivity.
Interestingly, this apparently is not true of Eme-
saya brevicoxa (Banks). Several specimens of this
reduviid found in cobwebs beneath the eaves of a
cabin were kept alive in a "breeding cage" for 5
months on various species of spiders. Although
supplied miscellaneous insects, they never were
observed to feed on them (Usinger 1941).


We thank the following individuals of Southern Illi-
nois University at Carbondale: J. A. Beatty (Depart-
ment of Zoology) for identification of spiders, Beth
Burke (Department of Zoology) for the laboratory cul-
ture of Drosophila sp., and Mike A. Mibb (Department
of Plant Biology) for identification of plants. We also
thank Tudi P. Smith (USDA Forest Service, Missoula,
MT; formerly Murphysboro, IL) for her assistance in
providing geological information and maps of the study
site. Finally, we are grateful to the USDA Forest Service
for granting permission to collect in the Shawnee Na-
tional Forest and to Ray G. Smith (USDA Forest Ser-
vice, Missoula, MT; formerly Vienna, IL), for his help in
obtaining the collecting permit.


BANKS, N. 1909. Notes on our species of Emesidae.
Psyche 16: 43-48.
BLATCHLEY, W. S. 1926. Heteroptera or true bugs of
eastern North America with especial reference to the
faunas of Indiana and Florida. Nature Pub. Co., In-
dianapolis, IN. 1116 pp.
BROWN, H. P., AND D. W. LOLLIS. 1963. Observations on
the life history and behavior of the thread-legged
bug Emesaya b. brevipennis (Say), (Hemiptera:
Ploiariidae). Proc. Oklahoma Acad. Sci. 43: 88-90.
ELKINS, J. C. 1951. The Reduviidae of Texas. Texas J.
Sci. 3: 407-412.
FROESCHNER, R. C. 1944. Contributions to a synopsis of
the Hemiptera of Missouri, Pt. III. Lygaeidae, Pyr-
rhocoridae, Piesmidae, Tingididae, Enicocephalidae,

September 2001

Phymatidae, Ploiariidae, Reduviidae, Nabidae. Am.
Midland Nat. 31: 638-683.
FROESCHNER, R. C. 1988. Family Reduviidae Latreille,
1807. The assassin bugs, pp. 616-651 in T. J. Henry
and R. C. Froeschner (eds.). Catalog of the Het-
eroptera, or true bugs, of Canada and the continen-
tal United States. E. J. Brill, New York. 958 pp.
GATES, D. E., AND L. L. PETERS. 1962. Insects in Kansas.
Kansas State Univ. Ext. Serv. B-94: 1-307.
HOWES, P. G. 1919. Insect behavior. Richard G. Badger,
Gorham Press, Boston, MA. 176 pp.
McATEE, W. L., AND J. R. MALLOCH. 1925. Revision of
the American bugs of the reduviid subfamily Ploiari-
inae. Proc. U. S. Natl. Mus. 67(1): 1-153 (inc. 9
MCPHERSON, J. E. 1992. The assassin bugs of Michigan
(Heteroptera: Reduviidae). Great Lakes Entomol.
25: 25-31.
NELSON, W. J., AND J. A. DEVERA. 1995. Geologic map of
the Jonesboro and Ware quadrangles. Union County,
Illinois. Illinois State Geological Survey, Map IGQ-
14. 1 p. (oversized).
READIO, P. A. 1926. Studies on the eggs of some Reduvi-
idae (Heteroptera). Univ. Kansas Sci. Bull. 16: 157-
READIO, P. A. 1927. Studies on the biology of the Redu-
viidae of America north of Mexico. Univ. Kansas Sci.
Bull. 17: 5-291.
TORRE-BUENO, J. R. de la. 1923. Family Reduviidae, pp.
677-692 in W. E. Britton (ed.). Guide to the insects of
Connecticut. Part IV. The Hemiptera or sucking in-
sects of Connecticut. Connecticut State Geol. Nat.
Hist. Surv. Bull. 34: 1-807.
TORRE-BUENO, J. R. de la. 1925. Methods of collecting,
mounting and preserving Hemiptera. Canadian En-
tomol. 57: 6-10, 27-32, 53-57.
UHLER, P. R. 1884. Order VI.--Hemiptera, pp. 204-296 in
J. S. Kingsley (ed.). The standard natural history.
Vol. II. Crustacea and insects. S. E. Cassino & Co.,
Boston, MA. 555 pp.
USINGER, R. L. 1941. Rediscovery of Emesaya brevicoxa
and its occurrence in the webs of spiders (Hemi-
ptera, Reduviidae). Bull. Brooklyn Entomol. Soc. 36:
WICKHAM, H. F. 1910. A note on Emesa longipes. Ento-
mol. News 21: 27-30.
WYGODZINSKY, P. W. 1966. A monograph of the Eme-
sinae (Reduviidae, Hemiptera). Bull. Am. Mus. Nat.
Hist. 133: 1-614.

Florida Entomologist 84(3)

Hallman: Irradiation Treatment Against Sweetpotato Weevil


USDA-ARS, Weslaco, TX 78596


An ionizing irradiation quarantine treatment of 165 Gy was approved by the California De-
partment of Agriculture against sweetpotato weevil, Cylas formicarius elegantulus (Sum-
mers), infesting sweetpotatoes from Florida. The first commercial shipment was made in May,
2000. At >400 Gy, 'Picadito' white-fleshed sweetpotatoes sometimes showed noticeable discol-
oration of cooked flesh. Therefore, there is not a large margin between the minimum absorbed
dose required for quarantine security (165 Gy) and the minimum dose which might cause ob-
jectionable loss to commodity quality (about 400 Gy); it can be expected that the absorbed dose
range absorbed by sweetpotatoes irradiated on a full pallet when the minimum target dose is
165 Gy will be 165-500 Gy. To be safe, sweetpotatoes should be irradiated in smaller units than
pallet loads, which could result in higher processing costs compared with irradiation on stan-
dard pallets. This is the first instance of an irradiation quarantine treatment being approved
and used against a non-fruit fly where live adults can be found by inspectors and indicates a
significant advance in the transfer of this promising quarantine treatment technology.
Sweetpotato weevil adults irradiated with a target absorbed dose of 150 Gy (maximum ab-
sorbed dose was 165 Gy) lived for 32 days, while at 32 days unirradiated weevils had suffered
57% mortality.

Key Words: Cylas formicarius elegantulus, boniato, disinfestation, gamma


Un tratamiento cuarentena de irradiaci6n ionizante de 165 Gy fue aprobado por el Departa-
mento de Agricultura de California contra el picudo de batata, Cylas formicarius elegantulus
(Summers), infestando batatas de Florida. El primer envio commercial fue hecho en mayo del
2000. A >400 Gy, batatas con care blanca 'Picadito' a veces demostraron decoloraci6n evi-
dente de care cocida. Por lo tanto, no hay un margen amplio entire la dosis minima absorbida
requerida para seguridad de cuarentena (165 Gy) y la minima dosis que pueda causar per-
dida objecionable a la calidad del product (alrededor de 400 Gy); es de esperarse que la gama
de dosis absorbida por las batatas irradiadas en una paleta cuando la dosis objetivo minima
debe ser 165 Gy sera 165-500 Gy. Para estar seguros, las batatas deberian ser irradiadas en
unidades mas pequenas que en cargas de paleta, lo cual puede resultar en costs de procesa-
miento mayores comparado con irradiaci6n en paletas estandar. Esta es la primera instancia
que un tratamiento cuarentena de irradiaci6n es aprobado y usado contra una mosca no fru-
tal donde adults vivos pueden ser encontrados por inspectors e indica un advance significa-
tivo en la transferencia de esta prometedora tecnologia de tratamiento cuarentena.
Adultos del picudo de batata irradiados con una dosis objetivo de absorber 150 Gy (dosis
maxima absorbida fue 165 Gy) vivieron por 32 dias, mientras que en el dia 32 picudos sin
irradiar sufrieron una mortalidad de 57%.

