Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00006
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 1995
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: VID00006
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access
Resource Identifier: isbn - 0015-4040
issn - 1938-5102
oclc - 33223434

Full Text

Weissling & Giblin-Davis: Rhynchophorus cruentatus diets 225


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


Several artificial diets were evaluated as alternatives to decomposing pineapple
[Ananas comosus (L.) Merrill] for culture of Rhynchophorus cruentatus (F) larvae.
The most suitable diet tested for larval growth and survival was a combination of
canned pineapple, oats, sucrose, molasses, brewers yeast, Wesson's salts, vitamins,
and preservatives. Diets that were not supplemented with brewers yeast provided
poor larval growth and survival. Larvae cultured from artificial diets were placed in
sugarcane (Saccharum officinarum L.) for pupation.

Key Words: Palm weevil, palmetto weevil, laboratory rearing, artificial diets.


Varias dietas artificiales fueron evaluadas como alternatives a la pina descom
puesta [Ananas comosus (L.) Merrill] para el cultivo de larvas de Rhynchophorus
cruentatus (F). La mejor dieta probada para el crecimiento y la supervivencia larval
fu6 una combiaci6n de pina enlatada, cebada, azfcar, melaza, levadura de cerveza, sa
les de Weson, vitamins y preservatives. Las dietas sin levadura de cerveza produce
ron poco crecimiento larval y sobrevivencia. Las larvas criadas en dietas artificiales
fueron colocadas en cana de azfcar (Saccharum officinarum L.) para su pupaci6n.

Rhynchophorus cruentatus (F) is the only species of palm weevil in the continental
United States (Wattanapongsiri 1966). Unlike several of its congeners, R. cruentatus
is not considered a major pest of palms. However, this species will attack transplanted
or otherwise stressed ornamental palms (Giblin-Davis & Howard 1988, 1989). In
Florida, R. cruentatus is sympatric with the native cabbage palmetto, Sabalpalmetto
(Walter) Loddiges ex Schultes (Woodruff 1967), a palm often used as mature speci
mens in landscaping due to its low cost, natural abundance, and high transplanting
Semiochemicals emanating from wounded or dying palms (Chittenden 1902, Wat
tanapongsiri 1966, Weissling et al. 1992, Giblin-Davis et al. 1994) and conspecific
weevils (Weissling et al. 1993, 1994) are attractive to R. cruentatus adults. Females
lay eggs in the leaf bases or directly into the wounds of dying host palms. The larvae
molt several times and are voracious consumers of tissue within the bud and stem.
Last-instar larvae migrate to the periphery of the host, prepare a cocoon from the fi

'Present Address: United States Department of Agriculture, Agricultural Research Service,
3706 W Nob Hill Blvd., Yakima, WA 98902
'To whom correspondence should be addressed.

Florida Entomologist 78(2)

ber, and enter a pupal stage. Following eclosion, adults emerge from the tree. The life
cycle of this weevil usually is completed in less than 84 d (see Giblin-Davis & Howard
1989 for detailed description).
Research on the biology of R. cruentatus, as well as its vector potential for the red
ring nematode [Bursaphelenchus cocophilus (Cobb)], requires the collection of adults
in the field, an expensive and time consuming endeavor. Thus, a laboratory rearing
method for R. cruentatus was needed. Three other species of palm weevils have been
reported to be cultured in the laboratory. Rhynchophorus ferrugineus (Olivier)
(Rananavare et al. 1975) and R. palmarum (L.) (Wilson 1963) can be cultured using
cut petiole or stem tissue of coconut palms, and R. cruentatus has been reared on buds
of an alternate host, Serrenoa repens (Bartram) Small (Berger 1907). However, collec
tion of palm tissue for culture of weevils is also expensive. Rahalkar et al. (1972) re
ported that sugarcane is a good substitute for coconut stem for rearing R. ferrugineus.
This method was improved by incorporating sugarcane in nutrient agar for young lar
vae and whole sugarcane stem pieces for older larvae (Rananavare et al. 1975). Cul
ture of R. ferrugineus was further improved by development of an artificial diet
(Rahalkar et al. 1978, 1985) containing sugarcane bagasse (fiber), coconut cake, yeast,
sucrose, minerals, vitamins, and preservatives. Using a combination of corn flour,
sugarcane fiber, oats, sugar, coconut oil, and propionic acid for a larval medium and
sugarcane for pupation, Sanchez et al. (1993) successfully cultured R. palmarum.
Giblin-Davis et al. (1989) determined that R. cruentatus and R. palmarum could be
cultured on decomposing pineapple [Ananas comosus Merrill] syncarp for young lar
vae and sugarcane stem for mature larvae. However, variation in the quality and
availability of pineapple, and the inconvenience of working with decomposing mate
rial, prompted the development and evaluation of several agar-based diets for
laboratory-culture of R. cruentatus.



R. cruentatus adults were harvested as cocoons from infested S. palmetto, placed
individually in covered 100 ml plastic cups with moistened tissue paper (Giblin-Davis
et al. 1989) and stored at 29 C until adult emergence. Females and males were placed
as individual pairs in 500 ml covered containers with moistened tissue paper and a
slice of apple (Pyrus malus L.). The apple slices were replaced at 1-3 day intervals,
carefully dissected, and the eggs removed (Weissling & Giblin-Davis 1994). Eggs were
transferred to petri dishes (15 x 100 mm) lined with moistened filter paper, sealed
with parafilm, and stored at 29 C until neonate larvae emerged.

Diet Preparation

All diets included 50 g bacto-agar (Difco Laboratories, Detroit, MI), 12.5 crushed
vitamin tablets (Centrum; Lederle Laboratories Div., Pearl River, NY), and 1892 ml
water. Each vitamin tablet weighed 1.4 g and contained: vitamins A (5000 I.U.), E (30
I.U.), C (60 mg), B, (1.5 mg), B, (1.7 mg), B, (2 mg), B,, (6 pg), D (400 I.U.), and K, (25
ig), folic acid (400 pg), niacinamide (20 mg), biotin (30 pg), pantothenic acid (10 mg),
calcium (162 mg), phosphorus (125 mg), iodine (150 pg), iron (18 mg), magnesium (100
mg), copper (2 mg), zinc (15 mg), manganese (2.5 mg), potassium (40 mg), chloride
(36.3 mg), chromium (25 ig), molybdenum (25 ig), selenium (25 ig), nickel (5 ig), tin
(10 pig), silicon (10 pig), and vanadium (10 pig). In addition, all diets contained the pre

June, 1995

Weissling & Giblin-Davis: Rhynchophorus cruentatus diets 227

servatives: m-para-hydroxybenzoate [14% solution in 95% ethyl alcohol (25 ml)], sor
bic acid [12.5% solution in 95% ethyl alcohol (37.5 ml) in preliminary tests or 6.28 g
sorbic acid potassium salt in refined diets], and 4M potassium hydroxide solution (7.5
ml) (Rahalkar et al. 1985). All other ingredients and amounts are listed in Table 1. All
materials except the crushed vitamin capsules and a 500-ml aliquot of water were
blended for approximately 2 min and poured into a 5.7-liter stainless-steel bowl. The
blender was rinsed with the remaining water and poured into the bowl. The mixture
was then autoclaved for 20 min at 120 C. As the diet cooled, the crushed vitamin tab
lets were added and the media was stirred. Diets were poured into diet cups while still
warm. When diets cooled, a small hole was made in the diet surface to facilitate feed
ing and larvae were transferred one per diet cup with a fine camel hair brush. Diet
cups were covered with a lid vented by several small holes. After 24 h, diets were
checked and dead larvae were replaced. All tests were conducted in an environmental
chamber at 29 C with a photoperiod of 13:11 (L:D).

Preliminary Evaluation of Diets

Ten to twenty replicates of seventeen different diets (Table 1) were evaluated in
preliminary tests for larval and adult biomass gain and survival. We also included the
diet for culture of R. ferrugineus (Rahalkar et al. 1985). Experimental diets consisted
of 100 g of diet in 100-ml cups. Preliminary observations indicated the need for fresh
diet if larvae were left in culture for more than three weeks (unpublished data). Thus,
larvae fed diets 1-17 and the published diet were removed from cups after three weeks
and transferred to 100 g of fresh diet. During this transfer, larvae were rinsed with
water, patted dry with paper toweling, and weighed. After an additional two weeks,
larvae were removed, cleaned, and reweighed. To determine if larvae could be cul
tured for more than three weeks without a change in diet, we included diet 18 which
was 200 g of diet 3. However, 100-ml cups were too small for this amount of diet so the
media was placed in 500-ml cups. Larvae on diet 18 remained undisturbed for five
weeks, at which time they were cleaned and weighed. After the final weighing, larvae
from all diets were individually placed into holes (0.95 cm diam) drilled 10 cm into one
end of 23-25 cm lengths of sugarcane (Giblin-Davis et al. 1989). Stems were wrapped
in window screen secured at each end by elastic bands, and placed in an environment
tal chamber at 29 C. After three weeks, stems were split open and cocoons, if present,
were removed and placed in 100-ml vented cups with moistened tissue paper and
stored at 29 C until adult emergence. If larvae were still present, the cane was care
fully re-wrapped, placed back in the environmental chamber, and checked one week

Culture Technique Modifications

The biomass and survival of larvae reared for three weeks on diets 2, 3, 12, 13,
and 15 was good in preliminary tests (Table 2). Therefore, larvae were tested again
on these diets when reared in 100-ml cups (100 g of diet) and then transferred to
sugarcane after three weeks. This was done to determine if healthy weevils were pro
duced under this time and resource saving regimen. Based on acceptable growth of
larvae cultured on diet 18 in 500-ml cups (Table 2), we also compared larval growth
on the diets described above when 200 g of media was placed in 500-ml cups and the
larvae were allowed to feed undisturbed for five weeks before transfer to sugarcane.
Based on poor biomass gain after three weeks in culture, but high final biomass after
five weeks in preliminary tests (Table 2), larvae fed diets 14 and 17 were evaluated

Florida Entomologist 78(2)


N- L



- C


o CD


- C


o LCr~
c' CDLC]

cc1 L

o NC?

N- C

o LO


LO0 00 L

^ i l

m a

~5j = -
m 5O

s a a on C
U t/i (/ ,U Q(l2 t/

June, 1995

- 3 6,





G >oajn
^ 5 ^
| S

^ b ^
GI-3 -

6b O

.ec 0

t^ Ez

-c, M

O gZ roaoo
2 z-5 .
-S a. a C
" S ^-

' s^^a B~
- 3 -'- M c
I IJ0!0
I Ilil

t "811 s
C^-G gI S
.a g -5 '4 | -3 *
Q ^ 00 y

Weissling & Giblin-Davis: Rhynchophorus cruentatus diets 229


+1 +1 +1

blLO n



+1 +1 +1 +1
0 0L0
Cdo *dC

+1 +1 +1 +1

-Q -Q -Q

+1 +1 +1

00 C
O 0

0 -

+1 +1 +1 +1

+1 +1 +1 +1
C ooM O0
O C'


+1 +1 +1 +1
0 00cc
E 'Jc'

-0 -0-D

+1 +1 +1 +1

C, -0
m0 U

+1 +1 +1 +1

- D C'J -



+1 +1 +1 +1
C) cn C )

+1 +1 +1 +1


+1 +1 +1 +1
LO t-~

U 1 11

+1 +1 +1 +1

1 1 +1 1
+1 +1 +1 +1

+1 +1 +1 +1

- LO- c
0 ^-i ^-i C

0 C^C^

S000000000000 000000

P T i -i T- T i -i T- T i -i T-

+1 +1 +1

C0 c'J c

+1 +1 +1

+1 +1 +1

+1 +1 +1

c+ c+ c
+1 +1 +


+1 +1 +1



Florida Entomologist 78(2)

further in 500-ml cups with 200 g diet. The biomass of these larvae was determined
at five weeks before immediate transfer to sugarcane. Each diet and cup size combi
nation was replicated 20 to 40 times.


Larval and adult survival were based on the initial number of test insects. All data
except percentage values were square root (X + 0.001) transformed. Percentage data
were subjected to angular [arcsin (square root + 0.001)] transformation. Transformed
data were subjected to analysis of variance using the Statistical Analysis System's
general linear models procedure (SAS Institute 1985) for overall comparison of diets.
Means were separated by Student Newman-Keuls' (SNK) test where significant (P <
0.05) effects occurred. Diets evaluated in 100-ml cups were compared with diets in
500-ml cups using orthogonal contrasts (SAS Institute 1985).


Preliminary Evaluation of Diets

After five weeks in culture the average biomass of larvae and adults, as well as lar
val survival and percentage adult emergence, varied greatly (Table 2). The greatest
mean larval biomass was observed for diet 15 both at three (3.40 g) and five (4.29 g)
weeks but survival was relatively low (40%) (Table 2). The greatest larval survival
(100%) occurred with larvae fed diet 14; mean biomass at five weeks (3.13 g) was not
significantly different from that of diet 15 (Table 2). There was little gain in biomass
of larvae fed diets 1 and 11, however, survival on diet 11 was relatively high (55%),
whereas survival on diet 1 was low (10%) (Table 2). Adult emergence was greatest for
diet 14 (70%) while no adults emerged from larvae fed diets 1, 5, and 7 (Table 2). Total
time from placement of larvae in sugarcane to adult emergence varied from 48.5 days
(diet 2) to 65 days (diet 11) (Table 2).

Culture Technique Modifications

Overall, the greatest larval biomass was obtained from diet 2 in 500-ml (large)
cups (3.5 g) while larvae fed diet 17 had the least biomass (1.39 g) (Table 3). The great
est larval survival was on diet 14 (92.5%) (Table 3). Adults reared as larvae on diets
2 and 3 (large cups) had the greatest biomass (0.94 and 0.92 g, respectively) while the
greatest adult emergence was from larvae fed diet 12, small cups (55%) (Table 3). To
tal time from placement of larvae in sugarcane to adult emergence ranged from 33.8
days (diet 12, 500-ml cup) to 46.6 days (diet 17) (Table 3). Orthogonal contrasts indi
cated that overall larval and adult mass was greater for individuals cultured in 500
ml versus 100-ml cups (P < 0.01). However, adult emergence was greater for larvae
cultured in 100-ml cups (P< 0.03).


Using the diet described by Rahalkar et al. (1978, 1985), and several variations
of this diet using readily available materials, we were able to achieve growth and
development of R. cruentatus larvae. Sugarcane stem was, however, necessary for
completion of the life cycle. Of diets tested in the preliminary experiments, seven were

June, 1995

Weissling & Giblin-Davis: Rhynchophorus cruentatus diets 231

- i C~'CC N C Cf Cl
N- N-~' CC~' Lf 'J CD CC Cf Cl f
C C CC -C C -

o o C C C C C

- - c c~c~cN ~ c c

on o on n on on o

c'j cflC'J Cfl Lf cJ cL CJ fI Lf N

Florida Entomologist 78(2)

chosen for further evaluation in replicated tests based on larval and adult biomass
and survival. Diets containing soybean oil (diets 4-8, and 10) resulted in high mortal
ity of R. cruentatus larvae. The published diet (Rahalkar et al. 1978, 1985) was suc
cessful for culture of R. cruentatus larvae but was eliminated because of difficulty in
obtaining coconut cake. Finely chopped field (S. palmetto, diet 1) and laboratory (pine
apple, diet 11) hosts were unacceptable for growth, development, and survival of R.
cruentatus larvae when brewers yeast, sucrose, molasses and salts were omitted (Ta
ble 2). However, when brewers yeast was added to chopped pineapple (diet 16), the
diet was acceptable for larval growth, development, and survival (Table 2). This sug
gests that yeasts or some other undefined ingredients present in rotting host tissues
are an important component in R. cruentatus nutrition. The importance of microbes
is not surprising considering that many members of the Rhynchophorinae are associ
ated with fermenting and rotting host tissues (Vaurie 1971). Further modification of
the pineapple and brewers yeast diet (16) to include sucrose, molasses, oats, and salts
with a reduction in the amount of pineapple (diet 2) resulted in improved larval
performance. Supplemental cholesterol (diets 3, 4, 5, 6, 13, 14, 15, and 17) or the ad
edition of egg yolk did not improve larval growth.
Weevils cultured by the methods of Giblin-Davis et al. (1989) had a mean genera
tion time of 78-79 days, with 47-48 days from placement of larvae in sugarcane to
adult emergence. Mean time to adult emergence in this study ranged from 33.8 + 13.3
days (diet 12, large cup) to 46.6 2.9 days (diet 17, large cup). Adults (sexes combined)
emerging from field-collected cocoons average 1.14 g (Giblin-Davis et al. 1989), while
adults cultured by the pineapple/sugarcane method of Giblin-Davis et al. (1989) had
a mean biomass of 0.72 g (males) and 0.75 g (females). Weevils cultured by the pine
apple/sugarcane method of Giblin-Davis et al. (1989) for over two years had a mean
adult biomass of 0.89 g (n= 704). Biomass of R. cruentatus adults cultured on artificial
diets in this study (refined test) ranged from 0.59 g (diet 17, large cup) to 0.94 g (diet
2, large cup) (Table 3). These comparisons suggest that R. cruentatus cultured on ar
tificial diets in this study are of comparable biomass to weevils reared using the pine
apple/sugarcane method of Giblin-Davis et al. (1989), but adults are slightly smaller
than those collected from the field.
Sugarcane appears to contribute very little to the continued growth of mature R.
cruentatus larvae (Giblin-Davis et al. 1989). Using data from two years of culture on
pineapple and sugarcane, we determined that the biomass of sugarcane pieces used
for pupation is not correlated to biomass of adults (unpublished data). However, larval
biomass is positively correlated with adult biomass (unpublished data). The biological
significance of adult biomass and its importance in the fitness of R. cruentatus has yet
to be explored. Weevils fed artificial diets in large cups had significantly greater larval
and adult biomass than weevils reared in small cups. However, larvae were in culture
for two weeks longer in the large cups. Thus, if larger larvae or adults are required,
larvae should be cultured for longer periods of time with an increased amount of diet.
The best diet tested for culture of R. cruentatus was diet 2 (supplemented pineapple
and oats), in both sizes of diet cups (Table 3). The simplest diet tested that provided
larval growth was diet 16, a mixture of canned pineapple and brewers yeast. Further
testing indicated that by using a combination of 300 g of diet 3 in 500-ml cups and sug
arcane, we were able to culture R. palmarum larvae to the adult stage (n = 5, 6.8 g lar
val mass, 80% larval survival, 2.5 g adult mass, 40% adult emergence) (unpublished
In conclusion, we were able to successfully culture R. cruentatus larvae by using
readily available ingredients. We suspect that sugarcane is nothing more than a
source of fiber from which the cocoon is constructed by R. cruentatus larvae. A limiting

June, 1995

Weissling & Giblin-Davis: Rhynchophorus cruentatus diets 233

factor in the culture of R. cruentatus to adults in many geographic locations may be
the availability of sugarcane. We have tested alternative fiber sources added directly
to 500-ml diet cups after larvae were in culture for 5 weeks but larvae failed to con
struct cocoons. Pupal induction in R. cruentatus needs further study but appears to
require warm temperatures, low relative humidity, and a high fiber substrate.
The culture of R. cruentatus on artificial diets has several interesting implications.
The use of agar-based diets, although somewhat odorous, has greatly improved work
ing conditions in the rearing facility. In addition, larvae can be reared individually
and their history can be documented. This will be important in future studies to de
termine the vector relationship of R. cruentatus with the red ring nematode, B. coco
philus. Larvae of R. cruentatus are large and would be well suited for use in
physiological studies. For example, using the pupae of R. palmarum and Sitophilus
oryzae (L.), Rahbe et al. (1990) identified a new class of hemolymph storage proteins.
Finally, although we are aware of no human consumption of R. cruentatus larvae in
the U.S., larvae of R. palmarum (Woodruff 1967), R. phoenicis (F.), and R. ferrugineus
(DeFoliart 1990) are considered delicacies by some. The culture of R. cruentatus on ar
tificial diets could be a potential advancement in developing a niche for consumption
of our indigenous species by palm weevil gourmets or feeding burrowing owls in cap
tivity (Yosef & Deyrup 1994).


We thank J. Cangiamila and B. J. Center for technical assistance, N. M. Mendoza
(CATIE, Turrialba, Costa Rica) for rearing R. palmarum, and D. Hall (U.S. Sugar
Corp., Clewiston, FL) for providing sugarcane bagasse. We are also grateful to F W.
Howard and G. Wheeler of the University of Florida, Fort Lauderdale Research and
Education Center, for critically reviewing an earlier version of this manuscript, and
D. Horton (USDA-ARS, Yakima, WA) for statistical advice. This research was sup
ported in part by a USDA Special Grant in Tropical and Subtropical Agriculture
CRSR-90-34135-5233 to R. M. G.-D., R. H. Scheffrahn, and J. P. Toth. This manu
script is Florida Agricultural Research Station Journal Series R-03901.


BERGER, E. W. 1907. The palmetto weevil (Rhynchophorus cruentatus, Fab.). Florida
Agr. Exp. Sta. Rept. 1907: XXXIX.
CHITTENDEN, F H. 1902. The palm and palmetto weevils. USDA Entomol. Bull. 38:
DEFOLIART, G. 1990. Hypothesizing about palm weevil and palm rhinoceros beetle
larvae as traditional cuisine, tropical waste recycling, and pest and disease
control on coconut and other palms-can they be integrated?, pp. 1-6 in G. De
Foliart [ed.], The Food Insects Newsletter, Vol. III (2).
GIBLIN-DAVIS, R. M., AND F W. HOWARD. 1988. Notes on the palmetto weevil, Rhyn
chophorus cruentatus (Coleoptera: Curculionidae). Proc. Florida State Hort.
Soc. 101: 101-107.
GIBLIN-DAVIS, R. M., AND F W. HOWARD. 1989. Vulnerability of stressed palms to at
tack by Rhynchophorus cruentatus (Coleoptera: Curculionidae) and insecti
cidal control of the pest. J. Econ. Entomol. 82: 1185-1190.
GIBLIN DAVIS, R. M., K. GERBER, AND R. GRIFFITH. 1989. Laboratory rearing of Rhyn
chophorus cruentatus and R. palmarum (Coleoptera: Curculionidae). Florida
Entomol. 72: 480-488.
1994. Field response of Rhynchophorus cruentatus (Coleoptera: Curculionidae)

Florida Entomologist 78(2)

to its aggregation pheromone and fermenting plant volatiles. Florida Entomol.
77: 164-177.
RAHALKAR, G. W., M. R. HARWALKAR, AND H. D. RANANAVARE. 1972. Development of
red palm weevil, Rhynchophorus ferrugineus Oliv. Indian J. Entomol. 34:
RAHALKAR, G. W., A. J. TAMHANKAR, AND K. SHANTHRAM. 1978. An artificial diet for
rearing red palm weevil, Rhynchophorus ferrugineus Oliv. J. Plantation Crops
6: 61-64.
SHANTHRAM. 1985. Rhynchophorus ferrugineus, pp. 279-286 in P. Singh & R. F
Moore [eds.] Handbook of insect rearing. Elsevier, New York, NY. v. 1.
RAHBE, Y., B. DELOBEL, J. GUILLAUD, AND C. NARDON. 1990. Storage proteins in Co
leoptera: a new class of tyrosine-rich proteins from the pupae of two weevils, Si
tophilus oryzae and Rhynchophorus palmarum (Coleoptera: Curculionidae).
Insect Biochem. 20: 331-341.
Method for the laboratory rearing of red palm weevil, Rhynchophorus ferrug
ineus Oliv. J. Plantation Crops 3: 65-67.
SANCHEZ, P. A., K. JAFFE, J. V. HERNANDEZ, AND H. CERDA. 1993. Biologia y compor
tamiento del picudo del coctero Rhynchophorus palmarum L. (Coleoptera: Cur
culionidae). Bol. Entomol. Venezolana N.S. 8: 83-93.
SAS INSTITUTE. 1985. SAS user's guide: statistics, fifth edition. SAS Institute. Cary,
VAURIE, P. 1971. Review of Scyphophorus (Curculionidae: Rhynchophorinae). Co
leopt. Bull. 25: 18.
WATTANAPONGSIRI, A. 1966. A revision of the genera Rhynchophorus and Dynamis
(Coleoptera: Curculionidae). Department of Agr. Science Bulletin, Bangkok 1:
Trap for capturing and retaining Rhynchophorus cruentatus (Coleoptera: Cur
culionidae) adults using Sabalpalmetto as bait. Florida Entomol. 75: 212-221.
WEISSLING, T. J., R. M. GIBLIN-DAVIS, AND R. H. SCHEFFRAHN. 1993. Laboratory and
field evidence for male-produced aggregation pheromone in Rhynchophorus
cruentatus (F.) (Coleoptera: Curculionidae). J. Chem. Ecol. 19: 1195-1203.
JR., AND A. C. OEHLSCHLAGER. 1994. Aggregation pheromone of palmetto wee
vil, Rhynchophorus cruentatus (F) (Coleoptera: Curculionidae). J. Chem. Ecol.
20: 505-515.
WEISSLING, T. J., AND R. M. GIBLIN-DAVIS. 1994. Fecundity and fertility of Rhyn
chophorus cruentatus (Coleoptera: Curculionidae). Florida Entomol. 77:
WILSON, M. E. 1963. Investigations into the development of the palm weevil Rhyn
chophoruspalmarum (L.). Tropical Agr., Trinidad. 40: 185-196.
WOODRUFF, R. E. 1967. A giant palm weevil, Rhynchophorus cruentatus (Fab.), in
Florida (Coleoptera: Curculionidae). Florida Dept. Agr. Div. Plant Industry En
tomol. Circ. No. 63.
YOSEF, R., AND M. DEYRUP. 1994. Pellet analysis of burrowing owls in south central
Florida. Florida Field Nat. 22: 78-80.

June, 1995

Thomas & Mangan: Morbidity in Fruit Flies


USDA-ARS Subtropical Agriculture Research Laboratory
Weslaco, TX 78596


Heat treatments are used to disinfest fruit exported from regions where Anas
trepha fruit flies are indigenous. Larvae that survive the heat treatments typically
form misshapen puparia. The assumption that all of these puparia will die before eclo
sion of the adult is shown to be incorrect. Two types of malformed puparia are typi
cally induced by hot-water immersion. A larviform puparium is seldom viable with an
eclosion rate of <1% in both the Mexican and West Indian fruit flies. However, a bot
tle-nosed puparium will produce an adult about 50% of the time. It should be assumed
that if any larvae survive treatment to form puparia, some will give rise to adults.

Key Words: Anastrepha, morbidity, puparial malformations, heat treatments, quar
antine, Tephritidae.


Los tratamientos con agua caliente se utilizan en regions donde se encuentran
moscas del g6nero Anastrepha para desinfestar las frutas que van a ser exportadas.
Las larvas que sobreviven los tratamientos de calor produce pupas malformadas. Se
demostr6 que la suposici6n de que todas las pupas mueren antes de la eclosi6n del
adulto es incorrect. Dos tipos de puparios malformados se pueden inducir con el tra
tamiento de agua caliente: larviformes y con nariz en forma de botella. Los puparios
larviformes raramente son viables, con una tasa de eclosi6n menor del 1% tanto en la
mosca mexicana como en la mosca de las frutas de las Indias Occidentales. Sin em
bargo, las pupas con nariz en forma de botella produce adults en aproximadamente
el 50% de los casos. Debe asumirse que si algunas larvas sobreviven al tratamiento y
forman puparios tambien darn lugar a adults.

The Mexican fruit fly, Anastrepha ludens (Loew), and the West Indian fruit fly,
Anastrepha obliqua (Macquart), are quarantined insect pests. Importation of host
fruits to the United States from countries where these flies are indigenous requires a
disinfestation treatment. For example, Mexican mangoes are given a hot-water im
mersion for 1 h at 46C, a treatment reported by Sharp (1988) to disinfest the fruit of
tephritid immatures. USDA-APHIS quarantine restrictions mandate a treatment
which will cause mortality of the infesting insects at the Probit 9 level, equivalent to
99.9968% mortality (Baker 1939; see also Chew 1994, Robertson et al. 1994, Shannon
1994, for recent discussion).
In reviewing the published studies on hot-water immersion treatments of various
combinations of host fruits and infesting insects, we noted that in many cases, mal
formed puparia were counted in the mortality figures (Sharp 1986; Sharp et al.

Florida Entomologist 78(2)

1989a,b,c; Sharp & Hallman 1992). We were concerned that scoring malformed pu
paria as dead insects without supporting data might include an undefined risk of dis
infestation failure.
We, therefore, undertook experiments emphasizing marginally sublethal temper
ature levels designed to produce large numbers of malformed puparia. The purpose of
these experiments was to identify the kinds of malformations associated with hot-wa
ter immersion treatments and to quantify the frequency of these malformations and
their specific mortality rates.


All experiments were conducted at the USDA Subtropical Agriculture Research
Laboratory in Weslaco, Texas. Ten-day-old, third-instar larvae of A. ludens and A.
obliqua were obtained from colonies of these insects maintained at this laboratory us
ing rearing procedures described by Rhode (1957) and Rhode & Spishakoff (1965) for
the Mexican fruit fly, and by Mangan & Ingle (1992) for the West Indian fruit fly. Un
der laboratory conditions at room temperature, 10-day-old larvae are sufficiently ma
ture to pupariate and metamorphose successfully to adults. The colonies of both
species were founded with material originating in Mexico, the A. obliqua colony dat
ing from 1987 and the A. ludens colony originating in 1953.
All larvae, except controls, were immersed in a circulating water bath controlled
to + 0.02 C by a computer driven temperature control system. The computer software
program "Water Troll" was developed by J. J. Gaffney, USDA-ARS, Gainesville, FL.
The water baths consisted of a series of four 11-liter capacity stainless steel water
banks; each equipped with an electric stirrer motor with shaft and propeller, a 1,000
Watt electrical resistance heater and a Hart 1006 precision thermometer. The larvae
were free within a 150 cc capacity, organdy mesh basket.
The experimental temperatures used were 38, 40, 41, 42, 43 and 44'C. The water
baths were brought to the selected temperature before immersion of the larvae, and
all immersions were for a duration of 1 h. The number of replicates performed varied
among temperatures and between species because of the difference in the numbers of
malformed puparia produced at the different temperatures and the necessity of using
those temperatures which produced the most malformations. Lower temperatures
which produced few malformations were replicated as a check against the possibility
that the malformations could have been caused by immersion in water alone. With A.
ludens there were eight replicates from two rearing batches at 38'C; five replicates
from five rearing batches at 40'C; nine replicates from six rearing batches at 41'C; 16
replicates from four rearing batches at 42'C; 23 replicates from eight rearing batches
at 43'C, and six replicates from three rearing batches at 44'C. With A. obliqua there
were 12 replicates from three rearing batches at 38'C, five replicates from five rearing
batches at 40'C; 59 replicates from 20 rearing batches at 41'C; 12 replicates from
three rearing batches at 42'C; 14 replicates from five rearing batches at 43'C, and
four replicates from one rearing batch at 44'C. The numbers of larvae varied among
replicates because of the necessity to avoid unnecessary handling and possibly detri
mental effects to the larvae prior to pupariation. Thus, batches of larvae were tested
as they became available and the precise numbers tested were counted after treat
ment during the immobile puparial stage. The mean treatment size was 289.6 larvae
per replicate with a total of 61 replicates for A. ludens, and 154.7 larvae per replicate
with a total of 106 replicates for A. obliqua.
Immediately following the hot-water immersion, both treated and untreated con
trol larvae were transferred to 10 cm wide, 250 cc capacity, plastic containers with

June, 1995

Thomas & Mangan: Morbidity in Fruit Flies

screened lids. The containers were half filled with clean, slightly moistened vermicu
lite (the pupariation medium) and held in an incubator at 25 + 2'C temperature, 75
+ 5% humidity and 12:12 L:D cycle. After 3-4 days the puparia were sifted from the
pupariation medium and scored as normal or malformed and separated. Dead larvae
were counted and removed. The puparia were then returned to the slightly moistened
vermiculite, placed in the incubator and held for emergence. After adult eclosion, the
puparia were examined to determine the percent eclosion for each type of malforma
In order to determine the viability of the adults from the malformed puparia, some
of the emergent flies were held and tested for fertility. Adults (males and females)
from eleven replicates each of A. ludens and A. obliqua were held in screen cages at
25'C and ambient light cycle with a diet of sugar and hydrolyzed yeast until an age
of 10 days. The treated, but normal, puparia were segregated from the malformed pu
paria. Thus, males from malformed puparia mated with females from malformed pu
paria, and males from treated, but normal, puparia mated with the corresponding
females. Untreated controls from the same rearing batches were also tested.When the
flies were ten days old, surviving females (up to 10 if available) from each test group
were placed in cages with an artificial gel oviposition medium and left overnight. The
numbers of eggs oviposited were counted on the following day. These eggs were held
and re-examined three days later to determine the percent hatch.
Differences in mean mortality between treatment and control groups were tested
using the t-test for paired comparisons (Sokal & Rohlf 1973). The relationship be
tween temperature and mortality and between temperature and frequency of malfor
mation was computed with the product moment correlation coefficient (Sokal & Rohlf
1973). The mean eclosion rate from normal vs. malformed puparia was tested using
single classification analysis of variance (Model I ANOVA) (Sokal & Rohlf 1973). The
probability value Pfor each F t and r statistic was calculated with the software pro
gram Speakeasy (Speakeasy Computing 1987).


A one-h exposure to a water bath at 44'C resulted in 100% total mortality in four
replicates of A. obliqua and 99.8 0.57% mean (SE) mortality in six replicates of A.
ludens. A total of two adults were produced from 1,132 A. ludens larvae tested, and
none from the 624 A. obliqua larvae tested. However, at 43'C, mean mortality was
79.2 + 12.5% in 23 replicates of A. ludens (n = 6,105) (Table 1). A. obliqua was more
susceptible to these temperatures; at 43'C only one adult was produced from 2,619
treated larvae (14 replicates), a mean mortality rate of 99.96 + 0.15%. At 42'C, only 5
adults were produced from 2,480 treated larvae, a mean mortality rate of 99.8 + 0.35%
(12 replicates). At 41'C, however, there was appreciable survival to the adult stage,
with a reduction in mean mortality to 75.6 + 12.0% (58 replicates) in A. obliqua (Table
Significant mortality was produced at all water temperatures tested except the
lowest. At the lowest test temperature, 38'C, there was 69.6 + 8.4% mean survival in
12 replicates of A. obliqua larvae (n = 2,526), whereas 85.9% of the controls from all
tests produced adults (Table 1.) The difference in mean survival, defined as the per
centage of adults emerging from puparia, was statistically significant using a
pair-wise t-test (t= 3.98; df= 11; P= .001). But, for A. ludens, the difference in sur
vival at the low temperature was much less. When subjected to 38'C for 1 h, mean
survival was 84.5 13.6%, compared to a mean survival of 84.5 + 12.9% in controls.

Florida Entomologist 78(2)


Normal Malformed

Test Total Dead Total Adults Bottle Larvi Peanut Adults
Temp n Larvae Pupae Eclosed nose form Pupae Eclosed

C 6059 110 5917 5500 25 3 4 13
38 2362 106 2172 1913 82 2 0 44
40 1154 68 1049 863 31 2 4 25
41 1911 258 1488 1085 164 0 1 72
42 5017 530 3834 3576 576 72 5 447
43 6105 2668 2281 1090 924 228 4 197
44 1132 1104 1 0 4 23 0 2

The difference in survival was not significant with a pair-wise t-test (t= 0.18; df= 7;
P 0.43).

Larval Mortality

Most of the mortality produced by the hot-water treatment was to the larval stage:
57% of total deaths in A. ludens and 55% of total deaths in A. obliqua. Larvae im
mersed in hot water stretched to their full length and became immobile. Surviving
larvae recovered mobility within 1-2 h and most pupariated eventually. Those which
did not recover mobility turned black and shriveled within 1-2 days.
At the lowest test temperature, 38'C, there was a mean mortality of 11.4 1.9%
in A. obliqua larvae and 4.2 + 2.9% in A. ludens larvae. However, even this low level
of mortality was significantly greater than for the untreated controls in which failure
to pupariate was 3.5 + 2.3%. Using a pair-wise t-test to compare larval mortality at


Normal Malformed

Test Total Dead Total Adults Bottle Larvi Peanut Adults
Temp n Larvae Pupae Eclosed nose form Pupae Eclosed

C 3455 73 3291 2945 73 15 6 24
38 2526 285 2078 1704 152 7 4 51
40 2454 176 2058 1354 179 13 4 81
41 7831 1846 2193 1001 2898 891 3 765
42 2480 1935 225 3 204 116 0 2
43 2619 2590 1 1 14 14 0 0
44 624 624 0 0 0 0 0 0

June, 1995

Thomas & Mangan: Morbidity in Fruit Flies



Z 60-

L 40


C 38 40 41 42 43 44
Fig. 1. Mortality in late third-instar larvae of Anastrepha ludens and A. obliqua
following one h exposure in hot-water bath at 38-44'C, and non-exposed controls (c).

38C, the t-value for A. ludens was 7.19 (df= 7, P= 8.5S"); and for A. obliqua t= 7.41
(df= 11, P= 6.73e6). Larval mortality rate increased sharply at 1 h exposures to 42C
for A. obliqua and at 43'C for A. ludens (Fig. 1).

Pupal Mortality

There was a 100% failure of larvae to pupariate at only one test temperature,
44'C, and for only A. obliqua. Some adults enclosed in all cases in which at least some
larvae pupariated. Pupal mortality was significantly correlated with temperature in
both species. The correlation coefficient (r) between temperature and mortality in A.
obliqua was 0.94 (P .0026). Mortality rose sharply at 41'C to 69.5 + 12.2% and at
test temperatures of 42'C, mean mortality was 99.2 + 1.5% in this species. For A.
ludens the correlation between temperature and pupal mortality was not as rigid but
still significant (r= 0.785, P= .032). The test temperatures did not induce high pupal
mortality in A. ludens until 43'C and above (Fig. 2).

Puparial Malformations

Pupariation in cyclorrhaphous Diptera is a four-step process (Zdarek & Frankel
1972, 1987). First, the anterior segments invert. Second, flexion of the integumental
musculature constricts the body into a barrel-shape. Third, the cuticle shrinks
(mainly by dehydration), and fourth, the cuticle becomes sclerotized through mela

Florida Entomologist 78 (2)



Z 60-

W 40



C 38 40 41 42 43 44

Fig. 2. Mortality in the pupal stages of Anastrepha ludens and A. obliqua from lar
vae subjected to hot-water immersion, and non-treated control larvae (c).

nization. Puparial malformations can be explained as dysfunctions of one or more of
these mechanisms (Thomas & Mangan 1990).
Three types of malformations were observed in the test groups of puparia. The rar
est form, a peanut-shaped puparium, was found in both treated and untreated
groups. This malformation was characterized by a mild constriction in the middle seg
ments. Its immediate cause was not determined but did not appear to be induced by,
or associated with, hot-water immersion. A total of 17 cases were seen in A. obliqua
(6 among the controls); and 18 cases in A. ludens (4 among controls). From the 35
peanut-shaped puparia, 15 adults emerged. These numbers were too small to treat
Larviform (Fig. 3a) and bottlenose (Fig. 3b) puparia were found to be associated
with hot-water immersed larvae in both species. The larviform malformation was
characterized by a failure to constrict into a typical barrel shape, and a failure of the
head segments to invert. Of the 1,056 larviform puparia of A. obliqua, none produced
an adult (100% mortality). However, of the 330 larviform puparia induced in A.
ludens, two adults closed (99.1% total mortality). Both were from the same treat
ment cohort subjected to 42 C. There was a significant correlation between tempera
ture and the percent frequency of larviform puparia in A. obliqua (r 0.88, P .0245).
For A. ludens, the correlation was not significant at the 95% confidence level (r= 0.64,
P .086). The lower mathematical correlation seemed to result from the abrupt in
crease in this type of malformation at the highest treatment temperature (Fig. 4), as
opposed to the linear increase in frequency seen in A. obliqua (Fig. 5).

June, 1995

Thomas & Mangan: Morbidity in Fruit Flies

Wiftt '

Fig. 3. Puparial malformations in Anastrepha ludens and A. obliqua: subjected to
hot-water immersion. (a) larviform puparia; (b) bottlenose puparia.
In the bottlenose malformation, the anterior-most segment of the puparium is ab
normally constricted, but otherwise morphologically asymptomatic (Fig. 3b). The fre
quency of the bottlenose malformation was significantly correlated with temperature
in both species: r= 0.77 (P .037) for A. ludens; r= 0.84 (P .038) for A. obliqua. The
greatest numbers of bottlenose puparia were produced at 40-41'C (A. obliqua) and


Florida Entomologist 78(2)

100 96.4.


z 60-

a_ 40 33.3

20- 14.4

0.5 3.7 3.0
0 .5

C 38 40 41 42 43 44
Fig. 4. Frequency of larviform or bottlenosed puparia following exposure of late
third-instar A. ludens larvae to one-h hot-water bath at 38-44'C, and non-exposed
controls (c).

41-42C (A. ludens). The survival of the pupal stage to eclosion as an adult was sig
nificantly less in the bottlenose puparia compared to the normal puparia at each test
temperature for both species. With A. obliqua, eclosion rate at 40C was 75.9 + 14.5%
for the normal puparia, but only 51.9 15.1% for the bottle-nose puparia. The differ
ence was significant tested by ANOVA (F= 9.19; df= 1, 12; P .01). Eclosion rate at
41 C was 37.4 + 13.5% for the normal puparia, but only 25.4 8.9% for the bottlenose
puparia, a significant difference (F= 7.67; df= 1, 26; P= .01). With A. ludens, adult
eclosion rate at 41C was 73.4 + 14.2% for the normal puparia, but only 50.5 20.4%
for the bottlenose puparia (F 7.59; df 1, 16; P .01). At 42C the adult eclosion rate
from the normal puparia was 92.7 + 5.3%. In comparison, the eclosion rate was lower,
78.7 + 12.7% for the bottlenose puparia, but the difference was not significant at the
95% confidence limit with ANOVA (F= 4.21; df= 1, 6; P .09).


Morbidity is the proportion of sick individuals in a population (Lapedes 1976).
Hot-water immersion of late third-instar larvae at sublethal temperatures resulted in
significant numbers of individuals which turned black and failed to pupariate or
which formed misshapen puparia in both A. ludens (Fig. 6) and A. obliqua (Fig. 7).
Adults failed to eclose from most of the malformed puparia. However, the results of
this study strongly suggested that morbidity can be equated with mortality only if the

June, 1995




0 40



Thomas & Mangan: Morbidity in Fruit Flies 24

- 96.6-

C 38 40 41 42 43 44
C 38 40 41 42 43 44

Fig. 5. Frequency of larviform or bottlenosed puparia following exposure of late
third-instar A. obliqua larvae to one-h hot-water bath at 38-44'C, and non-exposed
controls (c).
puparia are scored for the class of malformation. The larviform type of malformation
was lethal at a rate in excess of 99%. Conversely, a high percentage of the bottlenose
malformations are viable. In these experimental treatments, it was found that some
of the larvae that survived to pupariate eventually reached the adult stage, even un
der conditions in which the mortality of the larval stages was greater than 99% and
all of the puparia were malformed.
Followup tests with the adults that survived the hot-water treatments demon
strated that they were capable of attaining reproductive age, mating and ovipositing
at normal levels. This was true for adults from treated but normal puparia, as well as
for those emerging from malformed puparia. Mature A. ludens females (n = 66) from
normal puparia oviposited a mean of 26.6 + 25.0 eggs, of which 75.9% hatched. Fe
males (n = 69) from bottlenosed puparia oviposited a mean of 14.9 14.5 eggs, of
which 65.3% hatched. Control females (n = 35) laid a mean of 13.8 10.8 eggs, of
which 77.5% hatched. Mature A. obliqua females from normal puparia oviposited a
mean of 13.5 + 10.8 eggs (n = 66), of which 61.2% hatched. Females (n = 65) from bot
tlenosed puparia oviposited a mean of 8.0 6.9 eggs, of which 64.5% hatched. Control
females (n = 67) oviposited a mean of 12.3 + 11.5 eggs, of which 65.0% hatched.


Miguel Diaz Jr. was responsible for the operation of the computer driven circulat
ing hot water bath. Jose Galvan and Reyes Garcia were responsible for maintaining

Florida Entomologist 78(2)



z 60
a. 40

22.1 23.6
2 .3
C 38 40 41 42 43 44
Fig. 6. Morbidity: larval death or puparial malformation following exposure of late
third-instar A. ludens larvae to one-h hot-water bath at 38-44'C, and non-exposed
controls (c).

the test insects and for data collection. Guy Hallman and Felix Guerrero provided
valuable reviews of the manuscript.


BAKER, A. C. 1939. The basis for treatment of products where fruitflies are involved as
a condition for entry into the United States. U.S. Dept. of Agriculture Circular
No. 551, Washington D.C.
CHEW, V. 1994. Statistical methods for quarantine treatment data analysis, pp. 33-46
in J. L. Sharp and G. J. Hallman [eds.]. Quarantine Treatments for Pests of
Food Plants. Westview Press, Boulder, CO.
LAPEDES, D. N. 1976. Dictionary of Scientific and Technical Terms. McGraw-Hill, New
York, NY.
MANGAN, R. L., AND S. J. INGLE. 1992. Forced hot-air treatment for grapefruit in
fested with Mexican Fruit Fly (Diptera: Tephritidae). J. Econ. Entomol. 85:
RHODE, R. H. 1957. A diet for Mexican fruit flies. J. Econ. Entomol. 50:215.
RHODE, R. H., AND L. M. SPISHAKOFF. 1965. T6cnicas usadas en el cultiv de Anastre
pha ludens (Loew). II Memoria del dia del Parasitdlogo. Chapingo, Mexico. Esc.
Nac. Agric. Pp. 23-28.
tistical analyses to estimate efficacy of disinfestation treatments, pp. 47-65 in

June, 1995

Thomas & Mangan: Morbidity in Fruit Flies


100- DEAD LARVAE 90.9

80- 72.0


20177 15.5

C 38 40 41 42 43 44

Fig. 7. Morbidity: larval death or puparial malformation following exposure of late
third-instar A. obliqua larvae to one-h hot-water bath at 38-44 C, and non-exposed
controls (c).

J. L. Sharp and G. J. Hallman [eds.] Quarantine Treatments for Pests of Food
Plants. Westview Press, Boulder CO.
SHANNON, M. J. 1994. APHIS, pp. 1-10 inJ. L. Sharp and G. J. Hallman [eds.] Quar
antine Treatments for Pests of Food Plants. Westview Press, Boulder, CO.
SHARP, J. L. 1986. Hot-water treatment for control of Anastrepha suspense (Diptera:
Tephritidae) in mangos. J. Econ. Entomol. 79: 706-708.
SHARP, J. L. 1988. Status of hot water immersion quarantine treatment for Tephriti
dae immatures in mangos. Proc. Florida State Horticult. Soc. 101:195-197.
SHARP, J. L., AND G. HALLMAN. 1992. Hot-air quarantine treatment for carambolas in
fested with Caribbean fruit fly (Diptera:Tephritidae). J. Econ. Entomol. 85:
1989a. Immersion of Florida mangos in hot water as a quarantine treatment
for Caribbean fruit fly (Diptera: Tephritidae). J. Econ. Entomol. 82: 186-188.
SHARP, J. L., M. T. OUYE, S. J. INGLE, AND W. G. HART. 1989b. Hot-water quarantine
treatment for mangoes from Mexico infested with Mexican fruit fly and West
Indian fruit fly (Diptera: Tephritidae). J. Econ. Entomol. 82: 1657-1662.
Hot-water quarantine treatment for mangoes from the state of Chiapas, Mex
ico, infested with Mediterranean fruit fly and Anastrepha serpentina (Wiede
mann) (Diptera: Tephritidae). J. Econ. Entomol. 82: 1663-1666.
SOKAL, R. R., AND F. J. ROHLF. 1973. Introduction to Biostatistics. W. H. Freeman,
San Francisco, CA.

246 Florida Entomologist 78(2) June, 1995

THOMAS, D. B., AND R. L. MANGAN. 1990. Puparial malformations in the screwworm
with special reference to the bola-boliche syndrome. Southwest. Entomol.
ZDAREK, J., AND G. FRAENKEL. 1972. The mechanism of puparium formation in flies.
J. Exp. Zool. 179: 315-323.
ZDAREK, J., AND G. FRAENKEL. 1987. Pupariation in flies: a tool for monitoring effects
of drugs, venoms and other neurotoxic compounds. Arch. Insect Biochem. Phys
iol. 4: 29-46.


Florida Entomologist 78(2)


Systematic Entomology Laboratory
Universidad del Valle de Guatemala
Aptdo. 82
Guatemala City, GUATEMALA

Petrejoides caralae New Species is described from wet lowland forest of eastern
Guatemala (450-950 m altitude). The aedeagus of Petrejoides michoacanae Schuster
is described.
Key Words: Aedeagus, Mexico.

Se describe Petrejoides caralae Nueva Especie de bosque humedo de tierras ba
jas, del este de Guatemala, entire 450-950 metros de altitude. Se describe el edeago de
Petrejoides michoacanae Schuster.

Only two of the 18 previously described species of Petrejoides are known from be
low 1000 m altitude (Schuster 1991). Here we describe a new species from a wet, low
land forest of two isolated mountains of eastern Guatemala (Fig. 1). We also comment
on the holotype of Petrejoides michoacanae Schuster and describe the aedeagus.