The sweetpotato weevil, Cylas formicarius ele-
gantulus (Summers), is considered the most seri-
ous pest of both orange-fleshed and white-fleshed
(boniato) sweetpotatoes, Ipomea batatas (L.) Lam.,
in much of the crop's growing range (tropics and
subtropics) including the southeastern United
States, Hawaii, and Puerto Rico. It was first noted
in the United States in Louisiana in 1875. Female
weevils oviposit in sweetpotatoes by chewing a
small cavity in the root or stem, depositing an egg,
and sealing the hole with frass. In the field they
tend to oviposit near the juncture of the stem and
tuber. In storage sweetpotato weevils infest all
over the roots until they are completely destroyed.
The complete life cycle requires about 35 days in

the warm sweetpotato-growing regions of the
world. Larvae usually pupate in the roots, and the
female has about a 7 day preoviposition period.
Sweetpotato-growing areas which do not have the
weevil, such as the southwestern United States
and the Mediterranean region, prohibit the impor-
tation of sweetpotatoes without a treatment that
ensures that all weevil stages present are dead.
Killing the weevils without harming the roots is
difficult (Hallman & Chalot 1993); a feasible treat-
ment has not been available.
Ionizing irradiation has proven to be a viable
quarantine treatment against fruit flies (Diptera:
Tephritidae) because much research with these
insects has been conducted and doses required to

Florida Entomologist 84(3)

control fruit flies are relatively low (Hallman
1999). The only commercial uses of irradiation as
a quarantine treatment have been against fruit
flies. An unfavorable property of irradiation quar-
antine treatments which sets it apart from all
other treatments that have been commercially
implemented is the fact that irradiation does not
provide acute mortality at the doses used on fresh
commodities. Inspectors may find live insects and
be unable to distinguish them from unirradiated
insects. The measure of efficacy of irradiation
quarantine treatments against fruit flies is pre-
vention of adult emergence from irradiated eggs
and larvae. Thus, inspectors will find no adult
fruit flies in imported fruits. Even though live lar-
vae may be found, a treatment which prevents
the presence of adults is easier to accept than one
which allows for the presence of live adults in
properly treated produce. All stages of the sweet-
potato weevil may be found in marketed roots.
The adult is invariably the stage of insects which
requires the highest radiation dose to control
(Hallman 2000). This has been substantiated for
sweetpotato weevil by Dawes et al. (1987), who
observed little reproduction after 30 pairs of 7
day-old adults were irradiated with 100 Gy. At
150 Gy, no reproduction occurred (n = 30).


After obtaining permission from the U.S. Dept.
Agric., Animal and Plant Health Inspection Ser-
vice, Plant Protection and Quarantine and the
Texas Dept. Agric., sweetpotato weevils were col-
lected from a 'boniato' sweetpotato field near
Homestead, Florida in the spring of 1999. They
were shipped to our laboratory and reared on or-
ange-fleshed sweetpotato roots purchased from
the local market and 'boniato' (larger, white-
fleshed) sweetpotato roots shipped from Home-
stead. Rearing conditions were about 25C,
75%RH, and a photoperiod of 16:8 (L:D).
Gamma radiation was applied with a 137Cs self-
contained, dry-storage irradiator (Husman Model
521A, Isomedix, Inc., Whippany, NJ) which was
delivering a centerline absorbed dose rate of
about 40 Gy-min-1 during the time of this research.
Reference standard dosimetry was done in 1997
using the Fricke system. Routine dosimetry dur-
ing our research was done with radiochromic film
(Gafchromic MD-55, ISP Technologies, Inc.,
Wayne, NJ), and absorbance at the 510 nm wave-
length was read with a spectrophotometer (Milton
Roy Spectronic 401, Ivyland, PA) using the Fricke
centerline determination as the standard.
Adult sweetpotato weevils up to 3 weeks old
were placed with pieces of sweetpotato root in
clear plastic cylinders (29 cm x 4 cm diameter) in
the center of perforated stainless steel mesh cyl-
inders (11.4 cm inside diameter, 50 cm long)
which were placed in the irradiator for sufficient

time to achieve the target dose of 125 Gy. Routine
dosimetry readings yielded the absorbed dose
range. Irradiated weevils and unirradiated con-
trols were placed with sweetpotato roots, which
were changed every 3-4 days, until all irradiated
weevils were dead. Data recorded were death of
weevils and the number of new insects found in
sweetpotato roots exposed to both irradiated and
unirradiated weevils. Eight replicates of 200-600
weevils (sex ratio of about 1:1) with a total of
3,250 weevils were irradiated with 125 Gy. All ir-
radiated insects within each replicate were held
together to maximize the probability that fertile
adults would mate. Subsequently 14 replicates
with 1,000-3,950 adult weevils per replicate (total
30,655) were treated with a target dose of 150 Gy
and counts of mortality and reproduction were
made as before.
To be viable a quarantine treatment must not
only ensure near 100% efficacy but it must also
not lessen fruit quality excessively. McGuire &
Sharp (1995) found darkening of cooked sweetpo-
tato tubers at 400 Gy but not at 200 Gy; there
were no other negative consequences (concerning
appearance, rot, shelf life, weight, or organoleptic
qualities) from irradiation up to 1 kGy. Therefore,
sweetpotato quality research was only focused on
color of cooked roots; organoleptic preference
tests were conducted because color can influence
an organoleptic rating. 'Boniato' (cv Picadito) tu-
bers grown in Homestead, Florida and shipped to
Weslaco, Texas were irradiated with 0 (control),
200, 300, 400, and 500 Gy, held at about 24C, and
cooked 4 days after irradiation using the follow-
ing recipe: Roots were washed, peeled, cut into
slices about 2.5 cm thick, and placed in about 2 li-
ters of about 24C water with about 2 ml of lemon
juice for about 10 minutes while the cooking pot
was prepared. In the cooking pot the root chunks
were placed in about 2 liters of 100C water with
about 4 g of salt and kept in the boiling water un-
til they were tender upon which they were re-
moved from the water and kept in a covered dish.
Qualitative observations on color of cooked roots
were made and an informal panel was assembled
from laboratory personnel (7-10 persons) and
asked to rate organoleptic qualities of sweetpo-
tato pieces by marking on a 9 cm-long line with
'extremely dislike' written on the left end and 'ex-
tremely like' written on the right end of the line.
Data were recorded as distance (cm) from the left
end of the line to the mark. There were 4 repli-
cates done on different dates. These data were not
analyzed statistically because they are qualita-
tive and may not be normally distributed, but are
reported as mean and SEM of the 4 replicates.