Petrejoides caralae Cano & Schuster NEW SPECIES
Figures 2, 3

Head: anterior border of labrum concave, anterior angles rounded. Clypeus in
clined, short (anterior-posterior), rectangular, anterior border linear slightly arcuate,
anterior angles rounded and directed downward, smooth and brilliant. Frontoclypeal
suture slightly arcuate. External tubercles distinct, rounded.
Frontal area short, without inner tubercles. Frontal ridges poorly marked. Frontal
fossae glabrous or with 1-2 setae. Median frontal structure of "falsus" type (see Reyes

June, 1995

Cano & Schuster: Mesoamerican Petrejoides



Fig. 1. Distribution of Petrejoides caralae New Species. Square = Sierra de Caral,
circle = Cerro San Gil. Elevations: 0-500 m = irregular dots, 500-1500 m = regular
dots, 1500-2500 m = horizontal lines, >2500 m = vertical lines.

Castillo 1970), center horn short with apex not free, without median longitudinal
groove posteriorly; lateral ridges curved slightly forward, with terminal tubercles. Oc
capital groove well marked, terminating in supraorbital ridges.
Anterior 1/2 of supraorbital ridge bituberculate, posterior 1/2 not bifurcate. Ante
rior cephalic angle rounded. Canthus not swollen distally, apex rounded, not, or
barely, reaching lateral eye margin. Dorsal width of an eye 1/8 head width.
Ligula between insertions of labial palps wide, convex or slightly convex. Lateral
lobes of mentum with anterior external border rounded, whole surface punctate and
pubescent, lateral border straight. Medial basal mentum bare without punctures, an
terior border slightly biconvex. Hypostomal process narrow without lateral depres
sion. Infraocular ridge indistinct, punctate and pubescent.

Florida Entomologist 78(2)

Mandible with only 2 apical teeth. Mandibular tooth occupies more than 1/2 man
dible length. Internal teeth bifid.
Thorax: Lateral fossa of pronotum without punctations, 0-2 other punctations on
lateral pronotum outside fossa. Arcuate (3 specimens) or linear (1 specimen) scars
dorsally on posterior 1/3 of pronotum, 1 on each side. Pronotum with marginal groove
very narrow; anterior angles rounded. Prosternum rhomboidal with posterior apex
Mesosternum mostly brilliant throughout, narrow rugose line bordering mesepis
ternum, without or with a few punctations, with a few hairs. Mesosternal groove
lightly present. Mesepisternum not rugose, brilliant. Mesepimeron hairy.
Metasternum anterior angles pubescent. Lateroposterior sides of disk delimited
by 11-38 punctations on each side; marginal fossa wide, very pubescent.
Anterior elytral profile convex; elytral striations marked uniformly with small,
round, light punctations, somewhat heavier in lateral striations; junction of striations
1 and 10 with many extra punctations forming a double line, with many fine hairs.
Wings: as in Petrejoides recticornis (Burmeister), not reduced (see Fig. 5 in Castillo
& Reyes-Castillo (1984)).
Legs: Femur I with anterior-ventral groove almost absent or slightly marked.
Tibia II with long dorsal ridge.
Abdomen: Marginal groove complete around last sternite.
Aedeagus: Median lobe large and globose, with many pigment punctations. Ven
tral view: parameres and basal piece not completely separated by a distinct suture,
the tegmen is narrower at the base than the apex. Dorsal view: Basal piece small, all
the way round.
Dimensions (mm). Total length, mandibles to tip of elytra 33-34.2, x = 33.5, males
33-34.2, x = 33.5, female 33.6; elytral length 18.4-18.7, x = 18.6; pronotal length 8.1
8.6, x = 8.4; pronotal width 10.8-11.0, x = 10.9; humeral width 9.6-10.2, x = 9.9; head
width 7.4-7.7, x = 7.5; aedeagal length 3.6.

Material Examined
Four whole specimens, including one general female, two almost black males and
one black old male.

Type Material
Holotype male, Guatemala, Izabal Dept., Morales, Sierra de Caral, altitude 450
m., 28-X-1992, J. Monz6n collector.
Paratypes: Two from same location and collection data as holotype; one from Iza
bal Dept., Puerto Barrios, Cerro San Gil, 950 m., VI-1993, J. Monz6n.
Types at present are in the Arthropod Collection of the Universidad del Valle de
Guatemala. We plan to deposit paratypes in the collection of the Instituto de Ecologia,
Xalapa, Mexico and in the Natural History Museum of Paris.

Refers to the mountain from which 3 of the 4 specimens are known, Sierra de
Caral, Guatemala.

Known from a wet, lowland forest of two mountains of eastern Guatemala, the Si
erra de Caral near the Guatemalan-Honduran border, and Cerro San Gil near Puerto
Barrios (Fig. 1).

June, 1995

Cano & Schuster: Mesoamerican Petrejoides


a b c

Figs. 24. 2 -Petrejoides caralae New Species dorsal view of head, 3 -Petrejoides
caralae New species: 3a, aedeagus, ventral view; 3b, aedeagus, lateral view; 3c,
aedeagus, dorsal view. 4 -Petrejoides michoacanae Schuster: 4a, aedeagus, ventral
view; 4b, aedeagus, lateral view; 4c, aedeagus, dorsal view.

At present, apparently no gene flow occurs between these two areas. Sierra de
Caral and Cerro San Gil are separated by the wide, lowland valley of the Motagua


Florida Entomologist 78(2)

river. This valley could have been easily crossed during glacial times when montane
forests were lower (MacVean & Schuster 1981).
Other species found at the collection sites in the Sierra de Caral include: Verres
corticicola (Truqui), Passalusjansoni (Bates) and Passalus punctiger Lepeletier and
Serville. On Cerro San Gil, they include: Proculus opacipennis (Thompson), Popilius
eclipticus (Truqui), Passalus caelatus Erichson, Odontotaenius striatopunctatus
(Perch.) and Paxillus leachi MacLeay.


P caralae is the largest species of Petrejoides. It seems most related to P. michoa
canae Schuster. These species are in the "recticornis" group of Castillo & Reyes
Castillo (1984). P caralae can be differentiated from other Petrejoides by the following
modification in the key of Schuster (1991):
8(7') Mandible with 2 apical teeth, pronotum with a pair of dorsal scars posteriorly;
frontal fossae glabrous or with 1-2 setae, metasternum with lateral fossae
wide, juncture of elytral striae 1-10 with many fine hairs ..................................
............. ........ ................................. ... Petrejoides caralae New species
8' Mandible with 3 apical teeth. Pronotum without dorsal scars posteriorly; fron
tal fossae pubescent; metasternum with lateral fossae narrow, junction of ely
tral estriae 1-10 glabrous .............................Petrejoides michoacanae Schuster

Petrejoides michoacanae Schuster

Schuster (1991) describes this species based on one specimen from Michoacan,
Mexico. On reviewing the holotype, we noted that the humeral width is 9.9 mm, not
6.3 mm, and it is male. The aedeagal description is as follows (Fig. 4):
Ventral view: Aedeagus asymmetrical, almost as long as wide, 2.68 mm long. Me
dian lobe almost as long as tegmen. Tegmen divided, narrower at the base than at the
apex; lateral lobes more short than basal piece. Dorsal view: Median lobe and basal
piece with the sclerotized area lightly visible.


We thank Jose Monz6n for collecting the specimens and Universidad del Valle de
Guatemala for support.


CASTILLO, C., AND P. REYES-CASTILLO. 1984. Biosistematica del g6nero Petrejoides
Kuwert (Coleoptera, Lamellicornia, Passalidae). Acta Zool. Mexicana 4: 184.
MACVEAN C., AND J. C. SCHUSTER. 1981. Altitudinal distribution of passalid beetles
(Coleoptera, Passalidae) and Pleistocene dispersal on the Volcanic Chain of
Northern Central America. Biotropica 13: 2938.
REYES-CASTILLO, P. 1970. Coleoptera, Passalidae: morfologia y division en grandes
grupos; g6neros americanos. Folia Entomol. Mexicana 20-22: 1240.
SCHUSTER, J. C. 1991. Petrejoides (Col: Passalidae): four new species from Mesoamer
ica and Mexico with a key to the genus. Florida Entomol. 74: 422-432.

June, 1995

Shelly & Bailey: Katydid Movement 251


Department of Zoology
University of Western Australia
Nedlands, Western Australia 6009


Movement in relation to local food levels was compared between the sexes of the
zaprochiline katydid Kawanaphila nartee Rentz. Adults of both sexes were marked,
and their departure rates were estimated (1) early in the season for two plots having
high vs. low food density and (2) throughout the season for a single plot with changing
food levels. Based on the between-plot comparison, movement of males was indepen
dent of food density, whereas females were less likely to leave the rich patch than the
poor one. The within-plot comparisons revealed that, despite large changes in food
levels, male departure rates did not vary over the season. Females showed increased
movement over the season and by the end of the study had a departure rate similar
to that of males.

Key Words: Spacing, mate searching, Australia.


Se compared el movimiento de los machos y hembras de Kawanaphila nartee Rentz
(Tettigonidae: Zaprochiliinae) en relacidn con niveles locales de alimento. Adultos de
ambos sexos fueron marcados y se calculd la frecuencia de salida (1) en dos parcelas
con densidad de alimento alta y baja al comienzo de la temporada y (2) en una sola
parcela con niveles de alimento variables a lo largo de la temporada. La comparacidn
entire las parcelas indic6 que el movimiento de los machos era independiente de la
densidad de alimento mientras que las hembras fueron menos propensas a salir de la
parcela con alta densidad. Las comparaciones dentro de cada parcela revelaron que,
a pesar de los grandes cambios en los niveles de alimento, la salida de los machos no
varied a lo largo de la temporada. Las hembras mostraron incremento de movimiento
a lo largo de la temporada y al final del studio tuvieron una tasa de salida similar a
la de los machos.

Males of many insect species encounter and mate with females at resources, such
as food or oviposition substrate, that are critical to female reproductive success
(Thornhill & Alcock 1983). In many of these species, males aggressively defend terri
stories, and their mating success is directly related to the quality of resources con
trolled (McLain 1986; Villalobos & Shelly 1991). However, in other species, males do
not hold territories and exhibit exploitative (or scramble) competition for mates at re

'Present address: Hawaiian Evolutionary Biology Program, University of Hawaii, Honolulu,
Hawaii, 96822, USA

Florida Entomologist 78(2)

source sites (Banks & Thompson 1985; Hafernik & Garrison 1986). Theoretically,
these males are expected to have an ideal free distribution (Fretwell 1972), i.e., to
space themselves in such a way that each individual has the same probability of en
countering females (Parker 1974).
In an earlier article (Shelly & Bailey 1992) on the Australian katydid
Kawanaphila narteeRentz, we reported a difference between the sexes in their distri
bution relative to the local abundance of the primary foodplant. Censuses of nine plots
revealed that female abundance increased with increasing foodplant density, whereas
male numbers varied independently of foodplant density. Correspondingly, fe
male:male ratios increased with increasing resource richness in the plots. Because
males are not territorial (Bailey & Simmons 1991), the resource-dependent increase
in female:male ratio suggested that males had more mating opportunities in rich food
patches than small ones, i.e., they were not distributed in an ideal free manner.
Focusing on the same katydid species, we here examine movement tendencies of
males and females in response to resource abundance. Specifically, we estimated de
parture rates of the sexes (1) simultaneously for two plots with high vs. low foodplant
density and (2) over the adult season for a single plot whose foodplant density showed
marked temporal variation.


Study Site and Species

Field work was conducted during August-November 1990 in Kings Park, a large
area of bushland in Perth, Western Australia. The park is largely a shrub-grassland
with an open canopy dominated by various species of Banksia and Eucalyptus trees.
The life history, feeding biology, and mating system of the species have been de
scribed elsewhere (Gwynne & Bailey 1988; Gwynne & Simmons 1990; Simmons &
Bailey 1990; Bailey & Simmons 1991; Rentz 1993), and the following summary de
rives from these earlier studies. The species is univoltine, with adults sexually active
between September-November. The adults, which are flightless, feed exclusively on
pollen and nectar. A variety of foodplants are utilized, but kangaroo paws (Anigozan
those manglesii Endl.), which flowers from August to October (see Results), is the main
food source during the early and middle portions of the adult season. As kangaroo
paws decline, the zaprochilines switch their feeding to a species of grasstree (Xanth
orrhoea preissei Endl.), which flowers from August to November. Grasstrees are far
less abundant than kangaroo paws and typically occur as isolated individuals sepa
rated by distances of 2-10 m. However, each grasstree produces a large stalk bearing
thousands of flowers and thus represents a rich, albeit localized, source of pollen and
Mating behaviour involves the attraction of females to stationary males calling
within the vegetation. Males sing throughout the night, but sexual activity is re
stricted to the 3-h period immediately following sunset (1900-2200 hours). Males are
not territorial and are not physically aggressive toward one another. During mating,
the male transfers a large, proteinaceous spermatophore to the female, which she
then consumes. Both field observations and experiments reveal a seasonal change in
sex roles dependent on food availability. Females are apparently food-limited during
the kangaroo paws season, and males are the choosy sex at this time. Conversely,
flowering grasstrees provide superabundant food resources, and females are then the
discriminating sex. Females appear to mate multiply, though field data on remating
frequency are inexact.

June, 1995

Shelly & Bailey: Katydid Movement

Zaprochiline Abundance and Movement

During the kangaroo paw season, movement of zaprochilines was studied in 2
plots, each measuring 10 m by 10 m. The main plot (plot A) had a high density of kan
garoo paws and was monitored during 3 different periods corresponding to the early
(1 22 September), middle (25 September 16 October) and late (20 October 10 Novem
ber) parts of the kangaroo paws season. The other plot (plot B) had a much lower den
sity of kangaroo paws and was monitored only during the early part of the kangaroo
paws season. Counts of flowering kangaroo paws were made approximately one week
after the initial zaprochiline census for a given monitoring period. Plots A and B were
separated by a distance of approximately 30 m and contained the same sets of plant
During a given monitoring period, we censused a plot every 3 days for a period of
three weeks. Censuses were performed between 1900-2200 hours, usually under clear
skies, but occasionally during light rain (zaprochilines were still active under such
conditions). During a census, we slowly and systematically moved through a plot and
attempted to locate all K. nartee adults. Calling males produce an ultrasonic signal
(Gwynne & Bailey 1988) and were located using a "mini bat detector" (QMC Instru
ments) set at the average dominant frequency of the call (50 kHz). Females and non
calling males were located visually by scanning vegetation with a head lamp.
During the first 3 censuses of a given monitoring period, we marked all individuals
with unique color sequences of enamel paint on their abdomen and released them at
their original perch. Marking had no apparent effects on the subsequent behaviour of
the insects, and individuals typically resumed calling or feeding within minutes of be
ing handled. In the remaining censuses, we simply noted the presence of marked in
dividuals and counted unmarked ones.
Data on the presence or absence of marked individuals were used to calculate em
igration rates for both sexes. The numbers of individuals remaining in the plot
through time were log transformed to produce decay curves analogous to survivorship
curves used in demographic studies. Pairwise slope comparisons followed the method
of Zar (1974). Tests significant at p=0.005 were also found to be significant using ad
justed critical values of the Bonferroni correction (Kleinbaum et al., 1988).
In examining the presence/absence data, we found that some individuals had a
gap" in their record such that they were not observed on a given census date(s) but
were observed on the census dates immediately preceding and following the gap. In
almost all cases, the gap consisted of only one census date, and we assumed the indi
vidual was either missed or was immediately adjacent to the plot and therefore
treated the individual as being present throughout (i.e., operationally, we ignored the
gap). However, where the gap consisted of 2 or more consecutive census days, we as
sumed the individual moved away from the plot and considered the first day of ab
sence to be its departure date.
As the preceding paragraph indicates, we assume that the disappearance of
marked individuals from our study plots resulted primarily from emigration and that
mortality was of secondary importance. This assumption appears valid, because (1)
individuals typically live for several weeks in the laboratory (Simmons & Gwynne
1993), yet the numbers of marked individuals decreased markedly over just a few
days (see Results) and (2) compared to the relatively straightforward explanation in
volving variable food abundance, unusual (and therefore unlikely) patterns in preda
tion intensity would be required to account for the spatial and sexual differences
observed in population decay curves (see Results). In addition, because of the rapid
disappearance of marked individuals, successive censuses in a plot were not simply
repeated estimates for the same local population of insects. Though census data from

Florida Entomologist 78(2)

different dates were not completely independent (since some individuals were
counted on two or more censuses), the high turnover of individuals lessened this sta
tistical problem, and hence inter and intra-plot comparisons of abundance were con
sidered valid.
During censuses in the early part of the kangaroo paws season, we also marked
the position of each individual in plots A and B, respectively, by placing a numbered
tag immediately adjacent to the perch. On the day following a census, we recorded the
coordinates of these tags (to the nearest 0.1 m) on gridded maps of the study plots.
These data were used to investigate spacing patterns of both sexes via analysis of
nearest neighbor distances (Clark & Evans 1954).


Early season censuses in plots A and B revealed a differential effect of kangaroo
paws density on the local abundances of male and female zaprochilines. Presumably
owing to increased food levels, female abundance in plot A during the early season
was approximately 50% higher than that recorded in plot B (Mann-Whitney test;
U=58; P < 0.01; Table 1). In contrast, male abundance was, on average, not signif
icantly different between plots A and B (Mann-Whitney test; U=22; P> 0.05; Table 1).
An intersexual difference in response to changing food levels was also apparent
when seasonal trends in abundance were examined for plot A (Table 1). Female abun
dance varied significantly among monitoring periods (Kruskal-Wallis test; H=15.3; P
< 0.001). Corresponding to an increase in flowering kangaroo paws, female numbers
increased significantly between the early and middle monitoring periods (nonpara
metric multiple comparisons test; Zar, 1974; q=4.6; P < 0.005), whereas male abun
dance was similar between these 2 periods (q=1.8; P> 0.05). The number of flowering
kangaroo paws declined dramatically by late season, and both male (q=4.2) and fe
male (q=5.6) abundance decreased significantly from their mid-season levels (non
parametric multiple comparisons test; Zar, 1974; P< 0.01 in both tests).
Differences in male and female emigration rates for plots A and B were consistent
with the inter-plot differences in abundance. In the early part of the kangaroo paws
season, females emigrated from plot B at a greater rate than from plot A (Fig. 1; Table
2). For example, 36% of the females marked in plot A remained there for at least 9
days compared to only 9% of the females in plot B. Average residency of females was
approximately 7 days for plot A but only 4 days for plot B. In contrast, no inter-plot
difference in departure rates was noted for males (Fig. 1; Table 2), and approximately
15% of the males in both plots remained for at least 9 days. Average male residency


Males Females Kangaroo Paws

Plot A
Early 27 (7) 25 (5) 146
Middle 25 (6) 31 (5) 307
Late 16 (7) 14 (5) 93
Plot B
Early 25 (5) 16 (4) 50

June, 1995

Shelly & Bailey: Katydid Movement 255

1. Pio e ry e0 PlolA-early

0 3 6 9 12 15 0 3 6 9 2 15

SPlot A-middle 100 Plot A -Me

S3 6 9 12 15 0 3 6 9 12 15

Days since marking

Fig. 1. Survivorship data for marked males (solid circles/lines) and females (open
circles/dashed lines) in the two study plots. Numbers of individuals (log scale) remain
ing in the plot are graphed against days since marking. Linear regressions: Males
(solid lines) -(a) Y=1.72-0.08X; r2=0.93 (b) Y=1.73-0.09X; r2=0.94 (c) Y=1.67-0.09X;
r2=0.98 (d) Y=1.54-0.09X; r2=0.95. Females (dashed lines) -(a) Y=1.530.10X; r2=0.98
(b) Y=1.63-0.055X; r2=0.97; (c) Y=1.660.07X; r2=0.97 (d) Y=1.45-0.08X; r=0.95.

was about 5 days in both plots A and B. As this result suggests, during the early sea
son males left plot A at a greater rate than did females (Fig. 1; Table 2).
The degree to which emigration rates varied over the season also differed between
the sexes. In plot A, female emigration rates showed a consistent increase through
time: rates increased 30% between the early and middle period and 20% between the
middle and late periods, though the latter difference was not statistically significant
(Fig. 1; Table 2). Correspondingly, average female residency in plot A declined from 7
days in the early season to approximately 4 days by late season. In contrast, male em
igration rates from plot A were similar between all monitoring periods (Fig. 1; Table
2). Owing to this difference, male and female emigration rates became more similar
through time, with middle season values differing marginally and late season values
not differing at all (Table 2).
Early season mapping of the positions of individual zaprochilines revealed that on
most dates both males and females were randomly distributed in both Plots A and B.
Males in plot A and females in plot B were randomly distributed on all 8 dates (P=.05
in these and all subsequent tests; t-test following Clark & Evans, 1954). Female dis
tribution in plot Awas random on 6 dates, aggregated on one date, and regular on one
date. Male distribution on plot B was aggregated on one date and random on all oth

Florida Entomologist 78(2)


Be' Ae Am Al

Be ns ns ns
Ae *** ** ns ns Males
Females Am *** ms ns
Al ns ns ns

'On the matrix borders, upper case letters (A,B) designate plot; lower case letters (e,m,l) designate early, mid
dle, and late monitoring periods, respectively Within the matrix, elements above the diagonal represent com-
parisons between male populations, and elements below the diagonal represent comparisons between female
populations. Elements on the diagonal represent comparisons between male and female populations, ns -not sig
nificant; ms -marginally significant (P< 0.1); P< 0.05; ** -P< 0.005; *** P< 0.001.


Consistent with our previous study (Shelly & Bailey 1992), male and female K.
nartee exhibited different spatial distributions relative to food resources: in the early
season comparison of the two plots, male abundance was similar between rich and
poor food patches, whereas females were more numerous in the rich patch. Previously,
we suggested that this difference might have reflected a sampling bias. Females were
perhaps most conspicuous in dense patches of kangaroo paws, where they were often
feeding on exposed flowers well above the ground. In contrast, males usually perched
close to the ground regardless of the vegetation, and therefore our ability to locate
males was, more or less, independent of the density of kangaroo paws. Although this
possibility cannot be dismissed completely, the intersexual difference in movement
described here is strong corroborative evidence that the intersexual difference in
spacing was real and was not simply a sampling artifact.
It appears unlikely that predation (rather than emigration) accounted for the de
cay curves observed for marked individuals. If, as seems likely, predators (e.g., spi
ders, mantids) displayed an aggregativee response" (Hassell et al. 1976) to the
zaprochilines (i.e., settled preferentially in areas of higher prey abundance), then we
might have expected decay curves to be steeper in plot A (where zaprochiline abun
dance was higher) than in plot B. Yet, the opposite trend was observed for females,
and decay curves for males were similar in the two areas. In addition, and indepen
dent of an aggregative response by predators, it remains difficult to explain why pre
dation would have a differential sex-bias in the 2 study plots, i.e., why (in the early
season) females would be more vulnerable to predation than males in plot B but less
vulnerable in plot A?
The "uncoupled" spatial distributions of males and females observed during the
kangaroo paws season contrasts dramatically with the situation described for the ka
tydid Orchelimum nigripes Scudder (Feaver 1983). In this species, males initially es
tablished territories in areas of high female eclosion. However, females later moved to
preferred oviposition sites, inducing males to relocate as well and establish new ter
ritories at the oviposition area. Moreover, males typically perched along the borders
of the oviposition areas, where they had a higher chance of intercepting females.
Based on these observations, Feaver (1983) concluded that males assessed territory
locations, not on the basis of the resources present (otherwise males would have set
tled within oviposition areas independently of females), but by directly monitoring fe
male abundance.

June, 1995

Shelly & Bailey: Katydid Movement

At present, it is not known what strategy male K nartee use to maximize encoun
ters with potential mates. Based on the present results, it appears unlikely that
males use food resources or conspecific calling as a cue to female abundance (see also
Bailey & Simmons 1991). It is possible that movement by females may have greatly
reduced spatial variation in male mating opportunities. Despite their lowered emi
gration from rich food patches, females still displayed a relatively high level of move
ment: even when flowering kangaroo paws were abundant, 40%-50% of the females
moved out of plot A within 3 days of being marked. Consequently, males in poor food
patches may have encountered females travelling to new feeding or oviposition sites.
If female traffic were sufficiently high over the entire habitat, the ability of males to
differentiate areas of high and low female abundance may have been limited, leading
to the even distribution of males over the habitat.
However, this explanation simply raises another unanswered question, i.e., why
do female K. nartee move so frequently? The observation that female emigration from
plot A increased between early and mid season (despite an increase in flowering kan
garoo paws) suggests an age effect, perhaps related to increased searching for ovipo
sition sites. Any tendency to disperse eggs, and thereby spread mortality risks, would
further contribute to frequent movement by females. In short, it appears that under
standing the movement and distributional patterns of females (and thus males) may
require information on the oviposition behavior, particularly the identification of pre
ferred egg-laying sites.