Weevils irradiated with 125 and 150 Gy died at a
faster rate than unirradiated weevils (Table 1). This

September 2001

Hallman: Irradiation Treatment Against Sweetpotato Weevil


Mean + SEM adult progeny/female
Irradiation Mean + SEM days Mean % + SEM
dose (Gy) Irradiated Control' until 100%mortality control still alive

125 0.014 + 0.0078 41.8 + 7.4 32.6 + 3.4 53.0 + 7.3
150 0 33.3 + 5.9 31.5 + 1.6 57.2 + 8.3

'Control terminated when all irradiated weevils in same replicate died. Therefore, adult progeny/female in control expected to be greater.

is usually, but not always, the case for insects irra-
diated near the minimum doses that provide steril-
ity (Hallman 2000). Complete mortality of
irradiated weevils occurred at a mean of 32.6 days
at 125 Gy and 31.5 days at 150 Gy. During the re-
search done at 125 and 150 Gy, respectively, 53 and
57% of unirradiated weevils were still alive the day
the last irradiated weevils died. Reproduction,
based on the number of F, adult weevils found in
sweetpotatoes offered to irradiated weevils aver-
aged 0.014 and 0 per female at 125 and 150 Gy, re-
The upper range of dosimetry readings when
the target dose was set at 150 Gy was 165 Gy;
therefore, 165 Gy should be used as the recom-
mended dose for quarantine security of sweetpo-
tato weevil adults.
Even the unirradiated control sweetpotato tu-
bers showed some mottling 10 minutes after the
roots were removed from the hot water. However,
the roots irradiated with 500 Gy and then cooked
showed consistent and dark mottling. Cooked tu-
bers that had been irradiated with 300 Gy were no
more discolored than the control. In 2 of the repli-
cates at 400 Gy considerable discoloration, more
than in the control, occurred while in the other 2
replicates at 400 Gy the degree of discoloration after
cooking was not greater than in the control. The in-
formal organoleptic panel found no differences
among the cooked roots, even considering that those
exposed to 500 Gy did not look as pleasing as the
others. Mean (SEM) organoleptic values were 5.8
(0.2), 5.7 (0.3), 5.6 (0.3), 5.5 (0.3), and 5.6 (0.2) for 0
(control), 200, 300, 400, and 500 Gy, respectively.


The information from this study was submit-
ted via the Florida Department of Plant Industry
to the California Department of Agriculture to en-
able an irradiation quarantine treatment to be
applied to Florida sweetpotatoes, including 'boni-
atos' for shipment to California. It was approved
effective April 1, 2000, and the first shipments oc-
curred in late May. Although 30,655 adults were
eventually irradiated with a target dose of 150
Gy, authorities in California accepted the treat-
ment after research with only 18,800 of the adults
had been completed. California stipulated that
sweetpotatoes be packed in cardboard boxes with-

out holes before irradiation to reduce the chance
of post-treatment re-infestation.
Because sweetpotatoes irradiated with >400
Gy sometimes showed mottling after cooking,
there is not a comfortable margin between the
minimum dose required for quarantine security
(165 Gy) and the maximum which should be al-
lowed to prevent possible detrimental affects to
commodity quality. If sweetpotatoes were irradi-
ated in standard pallet-loads, some tubers in the
interior of the load would probably receive at
least 500 Gy and risk objectionable post-cooking
coloration. Sweetpotatoes will probably need to be
irradiated in narrower units than the standard
pallet, increasing the cost of treatment because of
the extra manipulation required to break down
and re-stack pallets.
This case is the first acceptance of an ionizing
irradiation quarantine treatment involving adult
insects and advances the transfer of this promis-
ing technology significantly because it sets a pre-
cedence for dealing with live adults found in
properly irradiated commodities. Acceptance of
live, but sterile, adults by inspectors shows a great
deal of confidence in irradiation and this research.

Evelio Sardina of Homestead, Fla., is thanked for
providing sweetpotato weevil and 'boniato' sweetpota-
toes. Miguel Diaz and Sandra Ramos, USDA, ARS,
Weslaco, are acknowledged for their technical help.
Walter Gould, USDA-ARS-Miami, is thanked for re-
viewing the manuscript.


AND P. A. LORETAN. 1987. Sensitivity of sweetpotato
weevil (Coleoptera: Curculionidae) to gamma radia-
tion. J. Econ. Entomol. 80: 142-146.
HALLMAN, G. J. 1999. Ionizing radiation quarantine
treatments against tephritid fruit flies. Postharvest
Biol. Technol. 16: 93-106.
HALLMAN, G. J. 2000. Expanding radiation quarantine
treatments beyond fruit flies. Agric. Forest Entomol.
2: 1-11.
HALLMAN, G. J., AND D. S. CHALOT. 1993. Possible quar-
antine treatments for Florida agricultural food com-
modities. Proc. Florida State Hort. Soc. 106: 240-243.
MCGUIRE, R. G., AND J. L. SHARP. 1995. Market quality
of sweetpotatoes after gamma-irradiation for weevil
control. HortSci. 30: 1049-1051.

Florida Entomologist 84(3)

September 2001


Department of Zoology, University of Florida, Gainesville, FL 32611-8525

Current address: 120 W 45th St., 39th Fl., New York, NY 10036


It is not known how pleometrosis (nest initiation in groups) and haplometrosis (nest initia-
tion alone) are both maintained in the paper wasp Mischocyttarus mexicanus (Saussure). To
answer this question, reliable measurements of the reproductive success of each tactic are
needed. It is shown here that females that begin nests alone are more likely to raise a few
daughters in rapid succession rather than many daughters at the same time. Females in
small groups or alone also tend to have smaller first daughters than those females working
in large groups. This difference in resource allocation between small and large groups causes
measurements of per capital rates of production to correlate differently with group size de-
pending on whether the number of cells, number of offspring, or weight of offspring added
per day is measured. These data are consistent with the observation that haplometrotic fe-
males receive more predator and conspecific attacks than pleometrotic females, and thus
produce their first daughters quickly to guard the nest. In addition the chronic mystery of a
negative correlation between per capital productivity and group size in social insects is
shown to be an expected outcome and not necessarily an indication that efficiency decreases
with an increase in group size.

Key Words: Mischocyttarus mexicanus, paper wasps, efficiency, social behavior, Polistinae,
per capital productivity


No se sabe como pleometrosis (iniciaci6n de nido en grupos) y haplometrosis (iniciaci6n de
nido solo) son mantenidos en la avispa de papel Mischocyttarus mexicanus (Saussure). Para
contestar esta pregunta, se necesitan medidas confiables del 6xito reproductive de cada tac-
tica. Se demuestra aqui que hembras que comienzan nidos solos son mas propensas a criar
unas pocas hijas muy rdpidamente en vez de muchas hijas al mismo tiempo. Hembras en
grupos pequenos o solas tambi6n tienden a hacer sus primeras hijas mas pequenas que
aquellas hembras trabajando en grupos grandes. Esta diferencia en asignaci6n de recursos
entire grupos pequenos y grandes causa que evaluaciones de producci6n promedio per capital
sean correlacionadas diferentemente con el tamaho del grupo dependiendo en que el numero
de c61lulas, numero de crias, o el peso de la cria sumado por dia sea evaluado. Estos datos apo-
yan la noci6n que hembras haplometroticas reciben mas ataques de predadores y conespeci-
ficos que hembras pleometroticas, y por lo tanto produce su primera hija rdpidamente para
proteger el nido. Adicionalmente, el misterio cr6nico de una correlaci6n negative entire pro-
ductividad per capital y tamaho del grupo en insects sociales es demostrado ser un resultado
esperado y no una debilitaci6n de la hip6tesis que grupos mayores son mas eficientes.

When females of Mischocyttarus mexicanus
(Saussure) (a Polistine paper wasp) begin nests,
they can be found doing this alone or in groups of
sisters (Litte 1977). The existence of solitary nest-
founding (haplometrosis) together with group
nest-founding (pleometrosis) is common in paper
wasps (West-Eberhard 1967; Litte 1981; Strass-
mann 1983; Reeve 1991; Gadagkar 1996) and
some other hymenopterans (Michener 1964;
Mintzer 1979; Tschinkel and Howard 1983; Mint-
zer & Vinso 1985; Rissing & Pollock 1987, 1988;
Stark 1992). Since haplometrotic and pleome-
trotic sisters can often be found working near one
another, it is compelling to hypothesize that these
two modes of nest initiation are the result of deci-

sions made by females: they must decide whether
to join a sister, and sisters must decide whether to
accept help (Strassmann 1996; Clouse 1997). How
these two tactics are maintained in the population
is an exciting topic for those who study the selec-
tive advantages of social and solitary behavior.
It has become increasingly clear that for many
Polistines, nests require guarding to survive, and
this may be a driving force behind the evolution
and maintenance of pleometrosis. We know al-
ready that nests of M. mexicanus and other paper
wasps suffer continuous intrusions by conspecifics
which prey on larvae and/or usurp the current
foundress, and we know that lone females suffer
more from these attacks (Gamboa 1978; Makino &

Clouse: Output of Mischocyttarus mexicanus Nests

Sayama 1991; Kasuya et al. 1980; Kasuya 1982;
Klahn 1988; Gamboa et al. 1992; Clouse 1995; Kat-
ada & Iwahashi 1996) as well as attacks by ants,
birds, and other predators and parasites (Yamane
1996). This is because lone females cannot both
guard their nests and forage, so their nests are left
vulnerable for part of every day. Moreover, since
initial attacks on haplometrotic nests are more
successful than those against pleometrotic nests,
haplometrotic nests probably receive a higher rate
of return attacks than pleometrotic ones. However,
haplometrotic females seem to compensate by
making numerous and brief foraging trips, and, at
least in some populations, haplometrotic females
are larger than even the highest-ranking pleome-
trotic females (Clouse 1997).
The observation that M. mexicanus nests suf-
fer regular intrusions is consistent with the find-
ing that adult females on a nest are significantly
less related than full sisters (Strassmann et al.
1995). Strassmann et al. (1995) also asserted that
M. mexicanus females mate only once, and Litte
(1977) observed that nests had only a single egg-
layer, so queen replacement-whether by daugh-
ters, co-foundresses, or outside usurpers-is the
most probable explanation for low relatedness.
Since hymenopteran sisters who share both par-
ents are more related to each other than to their
own offspring, low relatedness between hy-
menopteran females of any species disqualifies
perhaps the most elegant explanation for their
Even if pleometrotic females are raising rela-
tively unrelated nieces, if they are producing
many more of them than they would alone, low
relatedness may not matter; Strassmann et al.
(1995) suggest that Litte's (1977) data on nest
sizes, survivorship, and production rates support
this hypothesis. Indeed, many studies of social in-
sects (including M. mexicanus (Litte 1977)) have
focused on comparing the productivity of pleome-
trotic and haplometrotic nests (Table 3). Investi-
gators divided some measure of reproductive
output (cells, eggs, larvae, etc.) by the number of
foundreses, obtaining a per capital productivity
statistic for each female that could be compared
across groups of various sizes (Gadagkar 1996).
However, the results of such studies were almost
always that females could expect to produce fewer
offspring if they worked in larger groups (Brian
1953, 1956; Michener 1964; West-Eberhard 1967;
Gibo 1974; Hermann & Dirks 1975; Gibo 1978,
Noonan 1981; Strassmann 1981; It6 1987, Klahn
1988; Queller & Strassmann 1988; Wenzel &
Pickering 1991; Tschinkel 1993). Not only have
these results thwarted another hypothesis for the
evolution of social behavior in insects, but they
have also been interpreted as running counter to
the intuitive and theoretically defensible (Queller
1996) notion that the costs of working in groups
are offset by gains in efficiency. Thus, per capital

productivity data in social insects have become a
serious snag in our understanding of the evolu-
tion of social behavior in general.
Recognizing that there were fundamental prob-
lems with the measures of per capital productivity
to date, I did this study to obtain improved mea-
sures of per capital productivity for a social insect.
First, previous measures rest on the assumption
that all females have the same intrinsic reproduc-
tive potential, an assumption that is probably not
true and not testable (Clouse 1997); so I attempted
manipulating nests such that females could not
control the size of the group to which they be-
longed. Second, by discounting failed nests, previ-
ous studies did not count the output (albeit, zero)
for many foundresses, so I kept a record of nest
surviorship and presumed causes of mortality.
Third, the types of output measures chosen by pre-
vious studies were subject to different biases if fe-
males altered the way they allocated resources in
small versus large groups. It has been shown al-
ready that colonies of the fire ant Solenopsis in-
victa Buren produce smaller daughters when
foundress associations are large (Goodisman &
Ross 1996). Thus, I collected different types of out-
put data for the same nests. And finally, previous
measures often did not factor in the time required
to produce the measured output, so I calculated
output for all nests as a daily rate of production. In
addition, the interpretation of per capital produc-
tivity in the broader study of the evolution of social
behavior is revisited in the Discussion.


Mischocyttarus mexicanus is well-suited for
studying the selective advantage, and accordingly
reproductive output, of pleometrosis and haplom-
etrosis in social insects. Like other paper wasps,
they make open paper nests that can observed
and easily manipulated. Being a resident of the
Eastern subtropics, and having evolved from a
tropical genus, M. mexicanus females start new
nests year-round (Litte 1977, Hermann et al.
1985). In addition, all females (even the first
daughters) are apparently capable of being the
principal egg-layer on the nest. The females are
timid relative to other Polistines (Hermann &
Chao 1984), and they readily nest around build-
ings and on outdoor paraphernalia (wind chimes,
ladders, etc.).
I conducted all work at Archbold Biological
Station, Highlands County, Florida, between 10
May and 31 July, 1993. I used three different sets
of nest to measure various parameters of produc-
tion: Manipulated Pleometrotic and Haplome-
trotic, Restarted, and Unmanipulated nests. The
methods are arranged by nest type, and the re-
sults are arranged by production measurement.
All data are reported as (mean + standard error)
unless otherwise noted.