We thank P. R. Wycherley, Director of Kings Park and Botanic Garden, for his per
mission to conduct studies in the park. Lisa Masini provided capable field assistance,
and Emma Shelly and Ethel Villalobos assisted with the counts of flowering kangaroo
paws. Ethel Villalobos also supplied the resume. Comments by Leigh Simmons
greatly improved the manuscript. This research was supported by ARC grant


BAILEY, W. J., AND L. W. SIMMONS. 1991. Male-male behavior and sexual dimorphism
of the ear of a zaprochiline tettigoniid (Orthoptera: Tettigoniidae). J. Insect Be
hav 4: 51-65.
BANKS, M. J., AND D. J. THOMPSON. 1985. Lifetime mating success in the damselfly
Coenagrion puella. Anim. Behav. 33: 1175-1183.
CLARK, P. J., AND F. C. EVANS. 1954. Distance to nearest neighbor as a measure of
spatial relationships in populations. Ecology 35: 445-453.
FEAVER, M. N. 1983. Pair formation in the katydid Orchelimum nigripes (Orthoptera:
Tettigoniidae), pp 205-239 in D. T. Gwynne and G. K. Morris [eds.], Ortho
pteran mating systems: sexual competition in a diverse group of insects. West
view Press, Boulder, CO.
FRETWELL, S. D. 1972. Populations in a Seasonal Environment. Princeton University
Press, Princeton, NJ.
GWYNNE, D. T., AND W. J. BAILEY. 1988. Mating system, mate choice and ultrasonic
calling in a zaprochiline katydid (Orthoptera: Tettigoniidae). Behaviour 105:
GWYNNE, D. T., AND L. W. SIMMONS. 1990. Experimental reversal of courtship roles in
an insect. Nature 346: 172-174.
HAFERNIK, J. E., AND R. W. GARRISON. 1986. Mating success and survival rate in a
population of damselflies: results at variance with theory? American Nat. 128:

Florida Entomologist 78(2)

HASSELL, M. P., J. H. LAWTON, AND J. R. BEDDINGTON. 1976. The components of ar
thropod predation. I. The prey death-rate. J. Anim. Ecol. 45: 135-164.
KLEINBAUM, D. G., KUPPER, L. L., AND K. E. MULLER. 1988. Applied regression anal
ysis and other multivariate methods. PWS-Kent, Boston, MA.
McLAIN, D. K. 1986. Resource patchiness and variation in the intensity of sexual se
election in a resource defending polygynous insect. Oikos 47: 19-25.
PARKER, G. A. 1974. The reproductive behavior and the nature of sexual selection in
Scatophaga stercoraria L. IX. Spatial distribution of fertilization rates and evo
lution of male search strategy within the reproductive area. Evolution 28: 93
RENTZ, D. C. F. 1993. The Tettigoniidae of Australia. Volume 2. The Austrosaginae,
Zaprochilinae, and Phasmodinae. CSIRO, Melbourne.
SHELLY, T. E., AND W. J. BAILEY. 1992. Experimental manipulation of mate choice by
male katydids: the effect of female encounter rate. Behav. Ecol. Sociobiol. 30:
SIMMONS, L. W., AND W. J. BAILEY. 1990. Resource influenced sex roles of zaprochiline
tettigoniids (Orthoptera: Tettigoniidae). Evolution 44: 1853-1868.
SIMMONS, L. W., AND D. T. GWYNNE. 1993. Reproductive investment in bushcrickets:
the allocation of male and female nutrients to offspring. Proc. R. Soc. London
(B) 252: 15.
THORNHILL, R., AND J. ALCOCK. 1983. The evolution of insect mating systems. Har
vard University Press, Cambridge, MA.
VILLALOBOS, E. M., AND T. E. SHELLY. 1991. Correlates of male mating success in two
species of Anthidium bees (Hymenoptera: Megachilidae). Behav. Ecol. Socio
biol. 29: 4753.
ZAP, J. H. 1974. Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, NJ.


June, 1995

Florida Entomologist 78(2)


Everglades Research and Education Center,
University of Florida, IFAS,
Belle Glade, Florida 33430


Different staining procedures were evaluated for their time requirements and ef
fectiveness to differentially stain Liriomyza trifolii (Burgess)(Diptera: Agromyzidae)
eggs and wounds (i.e., stipples) made by the female flies ovipositing within cos lettuce
(Lactuca sativa L.) leaves. The best method for staining eggs within leaves was a lac
tophenol acid fuchsin solution that cleared the leaves of chlorophyll and stained the
eggs pink to dark red. Modification of this established procedure reduced the total
staining time to 7 min per leaf. Stipples were counterstained blue by immersing the
leaves in a lactophenol cotton blue solution for 1 min immediately following the egg
staining procedure. Larvae were not stained by either procedure. A new technique
was devised for highlighting stipples using solutions of silver nitrate, sodium carbon
ate, formaldehyde and citric acid. The procedure resulted in blackened stipples that

June, 1995

Nuessly et al.: Staining L. trifolii Eggs in Lettuce 259

were highly visible against the green leaves and did not kill developing larvae within
the leaves. Eggs could later be stained using the lactophenol acid fuchsin technique.
Sufficient contrast and color differences were produced by these procedures to suc
cessfully quantify stipples, eggs and larvae within stained and counterstained leaves
from digitized microscopic video images using computer image analysis software.

Key Words: Serpentine leafminer, Lactuca sativa, lactophenol, acid fuchsin, cotton
blue, silver nitrate.


Fueron evaluados various procedimientos para la tinci6n diferencial de huevos y he
ridas de Liriomyza trifolii (Burguess) producidos por la hembra en hojas de lechuga
de la variedad Cos (Lactuca sativa L.). Los criterios evaluados fueron el tiempo que
dur6 el procedimiento y la capacidad de tenir solamente los huevos y las heridas.
Hasta ahora el mejor m6todo para tenir los huevos dentro de las hojas es usar una so
luci6n de lactofenol y fuchsina acida para aclarar la clorofila y tenir los huevos de ro
sado a rojo intense. La modificaci6n de este process establecido redujo el tiempo total
de tenido a 7 min por hoja. Las punteaduras pueden ser contratenidas mediante la in
mersi6n de las hojas en una soluci6n de lactofenol y azul de algod6n por un minute in
mediatamente despues del process de tinci6n del huevo. Las larvas no fueron tenidas
mediante ningun process. Fue elaborada una nueva t6cnica para destacar las puntea
duras usando soluciones de nitrato de plata, carbonate de sodio, formaldehido y acido
citrico. El process di6 como resultado punteaduras negras muy visible contra el verde
de las hojas, y no mat6 las larvas en desarrollo dentro de las hojas. Los huevos pueden
luego ser tenidos usando la t6cnica del lactofenol con la fuchsina acida. Hubo sufi
ciente contrast y diferencias producidas por estos procedimientos para cuantificar
exitosamente las punteaduras, huevos y larvas dentro de las hojas tenidas y contra
tenidas a partir de imagenes de video usando software para analisis de imagenes de

Serpentine leafminers, Liriomyza trifolii (Burgess), are a major pest of lettuce
(Lactuca sativa L.) and other crops, particularly in Florida (Genung & Janes 1975,
Leibee 1981, Foster 1986). Damage to lettuce is in the form of stipples (i.e., scars from
feeding and oviposition punctures), leafmines, and pupae within the lettuce heads.
Hundreds of acres of lettuce are disked under annually before harvest, or left in the
field during harvest because of leafminers. Costs to the industry, including lost pro
duction costs and pesticide applications, probably exceed $1 million annually in south
Florida alone.
Host plant resistance is one method under examination to reduce leafminer pres
sure in lettuce. Field and laboratory studies indicated significant differences in stip
ple rates among commercial head and cos (i.e., romaine) lettuce cultivars (Nuessly &
Nagata 1993, 1994). However, oviposition rates (based on counts of 2-d-old larvae
within leaves) varied much less among romaine cultivars than was expected, based on
overall stipple counts (unpublished data). This suggested that variation among culti
vars mediated the feeding puncture to oviposition puncture ratio, or affected egg or
early instar mortality. In order to identify the source of the observed variation, it was
necessary to quantify both stipples and eggs.
Stipples on the distal areas of lettuce leaves can be observed using a 10X stereomi
croscope with either direct or transmitted light. However, callous tissue and exudates
that form over leaf punctures make it difficult to identify eggs, particularly in culti

Florida Entomologist 78(2)

vars with thick leaves or with strong wound responses. Eggs and stipples in the prox
imal area of lettuce leaves are difficult to quantify accurately even under microscopic
Many techniques are available for staining insects, nematodes, and fungal hyphae
within plant tissue. However, these techniques may require several hours to days to
complete the entire staining, destaining and counterstaining steps (e.g., Carlson &
Hibbs 1962, Gilstrap & Oatman 1976, Simonet & Pienkowski 1977). Also, extensive
variation in stippling and oviposition wounds within treatments on cos lettuce neces
states large sample sizes to determine meaningful differences among cultivars.
Leaves are similar in size and shape, and stipple rates vary significantly with their lo
cation on the leaves and plants (Nuessly & Nagata 1994), thus test leaves need to re
main separated to preserve their identity. This requires large holding spaces for test
leaves if the staining procedures require hours or days to complete. Additionally, stip
ples on individual leaves or plants, which can number in the thousands, take consid
erable time to quantify. Magnified images of stipples on leaves acquired with a video
camera and microscope, and then digitized into a computer, could be counted quickly
if the stipples and eggs could be enumerated based on color. The purpose of this study
was to compare several established staining techniques and to evaluate leaf clearing
and staining solutions and dyes, with the goal of finding techniques that would rap
idly differentiate (< 1/2 h) stained eggs and stipples on cos lettuce.


Four cos lettuce cultivars were used in the evaluation: 'Floricos 83', Valmaine',
'Tall Guzmaine', and 'Paris Island Cos'. Plants were grown in a greenhouse using
methods reported by Nuessly & Nagata (1994). The following methods were used for
exposing plants to flies, handling after exposure and evaluating L. trifolii mortality
and development in cos lettuce. Groups of four plants of the same cultivar were ex
posed to eight pairs of 48-h-old L. trifolii for 24 h within screened cages. After 72 h, all
flies were removed, and then the plants were carefully re-caged for another 72 h to in
sure that no further oviposition occurred. Since larvae emerge from eggs within 70 h,
any eggs detected in plants after this time could be assumed to be infertile. Time was
allowed also for callous formation at the leaf punctures and for larvae to emerge from
their eggs and start mining within the leaves. In addition, other leaves were pro
cessed within 24 h of oviposition to verify that the staining technique worked on both
viable and inviable eggs. The leaves were excised near the leaf axil and processed in
dividually. Stained leaves were examined with a stereomicroscope at 10 to 30 X using
direct and transmitted light.
Most staining techniques for insects, nematodes, or fungal hyphae involve boiling
the plant tissues in a solution to clear and stain them. This is followed by a destaining
process that may require additional boiling. Tissue disruption can occur during this
process. As a result, Parrella & Robb (1982) modified the lactophenol acid fuchsin pro
cedure of Simonet & Pienkowski (1977) to reduce the boiling time. This reduced tissue
disruption and improved staining of L. trifolii eggs in chrysanthemum, tomato and
celery leaves. However, their technique still required the leaves to steep in the solu
tion for >3 hr after boiling.
Beyond the time concerns, materials used in lactophenol staining procedures are
relatively expensive when used in the volumes required for our studies. For example,
the lactophenol acid fuchsin stock solution was made as follows: one part water; one
part lactic acid; one part phenol; two parts glycerin; and 0.5 g acid fuchsin per 0.5 liter
of solution. To address the cost concern, we evaluated other leaf clearing and de-stain

June, 1995

Nuessly et al.: Staining L. trifolii Eggs in Lettuce

ing agents (acetic acid, ethanol, glycerin, lactic acid, phenol, polyethylene glycol and
combinations) as well as insect and plant stains (cotton blue, methylene blue, acid
fuchsin, methyl red, scarlet red, saffarin, and gentian violet) to find a technique suit
able for our studies with lettuce. We also evaluated modifications of the following
techniques developed for staining insect eggs and plant parasitic nematodes in plant
tissue: lactophenol acid fuchsin (Parrella & Robb 1982), lactophenol cotton blue
(Franklin & Goodey 1959), water methyl red (Curtis 1942), and ethanol scarlet red
(Sugimoto 1976).
In addition, we evaluated a technique for highlighting stipples that involved excis
ing leaves, washing them in distilled water to remove any surface contaminants, and
then immersing them for 5 min in a 0.2% aqueous solution of silver nitrate. The leaves
were then washed in distilled water and immersed in a 7% aqueous solution of sodium
carbonate (with 0.375 ml 37% formaldehyde added per 100 ml solution) to develop the
stain. After 3 min in the developer, the leaves were removed and immersed in a 20%
aqueous solution of citric acid to stop the staining. A final distilled water wash com-
pleted the procedure. Care was taken not to handle the leaves without gloves, or to
abrade or tear the leaves prior to staining, as these areas would also become stained
and mask the stipples.


We found that the shape of the container used for the staining procedure ulti
mately affected the quality of the stain. For example, bending or folding the leaves
into beakers resulted in tissue destruction and too much dye infusion. Thus, to main
tain the structural integrity of the leaves, it was necessary to use a container large
enough for the entire leaf to lay horizontally in the solution. We found that a glass loaf
dish, 24 x 14 x 7.5 cm (L x W x D), was large enough for the cos lettuce leaves, and
small enough for the entire bottom surface of the dish to remain in contact with the
heating surface of a hot plate. One-half liter of stain solution provided adequate vol
ume for immersing the leaves during staining without touching the bottom of the
dish. The leaves were held immersed approximately 1.0 cm below the surface of the
test solutions with a strainer made from 5 mm diam glass rods.
Our goal was to develop procedures by which all staining, destaining and counter
staining could be completed in 15 min. We found it was necessary to boil the leaves in
order to achieve leaf clearing and egg-staining within this time limit, even though
this might result in some tissue disruption. Without boiling, the eggs were not suffi
ciently stained to discern them from the background color of the leaves.
Our preliminary tests with cos lettuce leaves indicated that boiling moderate to
older aged leaves >1 min, or young leaves >30 s, caused partial separation and tearing
of the abaxial and adaxial epidermal surfaces. This resulted in either excessive or lim-
ited staining of the compromised tissue, depending on the staining solution.
Leaves boiled in 50, 60, and 70% ethanol produced irregular clearing and the so
lution evaporated quickly. A boiling aqueous solution of 10% acetic acid and 50% eth
anol cleared the leaves well, but again this solution evaporated quickly and was
potentially dangerous because of alcohol's' flammability Adding stains to the acetic
acid-ethanol solution produced various results. Methylene blue stained the leaf tis
sues blue, but not the stipples or eggs. Methyl red heavily stained the leaf tissues and
stipples, but not the eggs. There was not enough contrast between eggs and stained
leaves to be able to easily identify the eggs. Gentian violet and saffarin poorly stained
the leaf tissues, stipples and eggs.
The egg staining methods of Curtis (1942) and Sugimoto (1976), who used final
staining solutions of water saturated with methyl red and 70% ethanol saturated

Florida Entomologist 78(2)

with scarlet red, respectively, did not stain L. trifolii eggs in lettuce. Lettuce leaves
were nearly destroyed by the three separate boiling steps used by Curtis (1942). Other
published methods of insect staining were not tested because they involved two sep
arate boiling procedures or long (i.e., > 24 h) staining times.
The lactophenol acid fuchsin solution of Simonet & Pienkowski (1977) provided
the best staining of L. trifolii eggs and the best overall clearing of leaves. Liquid phe
nol produced the most consistent results and was easier to work with than phenol
crystals. Modifications of their technique allowed us to successfully prepare the rela
tively soft lettuce tissue. The leaves were cleared and stained in a slow boiling solu
tion to minimize tissue disruption. Total staining time was ultimately reduced to 7
min per leaf.
The exact procedure we used for staining eggs in lettuce leaves was as follows:
leaves were placed in the boiling solution and held immersed for 1 min; young leaves
with soft tissue required boiling for only 30 s; the solution was removed from the hot
plate and the leaves allowed to steep in the stain for an additional 3 min; leaves were
removed from the stain and rinsed in warm water for 3 min to remove excess acid
fuchsin; stained leaves were then placed in cold water within 15 cm diam glass petri
dishes for microscopic examination.
This modified procedure stained the eggs pink to deep red and made them easy to
locate within the leaves. The 4 min staining and clearing procedure did not completely
clear the leaves of chlorophyll, but the degree of clearing was sufficient to make it easy
to locate the eggs. Leaves cleared better if they were boiled >1 min or steeped >3.5
min, but this resulted in greater destruction and darker staining of leaf tissues, re
spectively. Both of these conditions proved unsuitable for our purposes.
Efforts to remove or replace phenol from this procedure were unproductive. When
phenol was left out and the volume of water in the solution was doubled, or when phe
nol and water were replaced with an equal volume of polyethylene glycol, the leaves
became stained too heavily to be de-stained within several hours and the eggs were
poorly stained. The addition of 0.1 part acetic acid to either of these solutions without
phenol improved leaf clearing and egg staining, but the eggs were not stained as well
as when phenol was in the solution.
While this rapid lactophenol acid fuchsin technique produced well-stained eggs, it
did not stain all of the stipples. Many became indistinguishable once the leaves were
cleared. Parrella & Robb (1982) found that longer staining periods (> 3 h) with acid
fuchsin stained stipples pink to red on chrysanthemum, celery, and tomato leaves.
However, lettuce leaves steeped 10 min in lactophenol acid fuchsin became darkly
stained, while the stipples still could not be differentiated from normal leaf tissue. In
creasing steeping times made the eggs difficult to locate, so an alternative to longer
staining with acid fuchsin was desired. It was determined that stipples could be
quickly counterstained using a lactophenol cotton blue solution (Franklin & Goodey
1959) immediately following the egg staining procedure. The counterstain solution
was identical to the lactophenol egg staining solution except the acid fuchsin was re
placed with 0.4 g cotton blue per 0.5 liter of solution. The solution was initially boiled
and then allowed to cool to room temperature (25 + 1C) before use. Following the 3
min water rinse after the lactophenol acid fuchsin egg stain, leaves were placed in the
lactophenol cotton blue solution for 1 min. Leaves were then washed in warm water
for 1 min to remove excess stain. This procedure successfully stained the stipples blue
which provided a good contrast against the pink leaf tissue. Leaves left in the cotton
blue solution for >1 min absorbed excessive stain that blurred the boundaries of indi
vidual stipples. Although stipples could still be manually counted, the loss of indepen
dent boundaries around each stipple compromised efforts to use computer assisted

June, 1995

Nuessly et al.: Staining L. trifolii Eggs in Lettuce

image analysis for stipple counts. Efforts to combine the egg and stipple staining
steps failed. Leaves boiled in lactophenol acid fuchsin-cotton blue solution resulted in
the eggs being stained purple, while the stipples were unstained.
Larvae were not stained by acid fuchsin or cotton blue in lactophenol. However,
transmitted light made them appear yellow to greenish brown against the pink back
ground of the leaves following the egg and stipple staining steps. Tunnels within the
leaves were occasionally stained pink and blue.
While the lactophenol-cotton blue treatment was satisfactory for intact leaves, it
poorly stained stipples in areas of tissue disruption and in areas closely adjoining pri
mary leaf veins at the proximal region of leaves. Thus, the best overall technique for
highlighting stipples turned out to be the silver nitrate method. Black silver particles
adhered to all damaged portions of leaves and revealed all stipples, even if there was
no visible wound response. The blackened stipples provided enough contrast against
the green leaves for them to be successfully counted with computer image analysis
software. Using this technique offers several benefits. None of the solutions used in
this procedure masked or removed markings made on the leaves with felt tipped in
delible ink pens, so several leaves could be processed in the same container without
loss of leaf identity. These chemicals did not adversely alter leaf or egg tissues, so they
could be later stained using the lactophenol acid fuchsin solution to locate eggs. Since
there was no boiling involved, the procedure also allowed larvae within the leaves to
successfully complete development.
There was no obvious difference in leaf clearing or egg and stipple staining among
the four cos lettuce cultivars examined. Inviable eggs were stained the same color by
lactophenol acid fuchsin as viable eggs. Chorion of empty eggs absorbed little of the
acid fuchsin stain.
In spite of the costs of the lactophenol procedures, they worked the best with the
dyes to clear leaves and stain target tissues. The modified lactophenol acid fuchsin,
lactophenol cotton blue, and silver staining procedures all met our objectives of rapid
techniques that would differentially stain eggs and stipples. They produced adequate
contrast for counting eggs, stipples, and larvae in one step, and proved suitable for use
with computer image analysis software, e.g., Optimas (Optimas Corp., Edmonds,
Washington). They will be used to further our understanding of mechanisms involved
with host plant resistance to serpentine leafminer in lettuce.


Research was facilitated with support from the Wedgworth Family, Belle Glade,
FL, and from South Bay Growers Inc., South Bay, FL. Critical reviews of the manu
script were provided by R. Cherry, L. Datnoff, and J. Dusky (University of Florida).
This report published as Univ. of Florida Agricultural Experiment Station, Journal
Series no. R-03966.


CARLSON, O. V., AND E. T. HIBBS. 1962. Direct counts of potato leafhopper, Empoasca
fabae, eggs in Solanum leaves. Ann. Entomol. Soc. America 55: 512-515.
CURTIS, W. E. 1942. Method of locating insect eggs in plant tissue. J. Econ. Entomol.
35: 286.
FOSTER, R. E. 1986. Monitoring populations of Liriomyza trifolii (Diptera: Agromyz
idae) in celery with pupal counts. Florida Entomol. 69: 292-298.
FRANKLIN, M. T., AND J. B. GOODEY. 1959. A cotton-blue lactophenol technique for
mounting plant parasitic nematodes. J. Helminthol. 23: 175-178.

Florida Entomologist 78(2)

June, 1995

GENUNG, W. G., AND M. J. JANES. 1975. Host range, wild host significance, and in
field spread of Liriomyza trifolii and population build-up and effects of its par
asites in relation to Fall and Winter celery (Diptera: Agromyzidae). Belle Glade
AREC Res. Rpt. EV 1975-5, 18 p.
GILSTRAP, F. E., AND E. R. OATMAN. 1976. The bionomics of Scolothrips sexmaculatus
(Pergande) (Thysanoptera: Thripidae) an insect predator of spider mites. Hil
gardia 44: 2759.
LEIBEE, G. L. 1981. Insecticidal control of Liriomyza spp. on vegetables, pp. 216-220.
in D. J. Schuster [ed.], Proceedings of the IFAS-Ind. Conference on the Biology
and Control of Liriomyza Leafminers. IFAS, Univ. of Florida.
NUESSLY, G. S., AND R. T. NAGATA. 1993. Evaluation of damage by serpentine leaf
miner and banded cucumber beetles to cos lettuce. Everglades Res. and Ed.
Center Res. Rpt., EV 1993-2:76-77.
NUESSLY, G. S., AND R. T. NAGATA. 1994. Differential probing response of serpentine
leafminer, Liriomyza trifolii (Burgess), on cos lettuce. J. Entomol. Sci. 29: 330
PARRELLA, M. P., AND K. L. ROBB. 1982. Technique for staining eggs of Liriomyza tri
folii within chrysanthemum, celery, and tomato leaves. J. Econ. Entomol. 75:
SIMONET, E. E., AND R. L. PIENKOWSKI. 1977. Sampling and distribution of potato
leafhopper eggs in alfalfa stems. Ann. Entomol. Soc. America 51: 557-566.
SUGIMOTO, T 1976. On distribution of egg population of a leafmining fly, Phytomyza
ranunculi Schrank (Diptera, Agromyzidae) among leaves and in a leaf. Fac. Ag
ric. Kinki Univ. 9: 11-19.


Florida Entomologist 78(2)



'Departamento de Ciencia e
2Departamento de Enfermagem,
Universidade Estadual de Santa Cruz (UESC), 45660,
Ilh6us, Bahia, Brazil.

3Laborat6rio de Mirmecologia, Centro de Pesquisas do Cacau (CEPEC),
CEPLAC, C.P 7, 45600,
Itabuna, Bahia, Brazil.


The community structure of ants which infest houses in southern Bahia was stud
ied. Of the 31 species collected, 8 are considered as characteristic of human habitat,
even though some of them are also found in other conditions, such as in regional agro
systems. Pheidole megacephala was the dominant species and showed a negative as
sociation with most other ants, while Tapinoma melanocephalum was almost always
found with other ant species. The notable absence of the species L. humile in our tests
may be due to the antagonistic effects of Ph. megacephala.

June, 1995

Delabie et al.: House-Infesting Ants in Bahia

Key Words: Tapinoma melanocephalum, Pheidole megacephala, Linepithema humile,
Bahia, Brazil.


La structure de la communaut6 de fourmis qui infestent les habitations dans le
sud de Bahia a 6te 6tudiee. Parmi les 31 esp6ces rencontrees, 8 sont considerees ca
racteristiques de 1'habitat human, bien qu'elles soient aussi rencontrees dans d'au
tres conditions, telles que dans certain agrosystemes r6gionaux. Dans les
communaut6s, Pheidole megacephala est dominant et montre une association nega
tive avec la plupart des autres fourmis, alors que Tapinoma melanocephalum est a
peu pres toujours rencontree avec d'autres esp6ces. L'absence notable de Linepithema
humile de nos tests est probablement liee a son antagonisme avec Ph. megacephala.