Florida Entomologist 84(3)

Set I. Manipulated Pleometrotic and Haplometrotic

The main goal of studying Set I was to mea-
sure different rates of production on nests for
which I had manipulated the group size. I wanted
nests that were as close to the first day of initia-
tion as possible, and I wanted pleometrotic fe-
males to end up in small or large groups with
equal probability after manipulation. Nests were
found in saw palmetto (Serenoa repens) along
roadsides and paths, and only those that had only
eggs were included. Upon discovery, the initial
size and shape of each nest was recorded and
drawn. An attempt was made to control for group
size by removing foundresses at night. I removed
one female from nests that had two females, two
or one female alternately from nests that had
three females, and the appropriate number of fe-
males from larger nests to make nests with four
or one female alternately. Females were removed
by disturbing them with a pine needle until they
walked onto the needle or tried to sting it, where-
upon they were placed in a vial and frozen later.
Many nests required more than one night to re-
move the required number of females, since fe-
males often dropped off their nests when
disturbed. Females already found working alone
and whose nests had only eggs were harassed to
mimic the disturbance caused by removing
foundresses. They were touched with a stick at
night for several minutes, often to the point
where they left the nest for the rest of the night.
The number of females on each nest was recorded
every night, and these data were used to calculate
an average number of females working on each
nest per day.
When the most mature larva spun a cocoon in
which to pupate ("cell capping"), I collected the
entire nest. The number of cells and offspring
added since the nest was first discovered were re-
corded. Then the offspring were removed, dried at
60C for five hours, and weighed to the nearest
0.01 mg. Eggs adhered too tightly to the nest pa-
per to be removed and were included with the
weight of the nest paper. The nest paper was cut
back to the size upon discovery, and the paper
added since discovery was dried and weighed.
Four rates of daily per capital production were
generated from these measurements: number of
new cells, number of offspring, total weight of
nest product, and weight of offspring per female
per day. Only successful nests were used in final
calculations, and for one nest, ambiguities over
its size at collection forced me to exclude it from
measures of cell and offspring addition.

Set II. Restarted Nests

As the study of Set I progressed, it became
clear that most would not survive long enough to

obtain production data. The goal of studying Set
II was to obtain a sample of nests for which I had
determined the group size, and that had enough
females to survive to first pupation. At night I cut
down nests that were large enough to have pro-
duced daughters, and on the following night I
searched nearby leaves for the restarted nests.
For such nests, it is impossible to determine if the
foundresses were haplometrotic or pleometrotic,
since daughters and subordinate foundresses are
indistinguishable. The group sizes were altered to
form groups with either (1) four or more females
or (2) less than four females. The number of fe-
males was recorded each night, and when the first
cell capped on a nest, the entire nest was col-
lected. The offspring that capped their cells (pre-
pupae) were removed, dried, and weighed. When
a nest had more than one offspring cap its cell, the
pre-pupal weights were averaged to produce one
weight for each nest. Per capital rates of produc-
tion were calculated in the same way as for Set I

Set III. Unmanipulated Nests

The goal in studying Set III was to obtain a
large sample of unmanipulated nests from which
to determine how females in different sized
groups allocate resources differently among their
offspring. I conducted a survey of 51 pre-eclosion
nests between 10 May and 15 May 1993. Each
nest was censused at night and then collected,
whereupon the numbers of eggs, first through
fifth instar larvae, and pupae were recorded. Pre-
eclosion nests were easy to recognize by the fact
that they had their oldest offspring in the center
cells (the first cells built), and any cells large
enough to contain pupae did not show signs of
previous occupation (e.g., meconium).
It was obvious from the initial survey that
some foundresses had put their efforts into a few
offspring rather than continually adding new
ones. For example, a nest with one fifth-instar
larva and two eggs had clearly concentrated re-
sources on the one large offspring more than a
nest with one third-instar larva, two first and sec-
ond instars, and three or four eggs. However, it
was not possible to immediately compare a nests
that had a more scattered array of larval sizes.
For example, the degree of concentration of nests
that did not have any older larvae, just a few sec-
ond- or third-instars, was not easily compared to
nests that had older larvae and no eggs. Therefore
I used data on the size and number of offspring to
calculate a single measure of how concentrated
resources were in the oldest offspring for each
nest. I assigned each offspring to a size class be-
tween one and seven (egg = 1, first instar = 2, ...
fifth instar = 6, pupa = 7). I divided the size class
value of the oldest offspring on each nest by the
quantity of the youngest offspring. For example, if

September 2001

Clouse: Output of Mischocyttarus mexicanus Nests

a nest had six eggs, four second-instars, three
fourth-instars, and one pupa, I divided "7" (for the
pupa) by "6" (for the number of eggs). This mea-
sure I refer to as "concentration," and for this hy-
pothetical nest, the concentration is 1.17. It is not
as concentrated as another hypothetical nest that
has one pupa and two eggs (concentration = 3.5),
but it is more concentrated than a nest that has
two fifth-instars, two fourth-instars, and six
third-instars (concentration = 6 + 6 = 1).


I. Per capital rates of production

Ninety-nine nests were initially included in
Set I. Thirty-six percent were begun by one fe-
male, 26% were begun by two females, 21% by
three females, nine percent by four females, and
eight percent by five to nine females. The removal
of foundresses from pleometrotic nests was not ef-
fective in assigning females to group sizes with-
out respect to their initial group size: even after
the manipulation, the average number of females
on nests that originally had four or more females
(mean = 3.17, SD = 0.30, N = 17) was significantly
higher than the average number of females on
nests that originally had three females (1.73 +
0.58, N = 18; Fisher's PLSD; P < 0.01) and those
that originally had two females (1.39 0.09, N =
25; Fisher's PLSD; P < 0.01). In addition, 80% of
nests did not survive for more than 20 days.
There were enough survived pleometrotic
nests in Set I to measure productivity; however,
different methods for measuring productivity on
survived nests gave contradictory results. The
number of cells added per female per day did not
correlate with the average number of females
(Spearman Rank Correlation; N = 10; rs = 0.33;
P = 0.32). However, there was a significant posi-
tive correlation between the number of offspring
added per female per day and the average num-
ber of females (Spearman Rank Correlation; N =
10; rs = 0.71; P = 0.03). There was also a signifi-
cant negative correlation between the average
number of females per day and both the total mg
of nest product added per female per day (Spear-
man Rank Correlation; N = 11; rs = -0.70; P =

0.03) and the mg of offspring added per female
per day (Spearman Rank Correlation; N = 11; rs =
-0.61; P = 0.05).
I was successful in altering the group sizes in
restarted nests (Set II) such that the group sizes
before and after manipulation did not correlate
(Spearman Rank Correlation; n = 23; P = 0.34).
Analyzing just survived nests, there was no corre-
lation between the average number of females on
the nest per day and any of the four per capital
measures of daily production.
Manipulated haplometrotic females from Set I
were more productive than both manipulated ple-
ometrotic nests in Set I and restarted nests (Set
II) regardless of the production measure used
(Mann-Whitney U; P < 0.02 for all comparisons)
(Table 1).

II. The size of the largest offspring

The average weights of all pre-pupae from Sets
I and II combined were positively correlated with
the average number of females on their nest of or-
igin (Spearman Rank Correlation; N = 34; rs =
0.63; P = 0.02). The average weight (mg) of the
first pre-pupa on restarted nests (Set II) tended to
be positively correlated with the average number
of females on the nest (Spearman Rank Correla-
tion; N = 13; rs = 0.52; P = 0.07).
Restarted nests (Set II) had larger pre-pupae
than manipulated haplometrotic females (1.4
0.06 mg, n = 13 versus 1.3 0.03, n = 12; Mann-
Whitney U; P = 0.01), and manipulated pleome-
trotic females (Set I) (1.1 + 0.09, n = 10; P < 0.01).
Manipulated pleometrotic (Set I) and haplome-
trotic pre-pupae were not significantly different
in size (P = 0.16)

III. Concentration

Nests from Set III were more concentrated
when being built by fewer females. For collected
nests, concentration ratios for one-female, two-fe-
male, three-female, and four or more-female nests
were significantly different (Table 2; Kruskal-
Wallis; P < 0.025). From Set I, manipulated hap-
lometrotic nests had higher concentration mea-
sures than manipulated pleometrotic nests (4.31

Measurement Manipulated Manipulated N Restarted N
Haplometrotic Pleometrotic
# cells/female/day 0.28 + 0.02 12 0.07 + 0.01 10 0.15 + 0.01 15
# offspring/female/day 0.25 + 0.03 12 0.01 + 0.03 10 0.13 + 0.01 15
mg offspring/female/day 0.96 + 0.17 12 0.57 + 0.10 11 0.52 + 0.06 15
total mg/female/day 1.44 + 0.22 12 0.96 + 0.18 11 0.84 + 0.08 15

Florida Entomologist 84(3)

+ 0.41, N = 13, versus 2.34 + 0.50, N = 12; Mann-
Whitney U; P < 0.001). From Set II, restarted
nests with less than four females (N = 8, concent.
= 1.69 0.34) did not have significantly different
concentration ratios from restarted nests with
more than four females (N = 11, 1.35 0.29;
Mann-Whitney U; 0.25 > P > 0.15), although the
trend was similar to sets I and III.

IV. Time to cell capping

Restarted nests (the only nests I followed since
initiation), took longer to raise a daughter to pre-
pupal stage if there were less than four females on
average working on the nest (25.33 1.24 days
versus 22 + 0.45 days; Mann-Whitney U; P < 0.01).


The mortality rates of new nests make it clear
why measuring the final production of reproduc-
tive offspring has not yet been done: the mortality
rate for new nests is so high (80% failed within 20
days), one would have to follow several hundred
nests to have a few left for analysis in the final
stages. Moreover, it indicates that when address-
ing the question of what a female can expect to
produce, the chance of nest failure (producing
nothing) must be factored into the calculation.
Among nests that did survive, two processes
heavily influence measurements of production dur-
ing the pre-eclosion stage in M. mexicanus. First,
small nests seem to rush the production of their
first adult daughter. This is supported here by the
fact that (1) in Set I the per capital rate of adding
new offspring is larger in bigger groups, but these
bigger groups have a smaller per capital rate of add-
ing biomass, (2) the "concentration" ratio was
higher for nests attended by one female than by
several, and (3) smaller restarted groups lagged be-
hind large ones in the time to cell capping by only
two to three days. Each of these results is what we
would expect if females in small groups, especially
haplometrotic females, primarily fed their oldest
daughter and laid few additional eggs. The fact that
the first daughter on smaller nests tended to be
smaller than those from larger associations is also
consistent with the idea that small nests rush their
first daughter to eclosion; perhaps the first daugh-
ter herself decides to pupate early, since her own life
is at stake the longer the nest lacks extra guards.

The second factor biasing productivity mea-
sures is that surviving haplometrotic females have
much higher daily rates of production than ple-
ometrotic females, regardless of what type of out-
put one measures. It can be legitimately argued
that not having lost sisters or their original nest,
the lone females in this study were not nearly as
traumatized as the other females in this study
(and thus were more productive). But I disturbed
lone females to the point that I thought they might
abandon their nests, and they naturally have great
demands placed on them daily by the need to pro-
cure prey, water, paper, and nectar alone. Nonethe-
less, even while concentrating efforts on the oldest
offspring to a greater extent than any other nests,
they added more cells per female per day and more
biomass per day than any other group. Although
the daily per capital rate of cell addition declines as
group size decreases in pleometrotic nests (Set I),
it rises sharply again for lone females, and thus
lone females stand apart from the overall produc-
tion trends. Lone female production is so much
larger than group production in this study, produc-
tion analyses that assume that the only behavioral
difference between haplometrotic and pleometrotic
females is their choice in the number of nesting as-
sociates should not be accepted.
Since per capital productivity has been used to
compare the reproductive output of pleometrotic
and haplometrotic females, and productivity has
been the axis of discussions about synergy in in-
sect societies, per capital productivity has been
equated with "efficiency." Thus, for much of the
past forty-five years, productivity measures in so-
cial insects have led to discussions of the larger
question of why social behavior is apparently inef-
ficient. However, per capital productivity merely
measures the marginal productivity of each addi-
tional worker, and diminishing marginal returns
from adding sisters, or any other factor of produc-
tion, is quite expected (Krebs & Davies 1987). This
is because as one adds additional units of a produc-
tion input, while holding others constant, the ad-
ditional units become increasingly redundant.
(Interestingly, a few human examples exist of in-
creasing marginal returns during the initial
stages of production, and some wasp data reflect
this phenomenon when foundress group size in-
creases from one to two females (West-Eberhard
1967, Metcalf & Whitt 1977, Litte 1981, Noonan
1981, Strassmann 1981).)


Number of Foundresses 1 2 3 >3
N=28 N=8 N=7 N=8

Concentration 2.09 + 0.35 1.03 +0.24 0.70 +0.14 0.35 +0.16

September 2001

Clouse: Output of Mischocyttarus mexicanus Nests


Family Species meas. corr.

Halictidae Pseudagapostemon divaricatus1 F -
Augochloropsis sparsilis1 F -
Lasioglossum imitatum1 F -
Lasioglossum rhytidophorum1 F -

Apidae Apis mellifera1 CC -
Bombus americanum1 0 +

Formicidae Mymica rubra2 B, L, P -
Myrmica rubra macrogyna L -
Solenopsis invicta3 0, R, B, J -

Vespidae Polybia bistriata & P. bicyttarella E -
49 nests from 11 Polybine species1 E -
Polistes fuscatus4 C -
P. fuscatus' C, 0 n.s.
P. fuscatus6 0
P. fuscatus7 R n.s.
P. annularis8 C
P. annularis' I
P. annularis10 R
P. metricus11 E n.s.
P. metricus12 R +
P. metricus13 R n.s.
P. chinensis antennalis14 E, C -
Mischocyttarus mexicanus15 C -
M. labiatus16 C, P
Ropalidia fasciata" C -
R. marginata1" 0 -

'Michener (1964)
'Brian (1953, 1956)
Tshinkel (1993)
'West-Eberhard (1993)
'Gibo (1974)
0Gibo (1978)
'Noonan (1981)
'Hermann and Dirks (1975)
'Strassmann (1981)
"Queller and Strassmann (1988)
"Bohm (1977)
"Metcalf and Whitt (1977)
"Gamboa (1978)
"Hoshikawa (1979)
"Litte (1977)
"Litte (1981)
"It6 (1987)
"Shakarad and Gadagkar (1993)

True efficiency measurements in social insects
await refinement of a system by which total en-
ergy input can be accurately measured (such as in
Suzuki 1981), because "efficiency" is a ratio of out-
put to input (Brian 1953; Jeanne 1986). Using
output measures which encompass total output
and are thus free from the resource-allocation bi-

ases shown here should provide novel and inter-
esting efficiency data.
Moreover, if productivity, survivorship, and re-
latedness can be combined to calculate reliable
expectations of reproductive success for haplome-
trotic and pleometrotic foundresses, it could open
new doors of research into the maintenance of so-

cial and solitary behavior. One possibility is that
these tactics are evolutionarily stable strategies
in which mothers deliberately make some large
reproductive daughters who can keep the hectic
pace of working alone and some small daughters
who can work more slowly in groups. The overall
payoff could be the same for both types of females
if the large females suffer more from attacks and
lose more nests, but if they survive, their nests
produce more reproductive offspring in the end
than pleometrotic ones. Another possibility is
that one or the other strategy is more successful
but can be adopted only under certain circum-
stances. Haplometrosis-naturally desirable since
the female gets to lay all the eggs-may require a
minimal body size and fat store to defend the nest
and make numerous foraging trips; pleometro-
sis-also desirable since foundresses get to pro-
duce on a relatively well-guarded nest-may
require having and finding certain types of sisters
to minimize fighting between foundresses.