Ants which infest houses have only recently received attention from entomologists
in South America (Brown 1964, Fowler et al. 1992, 1993, Ketelhut et al. 1993; Bueno
& Fowler 1994) in contrast to the Northern hemisphere, where ants which occur in
houses and hospitals have been studied for at least two or three decades (see the re
views of Smith 1965, Edwards 1986, Eichler 1990, Thompson 1990). Compared to
temperate environments, urban tropical environments may yield ideal conditions for
a range of species for foraging and nesting. However, in both the temperate and trop
ical urban conditions of Americas, ants which occur in the human environment are re
garded as destructive or detrimental to human health, food conservation and quality,
wood conservation, electrical installations, and electronic equipment (Smith 1965,
Fowler 1990, Thompson 1990, Vinson & McKay 1990, Fowler et al. 1993a & b).
A higher frequency of occurrence of ants in houses in tropical regions of South
America compared to houses in temperate regions (Fowler et al. 1993, Bueno &
Fowler 1994) is expected because of the favorable climatic conditions and the charac
teristics of human habitation in the tropics. To understand the community structure
of house-infesting ants in tropical regions, a cooperative study was conducted in the
State of Bahia, Brazil, by the Universidade Estadual de Santa Cruz (UESC) and the
Myrmecology Laboratory of the Cocoa Research Center (CEPEC), at Ilheus, with the
aim of describing the ant situation in houses of this region. This study is the first step
toward a larger study on the interactions between ants and human health in tropical


One hundred houses were randomly sampled for ants in the region of Ilheus
(1445'S, 39"13'W), in the southern portion of the State of Bahia. Ants were caught in
small glass test-tubes (length: 50 mm; diam: 7 mm) using honey as bait. In each
house, three test-tubes were placed on the floor in each room: livingroom, bedroom,
kitchen, and bathroom. Tubes were collected in the morning, 8 to 12 hours after initial
placement. Collected test-tubes were transported to the Myrmecology Laboratory of
the CEPEC where the ants were identified to species.
Data were analysed using a house or a room as the study unit. An evaluation of in
terrelations between the most common house-infesting ant species and their possible
associations with other species were tested by chi-square analysis with Yates' correc
tion applied (Siegel 1956). This procedure has previously been used by Room (1971)
and Majer et al. (1994) to study the ant mosaic in cocoa plantations.

Florida Entomologist 78(2)

* Al1 7 na

r R2 0.97

e 20


1 2 3 4 5 6
Number of species per house

Fig. 1. Frequency of houses infested by ants as a function of the number of species
found per house.


All houses sampled were occupied by at least one species of ant. A total of 31
species belonging to 14 genera and 4 subfamilies were found: Ponerinae:
Gnamptogenys [1 sp.], Odontomachus [1 sp.]; Myrmicinae: Acromyrmex [1 sp.],
Crematogaster [1 sp.], Monomorium [2 spp.], Pheidole [5 spp.], Solenopsis [4 spp.],
Tetramorium [3 spp.], Wasmannia [1 sp.]; Dolichoderinae: Dorymyrmex [1 sp.],
Tapinoma [1 sp.] and Formicinae: Brachymyrmex [1 sp.], Camponotus [7 spp.]),
Paratrechina [2 spp.]. Eight exotic species (25.8%), Monomorium floricola (Jerdon),
M. pharaonis (L.), Paratrechina longicornis (Latr.), Pheidole megacephala (Fabr.),
Tapinoma melanocephalum (Fabr.), Tetramorium bicarinatum (Nyl.), T lucayanum
Wheeler and T simillimum (Fr. Smith) (Delabie 1993) were responsible for 68.9% of
the total number of occurrences. Most of these ants are known as "tramp ants" and
have characteristics in common, such as polygyny, low intraspecific aggressivity, and
the ability to change their nest site easily (Passera 1993).
One to six species were found per house, and the frequency observed for each of
these six classes decreased inversely to the number of ants found (Fig. 1). The greatest
number of ant occurrences (29%) and species (22) was observed in the livingroom (Fig.
2A,B). In Bahia, the livingroom is generally the largest room in the house and proba
bly also offers more possibilities for nest sites (e.g., plant containers) to a variety of
species. The other three rooms sampled (bathroom, bedroom and kitchen) showed a
similar range of species number (14-16) and ant species occurrence (22-27%) (Fig.
2A,B). The kitchen was the location with the highest number of species and occur
rences and the bedroom the lowest.
Among the 31 species collected, eight (Table 1) could be considered to be truly char
acteristic of human habitats because they were found at different sites and in all parts

June, 1995


Delabie et al.: House-Infesting Ants in Bahia


Ant species % of Houses % of Occurrences

Pheidole megacephala 47.0 31.7
Tapinoma melanocephalum 48.0 22.9
Solenopsis saevissima 23.0 9.3
Paratrechina longicornis 18.0 7.9
Wasmannia auropunctata 12.0 5.5
Camponotus (Tanaemyrmex) sp 13.0 4.8
Pheidole sp4 13.0 3.8
Tetramorium simillimum 9.0 2.9

of the house. These eight species were responsible for 88.9% of the total number of oc
currences. In Bahia, these ants are not found exclusively in houses, but are also com-
monly found in gardens, secondary vegetation, and crops. Their distribution tends to
be different at higher latitudes where few species are limited exclusively to human
dwellings, e.g., M. pharaonis (Eichler, 1990). This difference is probably related to the
year-round thermal stability in this region, which allows these ants to live both in and
out of human habitats.
The most abundant ant species were Ph. megacephala (47% of infested houses and
31.7% of occurrences) and T melanocephalum (48% of infested houses and 22.9% of
The community structure at the room level (Fig. 3) showed that Ph. megacephala
was the only species that was truly dominant (see Majer et al. 1994). This ant was
negatively associated with the species P longicornis, Solenopsis saevissima (Fab.), T
melanocephalum and the little fire ant Wasmannia auropunctata (Roger). Ph. mega
cephala was frequently the only species in a room or a house, with all potential com-
petitors being excluded (Fig. 4). In contrast, the second most common species, T
melanocephalum, usually occurred with other ants (Fig. 4). However, it seems that

Bathroom Bathroom
2% 1
Kitchen 22 Kitchen

29% Livingroom
Bed m"'- Br14

Fig. 2. Percent of ant occurrences (A) and number of species (B) in sampled rooms.

Florida Entomologist 78(2)

Tetramorlum simillimum

Tapinoma melanocephalum

Solenopsis saevissima Pheidole megacephala Wasmannia auropunctata

Pheidole sp4 Paratrechina longicornis

association i_


Fig. 3. Community structure of house-infesting ants of Bahia.

this species can act as a rapid colonizer, because it was the only species found when
houses were new.
The study of size classes shows a bimodal distribution (Fig. 5) in which the smaller
ants (< 5mm) were more common. Most of the larger ants were species of the Cam
ponotus genus which are generally more active at night in Bahia (see, for example,
Delabie et al. 1991). Establishment of these large ant species in the house environ
ment may relate to their nocturnal behavior which does not interfere with the human
activity and removes them from competition with diurnal species.



n 10

1 2 3 4 5 6
Number of species per house

SPh. megacephala M T. melanocephalum 0 P. longicornia
S. saevissima M W. auropunctata

Fig. 4. Relative abundance of the 5 commonest ant species in Bahia houses.

June, 1995

Delabie et al.: House-Infesting Ants in Bahia



r 6

c 4


0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10
Length (mm)

Fig. 5. Distribution of house-infesting ants as a function of body length.

Linepithema humile (Mayr) was completely absent from our house samples. This
absence has been confirmed in all of south Bahia and confirmed by other extensive ex
periments over the region (J. H. C. D., unpublished data). However, the same species
was commonly found forming discrete societies in cocoa plantations within the same
region. It seems that L. humile avoids human settings in Bahia and this may be due
to an antagonism with Ph. megacephala which is dominant in houses and other
strongly human anthropized habitats. The antagonism between these two species has
been documented in other situations (Haskins & Haskins 1965, Crowell 1968). Thus,
the main difference between house-infesting ant communities in Bahia and other
parts of Brazil is that L. humile can be found everywhere in the urban environment
(Brown 1964, Fowler et al. 1993), while it seems restricted to only a few agricultural
situations in Bahia.


We wish to thank Dr. David F. Williams and three reviewers for valuable com
ments and corrections on an earlier version of this paper. This research was granted
by the Universidade Estadual de Santa Cruz (project "House-infesting ants and hu-
man health"). Paper presented at the IV International Symposium on Pest Ants, Belo
Horizonte (Brazil), November 1993.


BROWN, W. L., JR. 1964. Some tramp ants of Old World origin collected in Tropical
Brazil. Entomol. News 75: 14-15.
BUENO, O. C., AND H. G. FOWLER. 1994. Exotic ants and the ant fauna of Brazilian
hospitals, pp. 191-198 in D. F. Williams [ed.] Exotic ants: biology, impact, and
control of introduced species, Westview Press, Boulder.

Florida Entomologist 78(2)

CROWELL, K. L. 1968. Rates of competitive exclusion by the Argentine ant in Ber
muda. Ecology 49: 551-555.
DELABIE, J. H. C. 1993. Formigas ex6ticas na Bahia. Bahia, Analise e Dados, CEI, 3:
DELABIE, J. H. C., F. P. BENTON, AND M. A. DE MEDEIROS. 1991. La polydomie chez
les Formicidae arboricoles dans les cacaoybres du Br6sil: optimisation de 1'occu
pation de l'espace ou strat6gie defensive? Actes Coll. Insectes Sociaux 7: 173
EDWARDS, J. P. 1986. The biology, economic importance and control of the pharaoh's
ant, Monomorium pharaonis L., pp. 257-271 inS. B. Vinson [ed.], Economic im
pact and control of social insects. Praeger Press, New York.
EICHLER, W. 1990. Health aspects and control of Monomorium pharaonis, pp. 671 675
in R. K. Vander Meer, K. Jaffe and A. Cedeno [eds], Applied Myrmecology: a
World Perspective, Westview Press, Boulder.
FOWLER, H. G. 1990. Carpenter ants (Camponotus spp): pest status and human per
ception, pp. 525-532 in R. K. Vander Meer, K. Jaffe and A. Cedeno [eds], Applied
Myrmecology: a World Perspective, Westview Press, Boulder.
FOWLER, H. G., F. ANARUMA FILHO, AND O. C. BUENO. 1992. Vertical and horizontal
foraging: intra and interspecific spatial autocorrelation patterns in Tapinoma
melanocephalum and Monomorium pharaonis (Hymenoptera: Formicidae).
Ciincia e Cultura 44: 395-397.
FOWLER, H. G., F. ANARUMA FILHO, AND O. C. BUENO. 1993a. Seasonal space usage by
the introduced pharaoh's ant, Monomorium pharaonis (L.) (Hym., Formicidae),
in institutional settings in Brazil and its relation to other structural ant spe
cies. J. Appl. Entomol. 115: 416-419.
FOWLER, H. G., O. C. BUENO, T. SADATSUNE, AND A. C. MONTELLI. 1993b. Ants as po
tential vectors of pathogens in hospitals in the State of Sao Paulo, Brazil. Insect
Sci. Applic. 14: 367-370.
HASKINS, C. P., AND E. F. HASKINS. 1965. Pheidole megacephala and Iridomyrmexhu
milis in Bermuda -equilibrium or slow displacement? Ecology 46: 736-740.
BUENO. 1993. Formigas caseiras de Rio Claro-SP (Hymenoptera: Formicidae).
Resumos, IV International Symposium on Pest Ants, Belo Horizonte, Brazil.
MAJER, J. D., J. H. C. DELABIE, AND M. R. B. SMITH. 1994. Arboreal ant community
patterns in Brazilian cocoa farms. Biotropica 26: 73-83.
PASSERA, L. 1993. Quels sont les caracteres 6tho-physiologiques des "fourmis vaga
bondes"? Actes Coll. Insectes Sociaux, 8: 39-45.
ROOM, P. M. 1971. The relative distribution of ant species in Ghana's cocoa farms. J.
Anim. Ecol. 40: 735-751.
SIEGEL, S. 1956. Nonparametric Statistics for Behavioral Sciences (Estatistica Nao
param6trica para as Ciencias do Comportamento -traducao, 1975), McGraw
Hill, Sao Paulo, 350 pp.
SMITH, M. R. 1965. House-infesting ants of the Eastern United States, their recogni
tion, biology and economic importance. Agricultural Research Service,
U.S.D.A., Tech. Bull. 1326, Washington, 105 pp.
THOMPSON, C. R. 1990. Ants that have pest status in the United States, pp. 51-67 in
R. K. Vander Meer, K. Jaffe and A. Cedeno [eds], Applied Myrmecology: a World
Perspective, Westview Press, Boulder.
VINSON, S. B. AND W. P. McKAY. 1990. Effects of the fire ant, Solenopsis invicta, on
electrical circuits and equipment, pp. 496-503 in R. K. Vander Meer, K. Jaffe
and A. Cedeno [eds], Applied Myrmecology: a World Perspective, Westview
Press, Boulder.

June, 1995

Greenberg & Thomas: Coleoptera of Sand Pine Scrub 271


'USDA Forest Service Southeastern Forest Experiment Station
and University of Florida, Department of Wildlife and Range Sciences
118 Newins-Zeigler Hall
P.O. Box 14524
Gainesville, Florida 32604

Florida Department of Agriculture and Consumer Services
Division of Plant Industry
Florida State Collection of Arthropods
P.O. Box 147100
Gainesville, FL 32614-7100


Coleopteran assemblages were sampled monthly for one year using pitfall traps in
replicated sites of three 5 to 7-year-old disturbance treatments and mature forested
sand pine scrub in the Ocala National Forest, Marion County, Florida. Disturbance
treatments were (1) burning at high-intensity and salvage-logging; (2) clearcutting,
roller-chopping and broadcast seeding, and; (3) clearcutting and bracke-seeding.
Community similarity of coleopterans was high. No differences in species richness, di
versity, density, or evenness were detected. Of 40 species captured, only seven were
common (n > 50). Predaceous beetles were numerically dominant followed by scaven
gers. Few xylophagous or herbivorous coleopterans were captured, probably due to
trap bias. Peaks of annual above-ground terrestrial activity varied among species. An
absence of differences among treatments may reflect similar plant communities or
structural habitat features. Additionally, a dearth of mature forest specialists might
be predicted in systems where mature forest was historically rare due to large-scale,
high-intensity, and low-frequency wildfire.

Key Words: Beetle assemblage, clearcutting, wildfire.


Durante un ano fueron muestreadas mensualmente comunidades de cole6pteros
usando trampas de suelo en sitios replicados que poseen tres tipos de tratamientos
con 5 a 7 anos de edad, y el matorral que crece bajo un bosque maduro de pino de
arena (sand pine) en el Bosque Nacional de Ocala, en el condado de Marion, en la Flo
rida. Los tratamientos fueron (1) quema de gran intensidad y preservaci6n de los
troncos; (2) tala, corte de los troncos en pedazos y siembra de semillas al voleo, y (3)
tala y siembra de semillas en los claros talados. La similaridad de comunidades de co
le6pteros fue elevada. No se encontraron diferencias en riqueza de species, diversi
dad, o densidad. De 40 species capturadas solamente siete fueron comunes (n > 50).
Los escarabajos depredadores fueron numericamente dominates, seguidos por los
comedores de carrona. Fueron capturados pocos cole6pteros xil6fagos o herbivoros,
probabemente debido a los tipos de trampas. Los picos de actvidad terrestre por en
cima del suelo variaron entire las species. La falta de diferencia entire los tratamien
tos puede reflejar comunidades de plants o formas de habitats estructurales
similares. Adicionalmente, la falta de especialistas de bosque maduro podria ser pre

Florida Entomologist 78 (2)

dicha en sistemas donde el bosque maduro es hist6ricamente raro debido al fuego a
gran escala, alta intensidad y baja frecuencia.

Coleopteran assemblages often are useful indicators for monitoring effects of land
management practices on biodiversity (Eyre et al. 1989; Kremen et al. 1993). Many
species have specific habitat requirements and are capable of dispersing as habitat
suitability declines (Refseth 1980). Also, the microcosm of food web dynamics, niches,
and habitat requirements represented within the order Coleoptera permits insight
into ecological shifts as land use changes.
Historical changes in climate and sea level, and subsequent isolation have led to
a high endemism of plants (Christman & Judd 1990), vertebrates (Auffenberg 1982),
and invertebrates (Deyrup 1989) in scrub. Knowledge of scrub arthropod distribution
sheds light on the biogeography of Florida scrub but is poorly documented (Deyrup
The scrub ecosystem occurs in xeric, infertile sandy soils (Kalisz & Stone 1984)
along coastal areas and inland ridges of Florida and extreme southern Alabama. The
thick, sclerophyllous shrub layer is dominated by myrtle oak (Quercus myrtifolia
Willd.), sand live oak (Q. geminata Small), Chapman's oak (Q. chapmanii Sarg.),
rusty lyonia (Lyonia ferruginea Nutt.), and two species of palmetto (Serenoa repens
Small and Sabal etonia Swingle ex Nash.). Herbaceous groundcover is scant. Sand
pine (Pinus clausa Vasey ex Sarg.) scrub is a habitat variant containing sand pine as
a dominant component.
The peninsular Florida variety of sand pine, Pinus c. clausa, has serotinous cones.
Here, the naturally even-aged, monospecific sand pine canopy is maintained by the
release of copious quantities of seed (recorded as > 2.47 million per ha, Cooper et al.
1959) following stand-replacing wildfire. Historically, low-frequency, high-intensity,
and large-scale wildfire created a forest mosaic of temporally shifting age-classes
(Rawlings 1933, Webber 1935, Bartram 1955, Myers 1990). An open, shrub-domi
nated habitat was maintained between wildfires, while mature forest probably ex
isted intermittently in time and space, especially in sites protected from fire. The
largest remaining area of scrub occurs in the Ocala National Forest in central Florida.
Current forest management of sand pine scrub there entails clearcutting patches of
approximately 8-24 ha. Heavy machinery used during the clearcutting operation
crushes and kills nearly all above-ground vegetation. Clearcutting is commonly fol
lowed either by roller-chopping and broadcast seeding or "bracke-seeding." Roller
chopper blades penetrate the soil to a maximum depth of 15 cm. Soil surface distur
bance with this method is nearly complete. Bracke-seeding entails direct seeding
along small, machine-created ridges (about 8 cm high). This method patch-scarifies
approximately 30% of the soil surface (Outcalt 1990). Because of the wood fiber value
and the possibility of large-scale, uncontrolled burns, fires in sand pine stands are
usually extinguished as rapidly as possible. Normally, burned sites are salvage
Plant community recovery and habitat structure of clearcuts are similar in many
respects to community recovery and structure following high-intensity wildfire
(Campbell & Christman 1982, Abrahamson 1984a, 1984b, Schmalzer & Hinkle 1992,
Greenberg et al. 1995). Major differences include (1) the absence of fire-associated
cues for attracting pyrophyllic coleopterans to clearcuts; (2) the presence of few stand
ing trees or snags in clearcuts versus an abundance of snags for several years follow

June, 1995

Greenberg & Thomas: Coleoptera of Sand Pine Scrub 273

ing a wildfire (unless salvage-logged, as in this study); (3) more slash piles and less
bole-sized woody debris in clearcuts (personal observation), and; (4) landscape pat
terns such as patch size and connectivity.
Because it occurs on ideal sites for citrus and urban development, sand pine scrub
is fast becoming and endangered ecosystem (Myers 1990). It is critical that the scrub
ecosystem on public lands be managed to maintain the characteristic species diversity
associated with the ecosystem and its driving processes.
This study is one portion of a larger study comparing plant, bird, and herpetofau
nal communities among four stand treatments: intense burning and salvage-logging;
clearcutting followed by either roller-chopping or bracke-seeding; and naturally re
generated, mature sand pine scrub. Here we describe and quantitatively compare co
leopteran assemblages among these treatments. The study also provides information
on annual cycles of commonly captured coleopterans.


Coleopterans were sampled using drift fences and pitfall traps in three replicated
5 to 7-year-old disturbance treatments and mature forested sand pine scrub (n
three sites each) in the Ocala National Forest, Marion County, Florida (Table 1). Dis
turbance treatments were (1) high-intensity burning, salvage-logging, and natural re
generation (HIBS); (2) clearcutting, roller-chopping, and broadcast seeding (RC), and;
(3) clearcutting and bracke-seeding (BK). Mature (> 55 yr) sand pine stands that had
naturally regenerated following a stand-replacing fire in 1935 were used as a control
(MF). All sites had similar elevation, topographic, and soil characteristics; same pre
treatment age, and (known) disturbance history (identical to MF); same post-treat
ment age for HIBS, RC, and BK sites (within 1.5 yr) (Table 1); were greater than 8.5
ha; and were more than 0.9 km from known water sources.
Sand pine density and height were measured in five 100m2 plots per site (or in a
20 m2 subplot if density was high). Three 10-m line transects were randomly estab
lished within quadrats to quantify percent cover of the vegetation and microsite char
acteristics by category, including herb, shrub, pine, woody debris, leaf litter, and bare
ground, using the line intercept technique (Mueller-Dombois & Ellenberg 1974). In
MF, sand pine canopy cover was estimated using a spherical densiometer at the mid
point of each line transect.
Trapping arrays were designed and concurrently used for herpetofaunal sampling
(Greenberg et al. 1994) but proved effective in sampling surface-active terrestrial ar
thropods as well. Arrays (modified from Campbell & Christman 1982) consisted of
eight 7.6-m lengths of erect 0.5-m-high galvanized metal flashing arranged in an "L
shaped pattern with a 7.6-m space between each length. Two black 18.9-liter plastic
paint buckets (pitfall traps) with 28.5-cm diam were sunk flush with the ground at
both ends of each fence (n = 16 pitfall traps per site). To improve drainage, 1.25-cm
holes were drilled into the bottoms of the pitfall traps. Drill holes were blocked with
sticks to prevent escape. No killing agents were used in pitfalls. Arrays were located
a minimum of 25 m from roads or stand edges (except for two drift fences of one array).
We assumed that consumption of arthropods by vertebrates was minimal due to low
capture rates of vertebrates during arthropod sampling periods. Any effects were con
sistent among treatments.
Arthropods were trapped for one 48-hour period each month from October 1991
through September 1992. Coleopterans were preserved in ethyl alcohol.
One-way analysis of variance (ANOVA) (SAS 1989) was used to determine dif
ferences among treatments in total numbers of individuals trapped, total numbers by

274 Florida Entomologist 78(2) June, 1995



7 cc

z -

S c cm c<




M c C cC
H P~

Greenberg & Thomas: Coleoptera of Sand Pine Scrub 275

species, Shannon's diversity indices, species richness, and evenness (Brower & Zar
1977). Horn's Index of Community Similarity (Horn 1966) was used to compare com-
munity overlap among treatments.
Each species was assigned to one of four feeding guilds: predator, scavenger (car
rion and dung beetles), herbivore (root-, stem-, foliage-, or flower-feeding beetles), and
xylophage based on adult food habits. ANOVA was used to detect differences in co
leopteran density for each feeding guild.
Annual above-ground terrestrial activity cycles were estimated from captures of
each commonly trapped species (n > 50 individuals caught over the 1 year period) for
each trapping period.


Mature forest differed structurally from disturbance treatments in having lower
stem density and greater foliar cover and height of sand pine. Mature stands also had
less bare ground and higher leaf litter, nonwoody plants (primarily lichens), and
shrub cover than disturbance treatments (Table 2) (see also Greenberg et al. 1995).
A total of 1,849 beetles representing 40 species in 14 families was captured (Table
3). Only seven species were commonly captured (> 50 individuals). Two carabid spe
cies of Pasimachus (P strenuous LeConte and P subsulcatus Say) were dominant fol
lowed by the tenebrionid Polopinus young Kritsky and an undescribed tenebrionid
species of Helops.
Three females of Romulus globosus Knull (Cerambycidae), a rare scrub endemic,
were captured in the July trapping period. Based on collection dates of other speci
mens, this species is most active in June and July (Thomas 1991). Peltotrupesyoungi
Howden, a species endemic to the Ocala National Forest scrub in Marion and Putnam
counties (Woodruff 1973), was relatively abundant. There were no differences in den
sity of individual species among treatments (Table 3).
There were no differences among treatments in density, diversity, or evenness of
captured coleopterans (Table 4). Species richness was lower in MF than in any of the
disturbance treatments, but differences were not statistically significant (Table 4).
Horn's Index of Community Similarity indicated a high degree of community overlap
among all treatments (Table 5).
Feeding guild structure did not differ among treatments (Table 3; Fig. 1). Preda
ceous beetles were dominant, composing 63-73% of total beetle numbers, followed by
scavengers (21-30%), herbivores (5-9%) and xylophages (< 1%).
Two general patterns of annual above-ground terrestrial activity cycles were ap
parent. Among completely trapped species, Peltotrupesyoungi Howden and Helops sp.
appeared to be most active above-ground from December-March and inactive mostly
from June-October. Several other species were active above-ground from at least May
October (Polypleurus sp. from April-August) but inactive during the cooler winter
months (Fig. 2).