Drs. H. J. Brockmann, J. F. Anderson, H. G. Hall, P.
Landolt, and M. Deyrup read and reread this manu-
script and offered crucial suggestions. Archbold Biologi-
cal Station, especially Drs. M. Deyrup and B. Ferster,
provided equipment and support. The University of
Florida Department of Zoology provided computer sup-
port. This project was funded in part by a grant from
Sigma Xi.


BOHM, M. K., AND K. A. STOCKHAMMER. 1977. The nest-
ing cycle of a paper wasp, Polistes metricus (Hy-
menoptera: Vespidae). J. Kansas Entomol. Soc. 50:
BRIAN, M. V. 1953. Brood-rearing in relation to worker
number in the ant Myrmica. Physiol. Zool. 26: 355-
BRIAN, M. V. 1956. Group form and causes of worker in-
efficiency in the ant Myrmica rubra L. Physiol. Zool.
29: 173-194.
CLOUSE, R. M. 1995. Nest usurpation and intercolonial
cannibalism in Mischocyttarus mexicanus (Hy-
menoptera: Vespidae). J. Kansas Entomol. Soc. 68:
CLOUSE, R. M. 1997. Are lone paper wasp foundresses
mainly the result of sister mortality? Florida Scien-
tist 60(4): 265-274.
GADAGKAR, R. 1996. The evolution of eusociality, includ-
ing a review of the social status of Ropalidia margi-
nata, pp. 248-271. In S. Turillazzi and M. J. West-
Eberhard [eds.] Natural History and Evolution of
Paper-Wasps. New York: Oxford University Press.
GAMBOA, G. J. 1978. Intraspecific defense: Advantage of
social cooperation among paper wasp foundresses.
Science. 199: 1463-1465.
1978. Division of labor and subordinate longevity in
foundress associations of the paper wasp, Polistes
metricus (Hymenoptera: Vespidae). J. Kansas Ento-
mol. Soc. 51: 343-352.

September 2001

SON, AND T. G. FISHWIND. 1992. Defense against in-
traspecific usurpation by paper wasp cofoundresses
(Polistes fuscatus, Hymenoptera: Vespidae). Cana-
dian J. Zool. 70: 2369-2372.
GIBO, D. L. 1974. A laboratory study on the selective ad-
vantage of foundress associations in Polistes fusca-
tus (Hymenoptera: Vespidae). Canadian Ent. 106:
GIBO, D. L. 1978. The selective advantage of foundress
associations in Polistes fuscatus (Hymenoptera:
Vespidae): a field study of the effects of predation on
productivity. Canadian Ent. 110: 519-540.
GOODISMAN, M. A. D., AND K. G. Ross. 1996. Relation-
ship of queen size and worker number in polygyne
colonies of the fire ant Solenopsis invicta. Insectes
Soc. 43: 303-307.
HERMANN, H. R. AND J.-T. CHAO. 1984. Nesting biology
and defensive behavior ofMischocyttarus mexicanus
cubicola (Vespidae: Polistinae). Psyche. 91: 51-65.
HERMANN, H. R. AND T. F. DIRKS. 1975. Biology of
Polistes annularis (Hymenoptera: Vespidae). I.
Spring Behavior. Psyche. 82: 97-108.
1985. Mischocyttarus mexicanus cubicola (Hy-
menoptera), distribution and nesting plants. Florida
Entomol. 68: 609-614.
HOSHIKAWA, T. 1979. Observations on the polygynous
nests of Polistes chinensis antennalis Perez
(Hymenoptera: Vespidae) in Japan. Kontyu. 47: 239-
ITO, Y. 1987. Role of pleometrosis in the evolution of eu-
sociality in wasps, pp. 17-34. In Y. It6, J. L. Brown
and J. Kikkawa [eds.] Animal Societies: Theories
and Facts. Tokyo: Japan Sci. Soc.
JEANNE, R. L. 1986. The organization of work in Polybia
occidentalis: costs and benefits of specialization in a
social wasp. Behav. Ecol. Sociobiol. 19: 333-341.
KASUYA, E. 1982. Take-over of nests in a Japanese pa-
per wasp, Polistes chinensis antennalis (Hy-
menoptera: Vespidae). Appl. Ent. Zool. 17: 427-431.
KASUYA, E., Y. HIBINO, AND Y. ITO. 1980. On "intercolo-
nial" cannibalism in Japanese paper wasps, Polistes
chinensis antennalis Perez and P. jadwigae Dalla
Torre (Hym., Vespidae). Res. Pop. Ecol. 22: 255-262.
KATADA, S., AND 0. IWAHASHI. 1996. Characteristics of
usurped colonies in the subtropical paper wasp,
Ropalidia fasciata (Hymenoptera: Vespidae). In-
sectes Soc. 43: 247-253.
KLAHN, J. E. 1988. Intraspecific comb usurpation in the
social wasp Polistes fuscatus. Behav. Ecol. Sociobiol.
23: 1-8.
KREBS, J. R., AND N. B. DAVIES. 1987. An Introduction
to Behavioral Ecology. Oxford: Blackwell Scientific
LITTE, M. 1977. Behavioral ecology of the social wasp
Mischocyttarus mexicanus. Behav. Ecol. Sociobiol. 2:
LITTE, M. 1981. Social biology of the Polistine wasp Mis-
chocyttarus labiatus: survival in a Colombian rain
forest. Smithson. Contr. Zool. 327: 1-27.
MAKINO, S., AND K. SAYAMA. 1991. Comparison of in-
traspecific usurpation between two haplometrotic
paper wasp species (Hymenoptera: Vespidae:
Polistes). J. Ethol. 9: 121-128.
METCALF, R. A., AND G. S. WHITT. 1977. Relative inclu-
sive fitness in the social wasp Polistes metricus. Be-
hav. Ecol. Sociobiol. 2: 353-360.

Florida Entomologist 84(3)

Clouse: Output of Mischocyttarus mexicanus Nests

MICHENER, C. D. 1964. Reproductive efficiency in rela-
tion to colony size in Hymenopterous societies. In-
sectes Soc. 4: 317-342.
MINTZER, A. 1979. Colony foundation and pleometrosis
in Camponotus (Hymnenoptera: Formicidae). Pan-
Pacific Entomol. 55: 81-89.
MINTZER, A., AND S. B. VINSO. 1985. Cooperative colony
foundation by females of the leaf cutting ant Atta
texana in the laboratory. J. New York Entomol. Soc.
93: 1047-1051.
NOONAN, K. M. 1981. Individual strategies of inclusive-
fitness-maximizing in Polistes fuscatus foundresses,
pp. 18-4. In R. D. Alexander and D. W. Tinkle [eds.]
Natural Selection and Social Behavior. New York:
Chiron Press, Inc.
QUELLER, D. C.1996. The origin and maintenance of eu-
sociality: the advantage of extended parental care,
pp. 218-234. In S. Turillazzi and M. J. West-Eber-
hard [eds.] Natural History and Evolution of Paper-
Wasps. New York: Oxford University Press.
QUELLER, D. C., AND J. E. STRASSMANN. 1988. Repro-
ductive success and group nesting in a paper wasp,
Polistes annularis, pp. 76-96. In T. H. Clutton-Brock
[ed.] Reproductive Success: Studies of Individual
Variation in Contrasting Breeding Systems.
REEVE, H. K. 1991. Polistes, pp. 99-148. In K. G. Ross
and R. W. Matthews [eds.] The Social Biology of
Wasps. Ithaca: Cornell University Press.
RISSING, S. W., AND G. B. POLLOCK. 1987. Queen aggres-
sion, pleometrotic advantage and brood raiding in
the ant Veromessor pergandei (Hymenoptera: For-
micidae). Anim. Behav. 35: 975-981.
RISSING, S. W., AND G. B. POLLOCK. 1988. Pleometrosis
and Polygyny in Ants, pp. 179-222. In R. L. Jeanne
[ed.] Interindividual Behavioral Variability in Social
Insects. Boulder, CO: Westview Press.
SHAKARAD, M., AND R. GADAGKAR. 1993. Why are there
multiple-foundress colonies in Ropalidia marginata.
In Proceedings of the XXI Annual Conference of the
Ethological Society of India, Tirupati.
STARK, R. E. 1992. Cooperative nesting in the large mul-
tivoltine carpenter bee Xylocopa sulcatipes Maa

(Apoidea: Anthophoridae): Do helpers gain or lose to
solitary females? Ethology 91: 301-310.
STRASSMANN, J. E. 1981. Wasp reproduction and kin se-
lection: reproductive competition and dominance hi-
erarchies among Polistes annularis foundresses.
Florida Entomol. 64: 74-88.
STRASSMANN, J. E. 1983. Nest fidelity and group size
among foundresses of Polistes annularis (Hy-
menoptera: Vespidae). J. Kansas Entomol. Soc. 56:
STRASSMANN, J. E. 1996. Selective altruism towards
closer over more distant relatives in colonies of the
primitively eusocial wasp, Polistes, pp. 190-201. In
S. Turillazzi and M. J. West-Eberhard [eds.] Natural
History and Evolution of Paper-Wasps. New York:
Oxford University Press.
1995. Genetic relatedness and population structure
in the social wasp, Mischocyttarus mexicanus (Hy-
menoptera: Vespidae). Insectes Soc. 42: 379-383.
SUZUKI, T. 1981. Flesh intake and production of off-
spring in colonies of Polistes chinensis antennalis
(Hymenoptera: Vespidae). II. Flesh intake and the
production of reproductive. Kontyu. 49: 283-301.
TSCHINKEL, W. R. 1993. Sociometry and sociogenesis of
colonies of the fire ant Solenopsis invicta during one
annual cycle. Ecol. Mon. 63: 425-457.
TSCHINKEL, W. R., AND D. F. HOWARD. 1983. Colony
founding by pleometrosis in the fire ant Solenopsis
invicta. Behav. Ecol. Sociobiol. 12: 101-113.
WENZEL, J. W., AND J. PICKERING. 1991. Cooperative
foraging, productivity, and the central limit theo-
rem. Proc. Natl. Acad. Sci. USA. 88: 36-38.
WEST-EBERHARD, M. J. 1967. Foundress associations in
Polistine wasps: dominance hierarchies and the evo-
lution of social behavior. Science. 157: 1584-1585.
YAMANE, S. 1996. Ecological factors influencing the col-
ony cycle inPolistes wasps, pp. 75-97. In S. Turillazzi
and M. J. West-Eberhard [eds.] Natural History and
Evolution of Paper-Wasps. New York: Oxford Uni-
versity Press.

Florida Entomologist 84(3)


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

6C. .... ... ..... ......... n. sp. is described from soldiers and imagos collected on St. Lucia, West
Indies. It is the seventh described species of West Indian Glyptotermes and is the largest spe-
cies among its Lesser Antillean congeners.

Key Words: taxonomy, new species, Neotropics, West Indies, Lesser Antilles

6. ,L........... amplus n. sp. es descrita de soldados e imagos colectados en St. Lucia, Antillas
Menores. Es la s6ptima especie descrita de la especie antillana Glyptotermes y la mayor es-
pecie entire sus congeneres de las Antillas Menores.

Glyptotermes Froggatt is a tropicopolitan ge-
nus that is the second most diverse in the family
Kalotermitidae after Neotermes Holmgren
(Krishna 1961). The soldier caste of Glyptotermes
is rather variable among species, but most can be
distinguished from those of other kalotermitid
genera by their rather cylindrical head capsule,
steep frons, and short, thickened mandibles. The
soldier head capsule of many Glyptotermes spe-
cies is also characterized by a pair of rounded
frontal protuberances that are separated by a me-
dian depression or cleft. Unlike those of the
closely aligned genus Calcaritermes Snyder, Glyp-
totermes soldiers do not have an enlarged apical
spur on the front tibia. Wing venation of the
imago of Glyptotermes is similar to that of Calcar-
itermes in which the sclerotized media runs close
and parallel to the radial sector to the wing tip.
Glyptotermes spp. are typically found in wet for-
ests infesting sound or rotting wood or wood scars
in live trees.
In the New World, all 25 known species of
Glyptotermes are Neotropical in distribution of
which 6 have been described from the West Indies
(Constantino 1998). Glyptotermes pubescens Sny-
der (1923) and G. liberatus (Snyder) (1929) were
described from Puerto Rico and Jamaica, respec-
tively. Glyptotermes liberatus is also known from
Puerto Rico (Martorell 1973). Glyptotermes adam-
soni, G. parvoculatus, and G tubifer were described
by Krishna & Emerson (1962) from Trinidad and
Tobago, Trinidad only, and St. Vincent, respec-
tively. The description of G. (= Calotermes) posti-
cus (Hagen) (1858) is based on a single dealated
female from St. Thomas U.S.V.I., and according to
Snyder (1929), is probably that of a Cryptotermes
species. Furthermore, Snyder (1929) determined
that the soldier described as Kalotermes posticus

by Banks (1919) was actually that of a new spe-
cies that Snyder (1929) renamed Kalotermes lib-
eratus. Nevertheless, Krishna (1961) included
posticus in his revised species list of Glyptoter-
mes. Two unidentified Glyptotermes spp. were
reported from Dominica and Martinique (Schef-
frahn et al. 1994).
During a 1998 expedition to St. Lucia, a new
species of Glyptotermes was collected. The de-
scriptions of the soldier and imago of Glyptoter-
mes amplus n. sp. are provided herein.


Morphometrics of specimens preserved in
85:15 ethanol: water were made with a stereomi-
croscope fitted with a calibrated ocular microme-
ter. Scanning electron micrograph prints were
scanned at 600 dpi, the digital image outline
traced using photograph-enhancing software
(Photo Magic, Micrografx, Inc., Richardson, TX),
the background converted to black, and the scale
bar digitally redrawn (Scheffrahn et al. 1999).
Latitude and longitude coordinates were mea-
sured at collection sites using a Garmin GPS
model 38 global positioning receiver (Garmin In-
ternational, Olathe, Kansas). Coordinates of col-
lection sites were converted to decimal degrees
and mapped (Fig. 3) using ArcView GIS version
3.0a software and relevant map data from Digital
Map of the World version 1.0 (Environmental Sys-
tems Research Institute, Inc. Redlands, CA).
The holotype soldier and morphotype imago
are deposited in the collection of the American
Museum of Natural History, New York [AMNH].
Paratype soldiers and imagos are deposited in the
National Museum of Natural History (Smithso-
nian Institution), Washington, D.C. [USNM]; the

September 2001

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