Several studies report decreases in forest specialists but increases in overall co
leopteran diversity following forest disturbance, fragmentation, or deforestation as
species of open habitat invade and mature forest generalists persist (Lenski 1982,
Baguette & Gerard 1993, Buse & Good 1993, Halme & Niemela 1993, Niemela et al.
1993). Conversely, Niemela et al. (1988) found few differences in carabid assemblages

r 0


z Z



< E-1


m 2

Cr V



E-1 <


M 0



+1 "
2 M

oa I
a s

^ f
a a
m >


June, 1995

Florida Entomologist 78(2)

-"I c 0 c'
0 ^ 06? C' C' cz C.6 l C^

U CDN- N Nc c
i- 3 QOC^Cno O iO 'mLO
m3 o ~cb ^cLo r c o
0M N M M



CC C.0

Q6 CfliCcli Q6 4' C Lfl 6z

fj o
C N-


Ln x x ,c.
c. yiI M c.0
CU -, Ct" N -
>c cc ccc b d cM i



C t OO r C C D l c'J tcfl D


E C-
C3 cb -^ c O Q
Ev O cD^Cl ^C

H H ^~U.

Greenberg & Thomas: Coleoptera of Sand Pine Scrub


c'J c'J -n



, CnC 0 -




- < 0,

CO ^ ,00 =D
CD -

-- -i i i CD N-- CD
0 Cf0

M i


| | | M 00
cn m tat
at c










Florida Entomologist 78(2)


CMM, c' CM ^- -

C'J -

N-~ N-



I' o




,a ;n
; -e~r e -v

June, 1995

Greenberg & Thomas: Coleoptera of Sand Pine Scrub

I C 2 I --

oi LoCD-Io
CO -

cn tN-C'JOtmfl Lfl

CmJ cM C Cj m C C


! l








c~ c~


Cn C'

V) n o V) V) ' V) O O] V) V) V) V)



- C






zo a'
M ^ S

280 Florida Entomologist 78(2) June, 1995


D 1 00 C
z U
a m
I z

o H
o L

Ll f C

z Q

D 0 C
Qo o g

2I I
H -0
0 0

< o I )

H? ^
2 0 t- C


3 S0 -
^ 3
0 >,/Wa
"^ K S !
^a a &^
MW 0 1

W 3
Mg "l y2c

Greenberg & Thomas: Coleoptera of Sand Pine Scrub 281


Richness Diversity Evenness

Burn-salvage (HIBS) 15.3 0.800 0.676
(1.9) (0.057) (0.019)
Chop (RC) 13.7 0.783 0.691
(1.5) (0.061) (0.026)
Bracke (BK) 14.7 0.763 0.655
(0.7) (0.018) (0.025)
Mature (MF) 10.7 0.681 0.667
(0.9) (0.048) (0.055)

P-value 0.098 0.385 0.899
F-value 3.00 1.15 0.19
df 2 2 2

between mature and successional coniferous taiga.
sand pine scrub as well.

This appears to be the case for

Several possibilities exist for the similarity in species composition and community
structure among disturbance treatments and mature sand pine scrub. The pitfall
trapping technique may have missed important species (Adis 1989), including non
terrestrial species such as many monophagous herbivores, xylophages, and mature
forest-specialists. However, similar studies using pitfall traps detected differences in
species composition among treatments (e.g. Lenski 1982, Baguette & Gerard 1993,
Buse & Good 1993, Halme & Niemela 1993, Niemela et al. 1993). Because trap bias
is consistent across treatments, comparisons using standardized trapping techniques
are valid.
Differences among treatments may have been present for the first few years fol
lowing disturbance but were not detected because stands were not sampled until 5-7


Burn-salvage Chop Bracke
(HIBS) (RC) (BK)

Burn-salvage (HIBS)
Chop (RC) 0.931
Bracke (BK) 0.945 0.939
Mature (MF) 0.913 0.821 0.874

2As Ro approaches one, community overlap increases.

Florida Entomologist 78(2)

PPredator EScavenger mHerbivore OXylophage




Burn-Salvage Chop Bracke Mature
Fig. 1. Mean proportion of Coleoptera in four feeding guilds in three disturbance
treatments and mature forest in sand pine scrub, Ocala National Forest, Florida.

years post-disturbance. Small stand size and time since disturbance permitted ample
opportunity for recolonization of disturbance treatments by coleopterans. Nonethe
less, similarity in coleopteran community composition indicates that suitable habitat
exists 5-7 years post-disturbance, whether by silvicultural means or by wildfire.
Similarity of many habitat features between 5-7-year-old disturbed and mature
sand pine scrub may also contribute to like coleopteran community composition
among treatments. Differences in pine age, height, and density, as well as percent
bare ground were the most prominent differences between MF and disturbance treat
ments. However, minimal differences in composition of dominant (woody) plant spe
cies exist between early and late-successional sand pine scrub or among disturbance
treatments (Greenberg et al. 1995). Infertile, acid sands and low water availability
render productivity and palatability of the sclerophyllous vegetation low in sand pine
scrub. Although several species of herbs occur within the study area (Greenberg et al.
1995), total cover is low. These factors may partially explain the low representation of
herbivorous coleopterans. Nonetheless, endemic, host-specific herbivores might be ex
pected where plant species distribution is restricted as is sand pine or endemic her
baceous species of scrub (Deyrup 1989).
Herbaceous plant species diversity is higher in disturbed scrub than in mature for
est. Among disturbance treatments, herbaceous plant community similarity, species
richness, and species diversity did not differ significantly (Greenberg et al. 1995).
However, disturbance treatments could differentially affect occurrences of some plant
species. Trap bias against plant specialists and xylophages could lead to potential dif
ferences among treatments going undetected.
The apparent absence of forest specialists within sand pine scrub could be due to
the historical prevalence of young forests due to low-frequency, high intensity wildfire
(Rawlings 1933, Webber 1935, Bartram 1955, Bonan & Shugart 1989). Even in the ab
sence of fire, sand pine stands begin to break up after about 50-70 years (Myers 1990)
due to disease or structural weakness. Historically, sand pine density in the study
area probably varied spatially and temporally as well; many scrubs have few to no

June, 1995

Greenberg & Thomas: Coleoptera of Sand Pine Scrub

(a) 60

U 50-

. 40-


o 20 -
E 10-

(c) 250

-- 200

.E 100
z 0

-*- Peltotrupes young Howden -0- Strategus anteus (Drury)

-*- Pasimachus strenuus LeConte
- 0- Pasimachus subsulcatus LeConte

Oct Nov Dec Jan Feb Mar Apr May May Jun Jul Aug Sept

Fig. 2. Annual above-ground terrestrial activity cycles of (a) tenebrionids; (b) scar
abs; and (c) carabids in the Ocala National Forest, Florida.

sand pine. Coleopterans are unlikely to have evolved into a specialized mature forest
niche where habitat availability was rare or unreliable.
The predominance of carabids, tenebrionids, and staphylinids reflects both the
trap bias toward ground-dwelling coleopterans and their relative abundance. The
overwhelming prevalence of predaceous Coleoptera (primarily carabids) may be due

Florida Entomologist 78(2)

to an abundance of arthropod prey within sampled sites (C. H. G. unpublished data).
Predaceous beetles may have consumed non-predaceous beetles in traps, but the ad
ditional presence of other arthropod prey reduces the likelihood that this was a sig
nificant problem.
Xylophagous species were under-represented in this study due to trap bias. Addi
tionally, this guild may have been more abundant had HIBS not been salvage-logged.
However, small-diam woody debris was available in the form of slash piles (from on
site delimbing) in HIBS and BK treatments. Woodpiles were less common in RC sites
because the roller-chopping fragmented and buried woody debris (Table 2). Few snags
were present in MF or in disturbance treatments.
Stand age may also account for low representation by xylophages which may have
been more abundant during the first 1 2 years post-logging. The absence of pyrophyl
lous species in the HIBS treatment was probably due to elapsed time since fire and
post fire salvage logging.
Monthly variation in above-ground terrestrial activity cycles suggests that trap
ping in all months is necessary for a complete census of coleopterans as well as for
gaining an understanding of their ecology.


We thank Ken Benfield for field assistance and the Ocala National Forest staff for
their time and cooperation. Roger Anderson offered useful suggestions for arthropod
sampling. Pat Outcalt and Sandra Coleman provided technical assistance with the
manuscript. Doria Gordon, Mark Deyrup, John Foltz, and Wayne Dixon reviewed an
earlier version of this manuscript and provided useful suggestions.
This study was funded by the USDA Forest Service Southern Research Station.


ABRAHAMSON, W. G. 1984a. Species responses to fire on Florida Lake Wales Ridge.
American J. Botany 71: 35-43.
ABRAHAMSON, W. G. 1984b. Post-fire recovery of Florida Lake Wales Ridge vegetation.
American J. Botany 71: 921.
ADIS, J. 1989. Problems of interpreting arthropod sampling with pitfall traps. Zoolo
gischer Anzeiger 202: 177-184.
AUFFENBERG, W. 1982. Florida environments and their herpetofaunas, part 3. Herpe
togeography Florida Herpetol. 4: 135.
BAGUETTE, M., AND S. GERARD. 1993. Effects of spruce plantations on carabid beetles
in southern Belgium. Pedobiologia 37: 129-140.
BARTRAM, W. 1955. The Travels of William Bartram. 1791. M. Van Doren [ed.], Dover,
New York.
BONAN, G. B., AND H. H. SHUGART. 1989. Environmental factors and ecological pro
cesses in boreal forests. Annu. Rev. Ecol. Syst. 20: 128.
BROWER, J. E., AND J. H. ZAR. 1977. Field and Laboratory Methods for General Ecol
ogy W. C. Brown Company, Dubuque, Iowa.
BUSE, A., AND J. E. G. GOOD. 1993. The effects of conifer forest design and manage
ment on abundance and diversity of rove beetles (Coleoptera: Staphylinidae):
implications for conservation. Biol. Conserv. 64: 6776.
CAMPBELL, H. W., AND S. P. CHRISTMAN. 1982. Field techniques for herpetofaunal
community analysis, p. 193-200 in N. J. Scott [ed.]. Herpetological communi
ties. U.S. Department of Interior, Fish and Wildlife Service Wildlife Research
Report 13.
CHRISTMAN, S. P., AND W. S. JUDD. 1990. Notes on plants endemic to Florida scrub.
Florida Scientist 53: 5273.

June, 1995

Greenberg & Thomas: Coleoptera of Sand Pine Scrub 285

generation on the Ocala National Forest. U.S. Department of Agriculture For
est Service Productivity Research Report 30.
DEYRUP, M. 1989. Arthropods endemic to Florida scrub. Florida Scientist 52: 254-270.
EYRE, M. D., M. L. LUFF, S. P. RUSHTON, AND C. J. TOPPING. 1989. Ground beetles and
weevils (Carabidae and Curculionoidea) as indicators of grassland manage
ment practices. J. Applied Entomol. 107: 508-517.
GREENBERG, C. H., D. G. NEARLY, L. D. HARRIS, AND S. P. LINDA. 1995. Vegetation re
cover following high-intensity wildfire and silvicultural treatments in sand
pine scrub. American Midl. Nat. 133: 372-412.
GREENBERG, C. H., D. G. NEARY, AND L. D. HARRIS. 1994. Effect of high-intensity
wildfire and silvicultural treatments on herpetofaunal communities in sand
pine scrub. Conserv. Biol. 8: 1047-1057.
HALME, E., AND J. NIEMELA. 1993. Carabid beetles in fragments of coniferous forest.
Annles Zoologici Fennici 30: 17-29.
HORN, H. S. 1966. Measurements of overlap in comparative ecological studies. Amer
ican Nat. 100: 419-424.
KALISZ, P. J., AND E. L. STONE. 1984. The longleaf pine islands of the Ocala National
Forest, Florida: a soil study. Ecology 65: 1743-1754.
JAYAN. 1993. Terrestrial arthropod assemblages: their use in conservation
planning. Conserv. Biol. 7: 796-808.
LENSKI, R. E. 1982. The impact of forest cutting on the diversity of ground beetles (Co
leoptera: Carabidae) in the southern Appalachians. Ecol. Entomol. 7: 385-390.
MUELLER-DOMBOIS, D., AND H. ELLENBERG. 1974. Aims and Methods of Vegetation
Ecology. Wiley, New York.
MYERS, R. L. 1990. Scrub and high pine, pp. 150-193 in R. L. Myers, and J. J. Ewel
[eds.]. Ecosystems of Florida. University of Central Florida Press, Orlando.
distribution of carabid beetles in fragments of old coniferous taiga and adjacent
managed forests. Annales Zoologici Fennici 25: 107-119.
NIEMELA, J., D. LANGOR, AND J. R. SPENCE. 1993. Effects of clearcut harvesting on bo
real ground-beetle assemblages (Coleoptera: Carabidae) in western Canada.
Conserv. Biol. 7: 551-561.
OUTCALT, K. W. 1990. Operational trials of a scarifier-seeder for regenerating Ocala
sand pine. South J. Appl. For. 14: 85-88.
RAWLINGS, M. K. 1933. South Moon Under. Scribners, New York.
REFSETH, D. 1980. Ecological analyses of carabid communities-potential use in bio
logical classification for nature conservation. Biol. Conserv. 17: 131-141.
SAS INSTITUTE. 1989. SAS User's Guide: Statistics, Version 6, Fourth edition SAS In
stitute, Cary, North Carolina.
SCHMALZER, P. A., AND C. R. HINKLE. 1992. Recovery of oak-saw palmetto scrub after
fire. Castanea 57: 158-173.
THOMAS, M. C. 1991. Rediscovery of Romulus globosus Knull (Coleoptera: Ceramby
cidae). Insecta Mundi 5: 127-128.
WEBBER, H. J. 1935. The Florida scrub, a fire fighting association. American J. Botany
22: 344-361.
WOODRUFF, R. E. 1973. The scarab beetles of Florida (Coleoptera: Scarabaeidae), part
1. The Laparosticti (Subfamilies: Scarabaenae, Aphodiinae, Hybosorinae,
Ochodaeinae, Geotrupinae, Acanthocerinae). Florida Department of Agricul
ture, Division of Plant Industry. Arthropods of Florida and Neighboring Land
Areas vol. 8.

Florida Entomologist 78(2)


'Laboratorio de Ecologia y Sistematica de Microartr6podos,
Depto. Biologia, Facultad de Ciencias, Universidad Nacional
Aut6noma de Mexico.

2Departamento de Biologia Animal y Humana. Facultad de Biologia.
Universidad de La Habana, Cuba.


Two new species of Deuterosminthurus are described and illustrated. One is found
in dry forest canopy at Chamela, Jalisco, Mexico, and the other is associated with sug
arcane at La Habana, Cuba. A key for identification of the American species is in

Key Words: Deuterosminthurus, taxonomy, species key


Se described e ilustran dos nuevas species de Deuterosminthurus. Una de ellas
encontrada en la canopia de la selva baja caducifolia de Chamela, Jalisco, Mexico, y
la otra asociada al cultivo de la cana de azucar en La Habana, Cuba. Se incluye una
clave para la identificaci6n de las species del Continente Americano.

Deuterosminthurus (Bdrner, 1901)

Body dorsally with a depression behind middle of great abdomen. Anogenital seg
ment strongly elongated and distinctly delimited from great abdomen. Integument
finely granulated. Setae unciliated, curved and moderately long. Great abdomen with
three pairs of fine and long trichobothria inserted in low papillae. Two trichobothria
on each side of genital segment. Ant. IV divided into 3 9 secondary joints in addition
to basal and apical parts. Eyes 8 + 8. Claw untoothed or armed with very small inner
and lateral teeth; tunica and pseudonychia absent. Unguicular filament sharply
pointed, or blunt at the tip, or knobbed. Tenent hairs clavate 3, 3, and 2. Dentes
smooth, about three times the length of mucro. Mucro spoon-like with smooth dorsal
edges, without mucronal seta. Secondary dimorphism in male: smaller body and head
length, longer antennae and different shape of anogenital segment; clasping organ on
anal segment absent. Female with subanal appendages. Type species: Sminthurus
bicinctus Koch, 1840.
The genus has 32 named species; 8 are known from the Neotropical Region and 7
from the Nearctic Region. Deuterosminthurus lippsoni Snider, 1978 is recorded in
both regions. It was described from Maryland and reported from Florida and Califor
nia. This species was also reported from Cuba by Banasco Almentero (1987). This ge

June, 1995

Palacios Vargas & Gonzalez: New Species ofDeuterosminthurus 287

nus was only recently recorded from Chamela, Mexico (Palacios-Vargas & G6mez
Anaya, 1993), as described here in more detail.
Deuterosminthurus is a cosmopolitan genus that lives in litter, epiphytic Bromeli
aceae and the forest canopy One species, D. russatus Maynard, 1951 is known to be
part of the transitory epineuston, so it can be found in lakes and other freshwater bod

Deuterosminthurus delatorrei New Species
(Figs. 1 -16)


Antennae light yellow, segment I with purple pigment dorsally Head with two
bands of pigmentation (Figs. 1 and 2), one posterior to eyes and one laterally; genae
with purple pigment extending to occiput. Body yellow with purple bands and patches
forming a "W". Abd. VI with two small dark purple patches. Legs and furcula without
Eyes 8 + 8; ocellus D half the diam of H (Fig. 3). Antennal segments female (n=9)
ratio 1:1.5:2:1:4.9 (Fig. 4); male (n=2) 1:2:2.5:5.7. Ant. IV subannulated into four in
termediates. Apical bulb simple; one sensorial small organ and one dorso-external mi
crosensillum (Fig. 5). Ratio head-antenna: female, 1:1.3; male, 1:1.4. Ant. III with

:2 2 2

I i I ,",

*''e *;

Figs. 1 and 2. Deuterosminthurus delatorrei sp. nov. 1. Dorsal distribution of pig
ments. 2. Habitus lateral view.

Florida Entomologist 78(2)

subapical sensory rods lying in shallow depressions; an accessory sensory rod slightly
oblique and posterior to sensory rods; setae normal (Fig. 6). Thoracic segmentation
not distinct. Metatrochanters without oval organs. Metafemora with two posterior
setulae. Tibiotarsi of the pro and mesolegs with three well-developed, appressed,
clavate tenent hairs; meta-tibiotarsi with two tenent hairs (Figs. 7 11). Pretarsus
with anterior setulae. Unguis lanceolate with inner tooth short and a small lateral
tooth. Unguiculus with a strong bristle, tapering to a strong knob (Figs. 9, 11). Ratio
unguiculus: unguis = 1:0.9. Sacs of ventral tube tuberculate. Rami of tenaculum tri
dentate (Fig. 12); anterior corpus with three apical setulae. Manubrium with 12 dor
sal setae. Dens with 5 vental setae, 6 lateral and 19 others, one of them in dorsal
position similar to a bothriotricum (Fig. 13). Mucro with rachis fused to lateral lamel
lae in a spoon shape, rachis forming a clear tip distally. Ratio mucro: dens = 1:2.7.
Anal papillae with normal curving setae; female subanal appendage setiform (37 pm),
tip may be acuminate or with 2 4 teeth (Fig. 14). Setae of head and body short and
curving. Male with eugenital setae longer than circumgenital setae (Fig. 15). Maxi
mum size of female (n=9) 0.7 mm and male (n=2) 0.47 mm. Head chaetotaxy as in Fig.
16. Male similar to female except for the lack of subanal appendages and smaller size.


Cuba: La Habana: Boyeros. Experimental sugarcane field.


Holotype 2 on slide, two 6 6 paratypes and eight Y 2 paratypes on slides. Two
paratypes will be kept at the Facultad de Biologia, Universidad de La Habana, the ho
lotype and other paratypes at Facultad de Ciencias, UNAM. Type material data:
sugar cane field, 80m altitude, yellow traps, 28-IV 1992, 4-VI-1992, V. Gonzalez, M.
Diaz and D. Prieto colls.


The number of teeth on subanal appendage (n=11) varied as follows (frequency in
parentheses): one (1), two (5), three (3), and four (2). In the same specimen the num
ber of teeth may vary from side to side. One bifid seta was observed on a trochanter.
Another specimen had one extra ventral dental seta.


This species is named after Dr. Salvador de La Torre, Cuban entomologist.


The new species is very similar to D. lippsoni Snider, 1978, which shares the pres
ence of a heavily knobbed unguiculus. Other species with a knobbed unguiculus are
D. xeromorphus Snider, 1978 and D. wexfordensis (Snider, 1969). D. lippsoni differs in
having five subsegments of antennal IV (after the original description) while in D. de
latorrei it has four. The only reduced ocellus in the new species is "D". D. lippsoni has
an oval organ in the metatrochanter, which is lacking in D. delatorrei. D. lippsoni has

June, 1995

Palacios Vargas & Gonzalez: New Species ofDeuterosminthurus 289


- 4

5 r 12



!\ '

8 ~



'.tb A

*m 2


'4 I,1
-:'' 14




Figs. 3 -16. Deuterosminthurus delatorrei sp. nov. 3. Right eyepatch. 4. Antennal
segments I to IV. 5. Ant. IV distal portion. 6. Ant. III distal portion. 7. Tibiotarsus II
external view. 8. Tibiotarsus II internal view. 9. Foot complex of leg II. 10. Tibiotarsus
III. 11. Foot complex of leg III. 12. Tenaculum. 13. Dens and mucro. 14. Female, anal
and genital region, with subanal appendages enlarged, lateral view. 15. Male, anal
and genital region. 16. Head chaetotaxy.




Florida Entomologist 78(2)

more pigmentation in a very different pattern. Subanal appendages in D. delatorrei
are thicker and apically toothed.
This species occurs exclusively in epiphytic environments. It was never found in
soil and litter samples taken monthly during a two-year study of sugarcane. This spe
cies was first caught when yellow traps were used.

Deuterosminthurus maassius New Species
(Figs. 17 29)


Antennae light yellow, segment I with dark purple pigment. Head with ocular
patches and posterior region with pigmentation (Figs. 17, 18). Great abdomen very
dark, except for a clear area in the middle thoracic region. Anal papilla with only dor
sal pigmentation dark purple; ventrally body is light, lacking of dark purple. Legs and
furcula lack pigmentation.
Eyes 8 + 8; ocellus D half the diam of H; F and G a little smaller than H (Fig. 19).
Antennal segments female, (n=8) ratio of 1:1.7:3:1:5.4 (Fig. 20). Ant. IV slightly sub
annulated into three intermediates. Apical bulb simple; one small sensorial organ and
one microsensilla dorso-external (Fig. 21). Ratio head: antenna as 1: 1.6. Ant. III with
subapical sensory rods lying in shallow depressions; accessory sensory rod slightly ob
lique and posterior to sensory rods; with ventral microsensillae (Fig. 22). Some setae
on Ant. I and II are long and thick. Thoracic segmentation not distinct. Metatrochant
ers with oval organs. Metafemora without posterior setulae. Tibiotarsi of the pro and
mesolegs with 3 strong, appressed, clavate tenent hairs; meta-tibiotarsi with 2 tenent
hairs (Figs. 23, 25). Pretarsus with an anterior setula. Unguis lanceolate with a small
apical inner tooth and small lateral teeth. Unguiculus lanceolate ending in an acumi
nate bristle (Figs. 24, 26). Ratio unguiculus: unguis = 1:2.3. Sacs of ventral tube tu
berculate. Rami of tenaculum tridentate; corpus with 3 apical setulae. Manubrium
with 12 dorsal setae. Dens with 3 ventral setae, 6 lateral and 13 others, one of them
in dorsal position similar to a bothriotricum (Fig. 27). Mucro with rachis fused to lat

1 7 ; EF>"1 tr. -..

Figs. 17 and 18. Deuterosminthurus maassius sp. nov. 17. Dorsal distribution of
pigment. 18. Habitus lateral view.

June, 1995

Palacios Vargas & Gonzalez: New Species ofDeuterosminthurus 291

Y 19

21 /

20 '

22 29

24 -/ 25 v 26

23 27K~
'28"^ \ 27

Figs. 19 29. Deuterosminthurus maassius sp. nov. 19. Right eyepatch. 20. Anten
nal segments I to IV. 21. Ant. IV distal portion. 22. Ant. III distal portion. 23. Tibio
tarsus II. 24. Foot complex of leg II. 25. Tibiotarsus III. 26. Foot complex of leg III. 27.
Dens and mucro. 28. Female, anal and genital region, with subanal appendages en
large, lateral view. 29. Head chaetotaxy.

Florida Entomologist 78(2)

eral lamellae in a spoon shape, edges somewhat undulating, rachis forming a thin tip
at the end of the mucro. Anal papillae with normal curving setae; female subanal ap
pendage large and very thick, apically rounded (67 pm long, 7.8 pm wide) (Fig. 28).
Setae of head and body long and thick, some of them appear as spines (Fig. 29). Two
pairs of anal setae are very thick. Maximum size of female (n=8): 1.22 mm. The male
is unknown.


Mexico: Jalisco: Chamela, Estaci6n de Biologia, Instituto de Biologia, Univer
sidad Nacional Aut6noma de Mexico.


Holotype 2 on slide, eight 2 2 paratypes on slides. Two paratypes will be kept at
the Institute of Biology, UNAM and two at Chamela Field Station, the remainder at
Facultad de Ciencias, UNAM. Data of type material: tropical dry forest,40 m altitude,
fogging, 18-VIII 1992, A. Pescador, A. Rodriguez Palafox and J. A. GdmezAnaya colls.


One specimen had four intermediate subdivisions on Ant. IV.


This species is dedicated to Dr. Manuel Maass, Centro de Ecologia, UNAM for his
contributions to tropical dry forest ecology.


Deuterosminthrus maassius sp. nov. is similar to D. tristani Denis, 1933 from
Costa Rica. The new species differs in having no pigmentation on the antennae and a
different color pattern. Head chaetotaxy is similar, with four small setae on the frons;
however, D. maassius lacks tubercles with setae. The subanal appendage of female D.
tristani is bifid, whereas that of D. maassius is thick and apically rounded. Compared
to the drawings of Denis (1933), there seems to be a different ventral dental chaetot
axy. Mucronal edges of the new species are somewhat undulate, while in D. tristani
species they are smooth. Both species share the presence of large stout setae,
spine-like on the head and body; most of the species in the genus have small to mod
erate and thin setae.
We collected soil and litter samples (n=960) during a two-year study at Chamela
and never found D. maassius sp. nov. However, several specimens were caught with
Malaise traps, and also with the aid of a fogger and contact insecticide. Therefore this
species is apparently also associated with epiphytes. In September 1992, we fogged
the canopy in an area covering 100 m2. Among other arthropods we collected more
than 1,013,000 specimens of Collembola. We identified 16 species of Collembola, in
cluding D. maassius sp. nov. However the great majority (98%) belonged to Salina
banks MacGillivray, 1894.


1. Needle of unguiculus not apically knobbed ........................................ ............. 2
N eedle of unguiculus knobbed ................................................ ...................... 3

June, 1995

Palacios Vargas & Gonzalez: New Species ofDeuterosminthurus 293

2. Dens with one subapical ventral seta; setae on head and body thick, spine-like
(Costa Rica)....................................................... .................... D. tristaniDenis
Dens with several subapical ventral setae (3-6); setae on head and body thin
a n d s h o r t ........................................................................ 4
3. Ant. IV with 4-5 interim ediate subsegm ents........................... ......................5
Ant. IV with 7 -15 interim ediate subsegm ents................................................. 10
4. Unguiculus as long as ventral edge of unguis ........................................ 12
Unguiculus short, 1/3 1/2 as long as ventral edge of unguis ...........................6
5. Dens with 5 ventral setae; female appendage with 2 (1-4) teeth
(Cuba)................................................................................. D delatorreisp. nov.
Dens with 6 ventral setae; female appendage without any teeth (USA and
Cuba) ....................... ........ .... .. ... ............... D lippsoni Snider
6. D ens w ith 3 ventral setae ................................................... ...........................7
D ens w ith 5 -6 ven tral setae .............................................................................. 8
7. Dens with 3 very short ventral setae; Ant. IV with 5-7 intermediate subseg
ments (USA) ............................................... ...................D. russatus M aynard
Dens with 3 long ventral setae; Ant. IV with 3 intermediate, poorly defined sub
segments; head and body with thick spine like setae
(M exico) ............................................................................... D m aassius sp. nov.
8. Dens with 5 ventral setae and 6 intermediate subsegments
(Brazil) ............................................................... ..................D. salinensis Arle
D en s w ith 6 ven tral setae ........................................................ ..........................
9. Ant. IV with 6 subsegments; tip of female appendage without any teeth; tibio
tarsus with spines (Brazil)..........................................................D. richardsiArle
Ant. IV with 6-9 subsegments; female appendage spatulate with 7 teeth; with
out spines on tibiotarsus III (USA) .......................................... D.luridaSnider
10. With ciliated setae on body; very small species (0.75 mm)
(USA) ..................................... D. xeromorphus Snider, 1978
With smooth setae on head and body; larger species
(1 .0 1 .3 m m ) ....................................... ....... ........... ................ ... .................. 1 1
11. Dens with 6 ventral setae; unguiculus with a very long tooth; female append
age spatulate with 5 teeth (USA) .................................. D. validentatus Snider
Dens with 5 ventral setae; unguis with small tooth; tip of female appendage
without teeth ...................................................... (USA) D wexfordensis (Snider)
12. Ant. IV with 6 intermediate subsegments (Brazil) ........................ auetiArle
Ant. IV with 14 15 intermediate subsegments
(U SA).......................... ......... .....................................D nonfasciatus Snider
*Dubious records and inaccurately described species are not included in this key.


Specimens of Deuterosminthurus from Cuba were collected by Lic. Magaly Diaz
and Dr. Dania Prieto; those from Chamela were collected by Dr. Alfonso Pescador, M.
Sc. Alicia Rodriguez and Jose A. G6mez Anaya. We are grateful to all of these collab
orators. Proyecto DGAPA IN2078/91 y Convenio de Intercambio Acad6mico
M6xico-Universidad de La Habana.


BANASCO ALMENTERO, J. 1987. Collembola in arable soils of Cuba. Colloq. Pedobiol.,
Soil Fauna and Soil Fertility. Nauka, Moscow, p. 618-621 (in Russian).

294 Florida Entomologist 78(2) June, 1995

DENIS, J. R. 1933. Contribute alla conoscenza del "microgenton" di Costa Rica, III.
Collemboles des Costa Rica avec un contribution aux esp6ce de 1'ordre seconde
note). Boll. Lab. Zool. Gener. Agrar. Fac. Portici, 27:222-322.
PALACIOS-VARGAS, J. G., AND J. A. GOMEZ ANAYA. 1993. Los Collembola (Hexapoda:
Apterygota) de Chamela, Jalisco, Mexico. Folia Entomol. Mex., 89:1 34.
SNIDER, R. J. 1978. New species of Sminthuridae from North America (Collembola:
Symphypleona). Great Lakes Entomol., 11:217-241.


Florida Entomologist 78(2)


Biosciences Research Laboratory
USDA, Agricultural Research Service,
P.O. Box 5674, SU Station
Fargo, ND 58105


A screwworm sample, Cochliomyia hominivorax (Coquerel), collected from Libya
in 1990, during the final phase of the eradication program, was analyzed for mito
chondrial DNA (mtDNA) RFLP patterns. The goal was to determine whether these
fertile flies represented reintroduction from either a new source or contaminants in
the shipments of sterile flies from the Mexican screwworm factory, (used for eradica
tion of flies in Libya) or progeny of the original introduction in 1988. Samples of lab
oratory-strains originating from South America, Mexican factory, Costa Rica, and a
1989 collection from Libya were also analyzed. These patterns were compared with
the previously published patterns of 30 laboratory strains originating from Mexico,
Central America and Jamaica and earlier sample of flies from Libya. The restriction
patterns (with EcoRV, Fnu4HI, HindIII, Hpal, Mspl, Scrfl and SstI) of 1990-Libyan
screwworm were similar (mitochondrial similarity, F=0.97) to those of South Ameri
can flies but different from those of Mexico (F=0.53), Jamaica and Central America
(F=0.71). The Hpal and Sspl digests of DNA amplified by primer flanking mitochon
drial CO-I to CO-II region produced patterns which were similar in Libyan and South
American flies but different from those of Mexican and Costa Rican samples. These
data confirmed that the flies collected in Libya in 1990 did not originate from Mexico,
Central America or Jamaica. Our data suggest some regions) of South America could
be the likely source of screwworm infestation in North Africa. However, it would be
necessary to analyze several geographical samples from South America to confirm
this tentative conclusion.

Key Words: Screwworm, Cochliomyia hominivorax, PCR, mitochondrial DNA RFLP,
geographical variation


Fueron analizados los patrons RFLP de DNA mitocondrial (mtDNA) de una
muestra de gusanos barrenadores del ganado, Cochliomyia hominivorax (Coquerel),
colectada en Libia en 1990 durante la fase final del program de erradicacidn. El ob
jetivo fue determinar si las moscas habian sido reintroducidas de una nueva fuente o

June, 1995

Narang & Degrugillier: Genetic Fingerprinting

si eran contaminants en los envios de moscas est6riles de la fabrica de gusanos ba
rrenadores del ganado de Mexico (usados para la erradicaci6n de las moscas en Libia)
o si eran parte de la progenie de la introducci6n original de 1988. Tambi6n fueron ana
lizadas muestras de cepas de laboratorio originales de America del Sur, de la fabrica
de Mexico, de Costa Rica, y de una colecci6n de Libia. Estos patrons fueron compa
rados con los patrons previamente publicados de 30 cepas de laboratorio de Mexico,
America Central y Jamaica y con muestrtas anteriores de moscas de Libia. Los patro
nes de restricci6n (con EcoRV, Fnu4HI, HindIII, HpaI, MspI, Scrfl y Sstl) de gusanos
de tornillo de Libia de 1990 fueron similares (similitud mitocondrial, F=0.97) a aque
llos de las moscas de America del Sur, pero diferentes de los de Mexico (F=0.53), Ja
maica y Am6rica Central (F=0.71). Los analisis de restricci6n de DNA con Hpal y SspI
amplificados por "primer flanking" a la region mitocondrial CO-I a CO-II originaron
patrons similares a los de las moscas de Libia y America del Sur y diferentes de los
de las muestras de Mexico y Costa Rica. Estos datos confirman que las moscas colec
tadas en Libia en 1990 no se originaron en Mexico, America Central o Jamaica. Nues
tros datos sugieren que alguna region o regions de America del Sur podrian ser la
fuente de la infestaci6n de Africa del Norte. Sin embargo, seria necesario analizar va
raias muestras geograficas de America del Sur para confirmar esta conclusion tenta

The New World screwworm (NWS), Cochliomyia hominivorax (Coquerel)
(Diptera:Calliphoridae), is a major agent of myiasis of livestock and wildlife in the
New World (Knipling & Rainwater 1937). Originally, it was widely distributed from
the southern U.S. to northern Chile. However, it was eradicated in the United States
by August, 1982, and in Mexico in February, 1991, (FAO 1992) by the sterile insect
technique (SIT), as described by Knipling (1955). In March, 1988, NWS were detected
in North Africa (Libya) causing wound infestations in livestock and humans. This
caused great concern that the screwworm might spread to Europe, throughout Tuni
sia and Egypt, and up the Nile into Sub-Saharan Africa. In 1991, the United Nations
Food and Agriculture Organization (FAO) declared eradication of screwworm from
Libya by the sterile insect technique.
The source of the 1988 screwworm infestation in Libya is not known. Taylor et al.
(1991) compared the mitochondrial DNA (mtDNA) restriction pattern of the Libyan
flies with the previously published patterns (Roehrdanz, 1989) of 30 strains original
ing from Mexico, Central America and Jamaica. They concluded that the 1989 collec
tion of Libyan flies had a unique mitochondrial genotype and that the 1988 infestation
in Libya did not originate from Jamaica, Mexico or Central America.
During May, 1991, we received a sample of flies from Libya (collected in November,
1990, near the end of the eradication program). We also received flies from South
America (Brazil) and Costa Rica. We analyzed mitochondrial DNA of these flies to de
termine whether or not: 1) the latest sample from Libya represented a new introduce
tion, 2) PCR-based DNA markers could be used to distinguish Libyan flies from South
American, Central American, Mexican and Jamaican flies, and 3) there were any ge
netic differences between Costa Rican samples collected in 1986 and 1991.


Collection of Samples
The 1990 Libyan sample was from a strain which originated from three egg
masses collected from wounds on sheep near Tripoli, Libya, in November of 1990, and

Florida Entomologist 78(2)

that were maintained in Libya for five generations. A sample of pupae was trans
ported from Libya to Fargo and the adults were used for mtDNA and PCR studies.
Frozen samples of the 1989 Libyan flies (LIB-89) were also analyzed. These flies orig
inated from a laboratory strain established from two eggs masses collected near Tri
poli in October, 1989. Flies were frozen for mtDNA analysis each generation up to 15
generations. The South American sample (BRA-90) was collected in August, 1990, by
exposing a wounded cow to feral flies in Rio de Janeiro, Brazil. About 500 larvae were
collected from the wound, reared to pupae and transported to Fargo. The adults were
combined and frozen samples from the ensuing F5 and F10 generations were used for
DNA analysis. The Costa Rican sample of 1991 (COR-91) was from a composite labo
ratory strain. It originated from eggs laid by twenty-two individual flies collected on
liver. Of these isofemale lines, only sixteen were established successfully in the labo
ratory These were later combined to establish a composite strain. After 2-3 genera
tions of culture in the laboratory, adults were frozen for mtDNA analysis. Flies
originating from the 1986 collection from Costa Rica (COR-86) were analyzed earlier
by Roehrdanz (1986). We compared mtDNA data of COR-91 with that of COR-86. The
OW-87 strain was used as a reference standard for comparison of our results with
those published by Roehrdanz (1989). This strain is currently being mass-produced in
Mexico for use in the sterile insect release program in Central America. The OW-87
was also released in Libya. This strain originated from egg masses collected in the vi
cinity of Orange Walk, Belize, in October, 1986 (refer to Taylor et al. 1991).

Restriction Analysis of mtDNA.

Total cellular DNA from individual flies was isolated by the method of Boyce et al.
(1989) with minor modifications. Individual flies were homogenized in 700 gl of CTAB
buffer (0.1 M tris HC1, pH 8.0, 1.4 M NaCI, 0.02 M EDTA, 2% CTAB -hexadecyl tri
methylammonium bromide, 0.2% 2-mercaptoethanol), incubated at 65'C for 45 min
utes and centrifuged. The supernatant was extracted with an equal volume of
chloroform:isoamyl alcohol and DNA was precipitated by adding 2/3 vol of isopro
panol. The DNA pellet was washed with cold 70% ethanol, vacuum dried and dis
solved in 50 pl 1X TE (10 mM tris, 1 mM EDTA, pH 8.0) containing 10 gg/ml DNAse
free RNAse A.
DNA samples were digested separately with 15 restriction endonucleases and
electrophoresed in 1% agarose gels in TBE buffer. Two molecular size markers, 1 Kb
ladder and lambda HindIII (purchased from BRL) were used in each gel for estima
tion of size of restriction fragments. The DNA gels were blotted on to nylon mem
branes according to Reed & Mann (1985). Purified mtDNA from laboratory reared
screwworm pupae of OW-87 (courtesy of Dr. Roehrdanz) was used to probe Southern
blots. The hybridization mixture contained 32P labeled probe and molecular size mark
ers in 0.5 M phosphate buffer (pH 7.0), 1 mM EDTA, 1% BSA and 7% SDS. After over
night hybridization at 65'C, membranes were washed and exposed to X-ray film. The
mitochondrial similarities (F) among geographical samples, based on the proportion
of identical fragments, were calculated by the method of Nei & Li (1979).

PCR Amplification.

Total DNA prepared from individual adults served as a source of mtDNA template
for the PCR. We used sequences from conserved regions spanning 12SrRNA,
16SrRNA, isoleucine t-RNA, CO-I and CO-II subunit genes of Drosophila (Simon et
al. 1991) and the honeybee (Hall & Smith 1991) as primers (Table 1). These oligonu

June, 1995

Narang & Degrugillier: Genetic Fingerprinting


Set No. Sequence Ref. Fly Sequence

1 I 5' AAA CTA GGA TTA GAT ACC CTA TTA T 3' 12sai (Simon et al. 1991)
4 5' ATG TTT TTG ATAAAC AGG CG 3' 16sa (Simon et al. 1991)
2 > 5'CGC CTG TTT ATC AAAAAC AT 3' 16sar (Simon et al. 1991)
4 5' CTC CGG TTT GAA CTC AGA TC 3' 16sbr (Simon et al. 1991)'
3 I 5' ATT TAC CCT ATC AAG GTAA 3' t-Iso (Simon et al. 1991)
4 5'CGG GCG ATG TGT ACA TAA TT 3' 12sfi (Simon et al. 1991)
4 I 5' ATT TAC CCT ATC AAG GTAA 3' t-Iso (Simon et al. 1991)
4 5'AGG GTA TCT AAT CCT AGT TT 3' 12sair (Simon et al. 1991)
5 > 5' TTG ATT TTT TGG TCA TCC AGA AGT 3' CO-I (Hall & Smith, 1991)
4 5'GAT CAA TAT CAT TGA CC 3' CO-II (Hall & Smith, 1991)

'Modified 16sbr

cleotide primers were synthesized (National Biosciences) and used for amplifications
of DNA segments. The PCR was performed as described by Simon et al. (1991). Each
reaction volume was 100 pl and contained about 0.1 pg of total cellular DNA, 10 Pl
PCR buffer (10X), 150 nmole MgCl2, 20 pmole of each primer, 20 nmole each of dATP,
dCTP, dGTP and dTTP, and 2.5 units Taq polymerase. The reaction profile consisted
of 2 cycles of 95 C for 2 min, 50 C for 2 min, 72 C for 4 min; 32 cycles of 93 C for 1 min,
50 C for 1 min, 72 C for 4 min, and a final cycle of 93 C for 1 min, 50 C for 1 min and
72 C for 10 min.
Aliquots (10 pl) of PCR product were digested with eight enzymes (EcoRV,
Fnu4HI, Hindlll, Hpal, Mspl, Scrfl, SstI, Sspl). The digests along with a molecular
size marker (0X174 HaeIII digest) were electrophoresed in 3.0% agarose and stained
with ethidium bromide. The sizes of fragments were estimated against the molecular
size marker.


Mitochondrial DNA Restriction Patterns

Data in Table 2 show the mitochondrial DNA restriction fragments observed with
15 enzymes. Table 3 and Figs. 1 to 5 summarize the restriction fragment patterns in
LIB-89, LIB-90, BRA-90 and COR-91 strains. The designations for restriction pat
terns correspond to those of Roehrdanz (1989). When our estimates of the sizes of one
or more fragments in a specific pattern differed slightly (500 bp or less) from those re
ported by Roehrdanz (1989), we maintained the same letter designation but indicated
it by parenthesis. Three new patterns, not reported in earlier studies (Roehrdanz
1989, Taylor et al. 1991) are indicated with asterisks (Tables 2 and 3).
The three new patterns included: pattern F for Mspl in LIB-89, LIB-90 and BRA
90 (Figs. 1 & 2); G for HindIII in COR-91 (Fig. 3); and C for Pvull in BRA-90 (not
shown). These three patterns were not reported before in LIB-89 (Taylor et al. 1991)

Florida Entomologist 78(2)


Enzyme Haplotype' Restriction Fragment Sizes (Kb)

EcoRI A 9.0, 3.5, 1.5, 1.2, 1.05
HindIII A 6.8, 4.2, 3.3, 0.52, 0.5, 0.45
D 6.8, 4.0, 3.8, 0.52, 0.5, 0.45
G' 5.5, 3.2, 2.8, 1.3, 0.52, 0.5, 0.45
HaeIII A 7.5, 4.8, 2.0, 1.1, 0.8
B 7.5, 7.0, 1.1, 0.8
MspI A 5.0, 4.9, 4.2, 1.4
B 6.1, 5.0, 4.2
F' 4.9, 4.8, 4.2, 1.4, 0.3, 0.2
EcoRV A 6.6, 5.0, 2.9, 1.5
B 6.6, 5.0, 4.5
PvuII A 7.8, 3.0, 2.9, 0.87, 0.57
C 8.8, 3.0, 2.9, 0.87, 0.57
Hpal A 9.5,6.5
B 16
SstI A2 9.0, 6.0, 0.7
B2 6.4, 5.2, 3.8, 0.7
Xhol A 16
Xbal A 16
PstI A 16
FnudII(BstuI) A 13, 2.7
Fnu4HI A 2.7, 2.6, 2.1, 1.9, 1.8, 1.4, 1.3, 0.8, 0.73, 0.4
C 2.7,2.6, 2.1, 1.9, 1.4, 1.3, 1.2,0.8,0.73,0.6,0.4
Sau96I A 5.8, 4.7, 2.0, 1.6, 1.1, 0.4, 0.3
ScrFI A 5.8, 4.8, 2.6, 1.4, 0.85, 0.75
C 8.2, 4.8, 1.4, 0.85, 0.75

'Indicates that these haplotypes have not been reported in previous publications (Roehrdanz, 1989, Taylor et
al., 1991).
2The haplotypes underlined indicate that fragment sizes are different from those reported in the previous

or in thirty screwworm lines (Roehrdanz 1989). Patterns A and F of Mspl were not dis
tinguishable (Fig. 1) until the gels were run for a relatively longer period of time (Fig.
2). In gels run for normal time (when the 0.6 Kb marker band of the molecular size
marker, lambda HindIII, approaches close to the anodic end of the gel), the two Mspl
fragments, 4.9 and 4.8 Kb of the pattern A appeared as a single band as did the 5.0
and 4.9 Kb fragments of pattern F (Fig. 1). These bands are better resolved in long
run gels as shown in Fig. 2. The other patterns shown in Figs. 4 and 5 include patterns
A and B of SstI, A of Sau96I, Pvull and EcoRI, A and B of Hpal, and A and B of EcoRV.

June, 1995

Narang & Degrugillier: Genetic Fingerprinting


Haplotypes in Screwworm Populations'2

Enzyme LIB-91 LIB-89 BRA-91 OW-87 COR-91 COR-86

HindIII D D D A D,G* A
MspI F* F* F* A A B
Pvull A A A,C* A A A
Hpal B B B A NT A
SstI (B) (B) (B) (A) (A),(B) A
Xhol A A A A A A
Xbal A A A A A A
PstI A A A A A A
FnudII(BstuI) A A A A A A
Fnu4HI C C C A A A
Sau96I A A A A A A
Scrfl C C C A C C

'The asterik (*) indicates that these haplotypes have not been reported in previous publications (Roehrdanz,
1989; Taylor et al., 1991).
2The haplotypes in parenthesis indicate that fragment sizes are different (refer to Table 2) from those re
ported in the previous publications. The data on CR-86 is from Roehrdanz (1989).

The mtDNA restriction patterns in Libyan flies for 7 (47%) enzymes (EcoRV,
Fnu4HI, HindllI, Hpal, Mspl, ScrlI and Sstl) were similar to those of South American
(BRA-90) flies (mitochondrial similarity, F=0.97). These patterns were different from
those of flies from Mexico, Jamaica and Central America. The South American sample
was polymorphic for Pvull sites (patterns A and C), whereas, the Libyan samples
(LIB-89 and LIB-90) were fixed for pattern A. Similarly, COR-91 was polymorphic for
HindlII (pattern D and G) and SstI (patterns A and B). The COR-91 flies differed from
Libyan (F 0.71) and South American flies (F 0.69) with Fnu4HI, HaeIII and Mspl.
There were more differences in restriction endonuclease recognition sites between the
laboratory-reared Mexican strain and the Libyan (F=0.53) and South American
(F 0.52) flies than between the latter two and flies from Costa Rica (COR-91). The
patterns obtained for Bstul, EcoRI, PstI, Pvull, Sau96I, Xbal and Xhol were not di
agnostic for any population.

PCR Amplified DNA Restriction Patterns

The results in Table 4 show the fragments obtained by digestion of amplified DNA
samples. Of the 8 restriction enzymes tested on the DNA samples amplified by primer
#5 (Table 1), flanking region between CO-1 and CO-II, only two (Hpal and Sspl pro
duced diagnostic patterns (Figures 6 to 8). The patterns obtained with Hpal and Sspl
were similar in Libyan and South American flies (Fig. 6). These patterns were differ

Florida Entomologist 78 (2)


Figures 15: 1 & 2) The mtDNA restriction fragment size differences between the
Mspl A pattern of Mexican (OW) and the pattern F of Libyan (LIB) and South Amer
ican (BRA) flies. Fragments, 5.0 and 4.9 bands of the pattern A and 4.9 and 4.8 of the
pattern F can only be distinguished in Fig. 2 (from long-run gel) but not in Fig. 1 (gel
run for normal period of time). Kb ladder (KB) and lambda HindIII digest were used
as molecular size markers; 3) HindllI restriction patterns showing the type A in Mex
ican (OW) the type D in Libyan (LIB) and South American (BRA) and types G and D
in Costa Rica (COR); 4) The patterns A and B of SstI in Costa Rica (COR) and the pat
tern A of Sau96I in Libyan (LIB) flies; 5) Restriction patterns for Pvull, Hpal, EcoRI
and EcoRV.

ent from those obtained in Costa Rican and Mexican samples. Hpal produced 2 frag
ments (0.91 and 0.38 Kb) in Costa Rican and Mexican strains but one uncut fragment
(1.37 Kb) each in Libyan and South American strains. Sspl produced 4 fragments
(0.55, 0.44, 0.15, 0.08 Kb) each in Costa Rican and Mexican strains but 3 fragments
(1.08, 0.15, 0.08 Kb) each in Libyan and South American strains. The patterns ob
stained with six other enzymes were similar in all the 4 samples tested. The digestion

June, 1995

Narang & Degrugillier: Genetic Fingerprinting

(I) () ())
S Ssp I jspl Hpal Hpal

1.35 1.33 1.35
1.08 1.1 .91

.6055 .55 .60
..44 .40

.31 .31

Figures 6 8. Restriction patterns obtained by digestion of DNA samples amplified
by PCR primer pair 5 (see Table 1). Hpal and Sspl patterns of LIB and BRA are dif
ferent from those of OW and COR.
of DNA amplified by primer #1 flanking region between 12SrRNA and 16SrRNA re
gion and primer #2 flanking internal sequences within 16SrRNA gene (Table 1) pro
duced similar patterns for 8 restriction enzymes in flies from the four populations.
The bands of DNA amplified by primers #3 and #4 were either weak or not

Our first goal was to determine whether the LIB-90 sample (collected toward the
end of the screwworm eradication program in Libya) represented a new infestation.
Our data on mtDNA RFLP and the PCR-amplification patterns showed no evidence
for new infestation(s). The mitochondrial genome of LIB-90 was almost identical
(F=0.97) to that of LIB-89. However, occasional reinfestation from the original geo
graphical source cannot be ruled out.
Roehrdanz (1989) identified 16 mitochondrial genotypes in 30 laboratory lines
originating from Mexico, Jamaica and Central America. Although restriction studies
on LIB-89 by Taylor et al. (1991), did not reveal any new patterns, the composite hap
lotype of LIB-89 differed from those 16 reported in Jamaican, Mexican and Central
American flies by Roehrdanz (1989). They designated this new haplotype as 17 and
determined the cladistic relationships by the unrooted dendrogram. Although our re
sults on restriction patterns (of LIB-89 and LIB-90 strains) for Mspl and SstI differed
(due to the improvement in resolution of fragments by our electrophoretic and south
ern blotting methods) from those of Taylor et al. (1991), they do not change the overall
relationship between haplotype 17 and the other 16 haplotypes. The mitochondrial
genome of South American (BRA-90) and LIB-90 flies is almost identical (F 0.97),
therefore, BRA-90 flies can tentatively be designated as haplotype 17.
We also wanted to determine whether some region of South America could be the
source of the original 1988 infestation in Libya. To address this type of question, anal
ysis of mtDNA RFLP is considered to be useful for monitoring population dispersal
and maternal lineages, because the mtDNA is maternally inherited without recombi

Florida Entomologist 78(2)


Restriction Fragment Sizes

Enzyme Primers (refer to Table 1)

1 2 5

BRA 90
BRA 90
BRA 90
BRA 90
BRA 90
BRA 90
BRA 90
BRA 90

.85, 0.5
.85, 0.5
.85, 0.5



.91, .38
.91, .38



1.08, .15, .08
1.08, .15, .08
.55, .44, .15, .08
.55, .44, .15, .08


.59, .48,
.59, .48,
.59, .48,

.29, .21
.29, .21
.29, .21

.71, .26,
.71, .26,
.71, .26,

June, 1995

Narang & Degrugillier: Genetic Fingerprinting

nation (Avise et al. 1987). Paternal contributions to the mtDNA gene pool are rare or
non-existent (Lansman et al. 1983). Although several geographical samples from
South America (we had access to only one strain from Brazil) should be analyzed, a
high level of mitochondrial similarity (F=0.97) between Libyan and Brazilian strains
indicate that some regions) of South America could very well be the source of the 1988
infestation in Libya. Clearly, additional studies are warranted.
The second goal was to find PCR-based DNA marker to distinguish Libyan or
South American flies from those of Mexico, Jamaica and Central American. As shown
in Table 4, total DNA amplified by primer pair 5 (Table 1) followed by digestion with
Hpal or Sspl produced patterns which can be used to distinguish Libyan and South
American strains from those of Mexico, Jamaica and Central America.
The third goal was to determine genetic differences, if any, between the Costa Rica
samples collected in 1986 and 1991. Our results (Table 3) showed that COR-91 was
polymorphic for HindIII and SstIII sites, whereas COR-86 was monomorphic (data
from Roehrdanz 1989). The COR-91 contained two patterns, D and G for HindIII and
A and B for SstI, whereas, COR-86 was fixed for pattern A for HindIII and pattern A
for Sstl. In addition, COR-91 was fixed for pattern A for Mspl, whereas, COR-86 was
fixed for pattern B for Mspl. These differences can be attributed to either livestock or
human-assisted migrations of flies or simply due to sampling error.
Variation in restriction sites in mtDNA within and among populations are com
mon (reviews by Avise & Lansman 1983, Wilson et al. 1985, Avise 1986, Avise et al.
1987, Moritz et al. 1987). Kessler & Avise (1985) reported significant spatial hetero
geneity in the distribution of mtDNA variants within populations (of cotton rats). In
addition, mtDNA length variation within an individual (heteroplasmy) and between
individuals have been reported in other organisms (reviewed by Moritz et al. 1987).
During our studies, except for one individual from COR-91, which produced a mixture
of patterns D and G for HindIII, we did not detect any evidence of heteroplasmy in
populations from South America, Libya, and a Mexican laboratory strain.
The extent of variability observed within and among populations greatly depends
upon the analytical power of the method used. For example, the maternally inherited
mtDNA and paternally inherited Y linked genes generally show less variation within
populations and more between populations than biparently inherited autosomal nu
clear genes (Dowling & Brown 1989). Furthermore, the levels of intra and inter-spe
cies mtDNA RFLP varies greatly among different taxa (Narang et al. 1994).
Therefore, it is desirable to use both multiple techniques as well as feral samples of
populations to obtain more reliable estimates of genetic variability. Our data provide
useful markers to identify geographical populations of screwworm, and should not be
regarded as indicators of level of genetic variability of respective geographical popu
nations or for genetic relationships among them.

Our thanks to Dr. W. Klassen, International Atomic Energy Agency, Vienna, Aus
tria for providing samples of 1990 collection of screwworm from Libya and to Dr. D. B.
Taylor for the Libya-1989, Brazil-1990, Costa Rica-1991 samples and a strain from
the Mexican Screwworm mass production facility. We are grateful to Drs. Sharon
Mitchell, Robert Sparks, Stephen Miller, Glenn Hall and Walter Tabachnick for criti
cally evaluating the manuscript.

AVISE, J. C. 1986. Mitochondrial DNA and the evolutionary genetics of higher ani
mals. Phil. Trans. Roy Soc. London Ser. B312: 325-342.

Florida Entomologist 78(2)

AND N. C. SAUNDERS. 1987. Intraspecific phylogeny: the mitochondrial DNA
bridge between population genetics and systematics. Ann. Rev. Ecol and Syst.
18: 489-522.
AVISE, J. C., AND R. C. LANSMAN. 1983. Polymorphism of mitochondrial DNA in pop
ulations of higher animals, pp. 165-190 in M. Nei and R. K. Koehn [eds.], Evo
lution of Genes and Proteins. Sinauer, Sunderland, MA.
BOYCE, T. M., M. E. ZWICK, AND C. F. AQUADRO. 1989. Mitochondrial DNAin the bark
weevils: size, structure and heteroplasmy. Genetics, 123: 825-836.
DOWLING, T. E., AND W. M. BROWN. 1989. Allozymes, mitochondrial DNA, and levels
of phylogenetic resolutions among four species of minnows (Notoropis:Cyprin
idae). Syst. Zool. 38: 126-143.
DOWLING, T. E., C. MORITZ, AND J. D. PALMER. 1990. Nuclear Acids II. Restriction site
analysis, pp. 250-317 in D. M. Hills and C. Moritz [eds.], Molecular Systemat
ics. Sinauer, Sunderland, MA.
FAO, 1992. The New World Screwworm Eradication Program. North Africa 1988
1992. pp.192. Food and Agriculture Organization of the United Nations.
HALL, G., AND D. R. SMITH. 1991. Distinguishing African and European honeybee
matrilines using amplified mitochondrial DNA. Proc. Natl. Acad. Sci. USA 88:
KESSLER, L. G., AND J. C. AVISE. 1985. Microgeographic lineage analysis by mitochon
drial genotype. Variation in the cotton rat (Sigmodon hispids). Evolution, 39:
KNIPLING, E. F 1955. Possibilities of insect control or eradication through the use of
sexually sterile males. J. Econ. Entomol. 48: 459-462.
KNIPLING, E. F., AND H. T. RAINWATER. 1937. Species and incidence of dipteran larvae
concerned in wound myiasis. J. Econ. Entomol. 23: 451-455.
LANSMAN, R. A., J. C. AVISE, AND M. D. HUETTEL. 1983. Critical experimental test of
the possibility of "paternal leakage" of mitochondrial DNA. Proc. Natl. Acad.
Sci. USA 80: 1969-1971.
MORITZ, C., T. E. DOWLING, AND W. M. BROWN. 1987. Evolution of animal mitochon
drial DNA: Relevance for population Biology and Systematics. Ann. Rev. Ecol.
Syst. 18: 269-292.
NARANG, S. K., W. J. TABACHNICK, AND R. M. FAUST. 1994. Complexities of population
genetic structures. Implications for biological control programs, pp. 19-52 in
Karl Narang, A. C. Bartlett and R. M. Faust [eds.]. Applications of Genetics to
Arthropods of Biological Control Significance. CRC Press, Boca Raton, FL.
NEI, M., AND W. H. LI. 1979. Mathematical model for studying genetic variation in
terms of restriction nuclease. Proc. Natl. Acad. Sci. USA 76: 5269-5273.
REED, K. C., AND D. A. MANN. 1985. Rapid transfer of DNA from agarose gels to nylon
membrane. Nucl. Acid Res. 13: 7207-7221.
ROEHRDANZ, R. L. 1989. Intraspecific genetic variability in mitochondrial DNA of the
screwworm fly (Cochliomyia hominivorax). Biochem. Genet. 27: 551-569.
SIMON, C., A. FRANKE, AND A. MARTIN. 1991. The polymerase chain reaction: DNA ex
traction and amplification, pp. 329 in Hewitt, G. M., A. W. B. Johnson and J. P.
W. Young [eds.], Molecular Taxonomy, Vol. H57. Nato Advance Study Institute,
Springer Verlag, Berlin.
TAYLOR, D. B., L. HAMMACK, AND R. L. ROEHRDANZ. 1991. Reproductive compatibility
and mitochondrial DNA restriction site analysis of New World screwworm, Co
chliomyia hominivorax from North Africa and Central America. Med. Vet. En
tomol. 5: 145-151.
STONEKING. 1985. Mitochondrial DNA and two perspectives on evolutionary
genetics. Biol. J. Linn. Soc. 26: 375-400.

June, 1995

Jones et al.: Termites of Mona Island 305


Southern Forest Experiment Station
P O. Box 2008 GMF
Gulfport, MS 39505

Entomology Department
North Carolina State University
Raleigh, NC 27695

Department of Biology
University of North Carolina
Chapel Hill, NC 27599

International Institute of Tropical Forestry
Call Box 25000
Rio Piedras, PR 00928
Department of Biology
Bayam6n Technological University College
Bayam6n, PR 00959


A survey of eight sites on Mona Island revealed four termite species in the family
Kalotermitidae: Incisitermes nr. bequaerti (Snyder), I. nr. incisus (Silvestri), Neoter
mes mona (Banks), and Procryptotermes corniceps (Snyder). Incisitermes nr. bequaerti
is a new record for the island. Identifiable wood hosts are reported. All species were
found in dead wood, which is typical of drywood termites, but N. mona and P cor
niceps were also collected from live wood. Examination of dead wood in three 9.3 m2
plots in a dense stand of Leucaena leucocephala (Lam.) de Wit. revealed that an aver
age of approximately three-quarters of the wood pieces were attacked by termites. Al
though the total volume of dead wood was almost six times greater in Plot 1 than in
the other two plots, the number of termites per unit volume of dead wood was very
similar (0.4 per cm3). Total numbers of individuals per colony ranged from 11 to 3,359
termites. Caste composition is reported for each colony, and large variations among
colonies were noted. Alates as well as eggs were more common in larger colonies. Sol
dier percentages ranged from 0.7% to 20.5% in 16 P corniceps colonies. The largest
colonies occurred when the greatest volume of dead wood was available.

Key Words: Caste, colony size, subtropical dry forest, West Indies, wood decomposi

1Current address: P.O.Box 58417, Salt Lake City, Utah 84158

Florida Entomologist 78(2)


Un muestreo de ocho sitios en la Isla Mona revel6 cuatro species de termitas de
la familiar Kalotermitidae: Incisitermesnr. bequaerti (Snyder), I. nr. incisus (Silvestri),
Neotermes mona (Banks), y Procryptotermes corniceps (Snyder). Incisitermes nr. be
quaerti es un nuevo registro para la isla. Son reportadas las maderas hospedantes. To
das las species fueron encontradas en madera muerta, lo cual es tipico para las
termitas de madera seca, pero N. mona y P corniceps fueron tambi6n colectados en
madera viva. El examen de la madera muerta en tres parcelas de 9.3m2 en un denso
grupo de Leucaena leucocephala (Lam.) de Wit. revel6 que un promedio de aproxima
damente tres cuartos de los pedazos de madera estaban atacados por las termitas. A
pesar de que el volume total de madera muerta fu6 casi seis veces mayor en la par
cela 1 que en las otras dos, el numero total de termitas por unidad de volume de ma
dera muerta fue muy similar (0.4/cm3). El numero total de individuos por colonia vari6
de 11 a 3359 termitas. La composicidn de las castas es reportada para cada colonia.
Fueron observadas grandes variaciones dentro de las colonies. Los porcentajes de sol
dados variaron de 0.7 a 20.5% en 16 colonies de P corniceps. Las mayores colonies fue
ron encontradas cuando el mayor volume de madera muerta era disponible.

Mona Island is a 55 km2 limestone plateau that rises 60 to 100 m above sea level
in the Mona Passage between the Caribbean islands of Puerto Rico and Hispaniola.
It has no permanent settlements and, since 1973, has been under the administration
of the Puerto Rican Department of Natural Resources, which has emphasized the
value of Mona Island as a wilderness and research site. An overview of the history, ge
ography, and ecology of Mona Island is provided by Cintrdn (1991). The vegetation of
Mona Island, which belongs within the subtropical dry forest life zone, is described in
Cintrdn & Rogers (1991).
A main purpose of this study was to continue a survey of the termites of Mona Is
land begun in the early 1990's by Jones (1991). It is the first since Ramos' (1946) ex
tensive survey of the insects of Mona Island in 1935 and 1944. Three of the four
termite species he collected were identified by A. E. Emerson as Incisitermes snyderi
(Light), Neotermes mona (Banks), and Procryptotermes corniceps (Snyder). The iden
tity of the fourth species was tentatively given as I. incisus (Silvestri), but Emerson
noted that he could not be certain because he had not examined the type specimens.
Incisitermes snyderi is distributed extensively, not only in the West Indies, but also
in Mexico and the southern United States (Scheffrahn et al. 1994, Snyder 1956,
Weesner 1965). Incisitermes incisus is reported from Barbados, Dominica, Guade
loupe, Montserrat, Puerto Rico, and the Virgin Islands (British: Beef, Eustatia,
Guama, Virgin Gorda; U.S.: St Croix) (Snyder 1956, Scheffrahn et al. 1994).
Procryptotermes corniceps, the only Neotropical representative of this genus, is re
ported from many islands in the West Indies (Scheffrahn et al. 1994). It is listed as the
most common kalotermitid in natural vegetation on Providenciales and Grand Turk
Islands in the British West Indies (Scheffrahn et al. 1990). The original description of
N. mona was by Banks (1919) from soldiers collected on Mona Island, which remained
for several decades the species' only reported locality (Ramos 1946, Snyder 1956).
Neotermes mona is now known to be common on Providenciales and Grand Turk Is
lands (Scheffrahn et al. 1990), and it also has been reported from the Dominican Re
public (Hispaniola) and Guana (British Virgin Islands) (Scheffrahn et al. 1994).
The four termite species found on Mona Island are drywood termites belonging to
the primitive family Kalotermitidae. Colonies live within their food sources of dry

June, 1995

Jones et al.: Termites of Mona Island

wood, chiefly dead branches and tree trunks. They do not require contact with the soil
for moisture. Rather, the termites obtain water as a metabolic by-product of cellular
metabolism and also from external water and, possibly, living plant tissue. Colonies
consist of a pair of primary reproductive (king and queen) or replacement reproduc
tives, soldiers, nymphs, pseudergates, larvae, and eggs. Winged imagoes (alates) may
be produced and, if so, they swarm seasonally to establish new colonies. Mature dry
wood colonies generally do not exceed a few thousand individuals (Nutting 1969, Lenz
1994). Because of their habit of living within dry wood and their relatively small col
ony sizes, it is possible to collect entire colonies. This made it feasible to collect entire
colonies on which to base the other major purposes of this study, which were to deter
mine: kalotermitid colony size, caste composition, and the relationship between the
number of termites and available dead wood volume.



Our survey of the termites of Mona Island was conducted from March 24 through
30, 1993. We intensively sampled in the vicinity of eight collection sites including the
airstrip, Carabinero Beach, El Faro, Pajaros Beach, Sardinera Beach, Uvero Beach,
Vereda del Centro, and Vereda India (Fig. 1). These sites were selected because of ease
of access via roads or trails.
Termites were removed from standing or fallen dead wood using hatchets, wood
chisels, saws, and forceps. Groups of termites with representative castes were placed
in vials containing 85% ethanol. We collected soldiers and pseudergates, and alates if
available. The host plant was identified whenever possible using keys and figures in
Little & Wadsworth (1964) and Little et al. (1974); we also relied on local expertise.

67 7'




Fig. 1. March 1993 termite collection sites (*) on Mona Island.


eia Island F

ine Veleda del
nabinero Centro Pajaros

Uvero Ver India



Florida Entomologist 78(2)

Termites were identified to species with the aid of taxonomic keys (Banks 1919, Sny
der 1956). A majority of determinations were verified by R. H. Scheffrahn (University
of Florida) and J. Krecek (University of Florida) and included comparisons with ref
erence specimens. Voucher specimens have been deposited in laboratory collections of
the senior author, as well as with the International Institute of Tropical Forestry in
Rio Piedras, Puerto Rico and with the Puerto Rican Department of Natural Resources
museum on Mona Island.

Volume of Available Dead Wood, Colony Density, and Caste Composition

A site was selected SSW of the airstrip in a large stand of leadtrees [Leucaena leu
cocephala (Lam.) de Wit.]. Live trees in the stand averaged 2.7 cm diam breast height
(n = 32) (Fig. 2). Three points were randomly selected, then used as centers for three
3.05 x 3.05 m plots with borders aligned along compass points. All dead wood >1.2 cm
in diam was collected (standing dead trunks, branches or trunks on the ground, dead
branches on live trees, and dead branches detached but suspended in undergrowth),
except for two dead branches that could not be reached in the crowns of trees in Plot
1 and thus were excluded from all calculations. The length (longest continuous dimen
sion) and midpoint of branches were measured and used to calculate wood volume
(r 2L). Standing dead trunks were measured at breast height. Because live wood was
extremely dense and difficult to split, it generally was not feasible to determine live
wood volume when termite galleries extended into living portions of trees.
Each piece of dead wood was split open and examined for any signs of termite ac
tivity. Termite galleries, termite fecal pellets, and/or body parts were noted as evi
dence of former termite activity. When live termites were observed, every attempt was
made to collect the entire colony into 85% ethanol. Each entire colony collection was

Fig. 2. A dense stand of Leucaena leucocephala along the airstrip on Mona Island

June, 1995

Jones et al.: Termites of Mona Island 309

later examined in the laboratory using dissecting microscopes. All termites were
sorted by categories: 1) primary reproductive, 2) alates and de-alates, 3) soldiers, 4)
pseudergates and nymphs with short wing pads (tip of mesothoracic wing pad not ex
tending beyond thoracic segments), 5) nymphs with long wing pads, and 6) larvae
(first three instars). We were unable to differentiate lightly pigmented replacement
(neotenic) reproductive. Exact counts of each category were made. The presence of
eggs was noted.



Seventy-seven termite samples were collected during the survey. Members of the
family Kalotermitidae were the only termites collected on Mona Island. Four species
were identified: I. nr. bequaerti (Snyder), I. nr. incisus, N. mona, and P corniceps. At
least one collection of each species included alates. Incisitermes nr. bequaerti repre
sents a new record for Mona Island. The other species collections corroborate previous
records for the island (Banks 1919, Ramos 1946, Jones 1991). However, termites re
ported as I. nr. snyderi by Jones (1991) should have been classified as I nr. incisus.
Procryptotermes corniceps was the most common and widely distributed species; it
was found at every site sampled. Our observations suggest that P corniceps more
readily tolerates hot, dry conditions than the other termites on Mona Island. Among
the hosts of P corniceps were dead Coccoloba uvifera (L.) L., Gossypium barbadense
L., Hippomane mancinella L., Tamarindus indica L., dead and live wood of L. leuco
cephala, and others that were unidentifiable.
Our collections suggest that N. mona is neither widely nor commonly distributed
and occurs primarily along the west coast of Mona Island. Of the sites sampled, this
species was found only in the Sardinera Beach area and the airstrip (Fig. 1). Neoter
mes mona was previously collected at Uvero Beach (Ramos 1946, Jones 1991) and the
airstrip (Jones 1991). We recovered N. mona from dead and live wood of L. leucoceph
ala and from dead wood of Bursera simaruba (L.) Sarg. We also identified this termite
species from a sample collected by G. Hernandez (P.R. Dept. Natural Resources) dur
ing April 1993 in Melicoccus bijugatus Jacq. at Sardinera Beach.
Incisitermes nr. incisus was found at all eight sites, and was particularly common
at El Faro. Many colonies included two soldier morphs: long and short-headed forms.
This termite was found in H. mancinella, L. leucocephala, and dead wood that was un
Incisitermes nr. bequaertiwas collected at Carabinero Beach, Uvero Beach, Vereda
del Centro, and Vereda India. It was collected from C. uvifera and dead wood that was
unidentifiable to species.

Volume of Available Dead Wood, Colony Density, and Caste Composition

The total volume of available dead wood in the three plots was 20,671 cm3; 3,242
cm3; and 3,153 cm3, respectively. A large percentage of the wood showed current or
past signs of termite occupation: Plot 1, 74.3%, n = 35; Plot 2, 87.5%, n = 24; Plot 3,
66.7%, n = 21. Overall, slightly more than three-quarters of the available pieces of
dead wood showed evidence of termites. Of these, however, current termite infesta
tions were less common than colonies that had met their demise (Plot 1, 26.9% live;
Plot 2, 28.6% live; Plot 3, 42.9% live).

Florida Entomologist 78(2)

-iC cL cMC
t-~ LO CD 0

-i LO LO
0' 0L
V( -cx

o n

c] cl]


c'Jc'Jb C'J

cct ~mb


,- ^- [>- cc CJ ~'f

c00 -
00 00-


c'Jcc h c

c'J v


t 0

^*f 0 Oib ^ ^ ^CD C
IN -1L mm L

-iM 0iO N C

0 00 CD'-i CM C


muo C o0t>- -I CD LO
-i c'Ji c'J
-ii -c'J

c'J c'J -cc-

-c'j -~

c'j Th- c'J
c'J h -

xO CD x- ,0 M0 M*f D t-*

-c0 o oo c.0
_i m ^-1 ^

c -i Cz) M Ln M -C C

30 c-O C o MD ;LO tc~ M *f-icT
cD) C MD D CD cD c.0 'cD0
-c' Ln p -


oD C

June, 1995



v E




-0~ E~


Jones et al.: Termites of Mona Island 311

Procryptotermes corniceps and I nr. incisus were the only species found in the
sample plots, although N. mona was found outside the plots in the same stand of
leadtrees. Procryptotermes corniceps was by far the predominant termite.
In some cases (n 8) of dead branches on live trees, termite galleries extended into
the living portion of the tree (Table 1). Because this live wood either was not sampled
or was incompletely sampled, totals for these colonies are underestimated. However,
our observations of N. mona and P corniceps in live wood indicated that colonies prob
ably were initiated in an attached dead branch, then extended into the living portion
of the tree as the colony grew. Whereas dead wood was extensively riddled with gal
leries, often just a single termite gallery penetrated into the core of the live wood. Dry
wood termite species that can extend foraging galleries from the originally colonized
dead wood into regions of live wood in a tree attain larger colony sizes than species
that remain solely in the dead wood that initially was invaded (Lenz 1994). Maximum
colony size of P corniceps greatly exceeded that of I nr. incisus, although the latter
species was much less common in the plots (Table 1).
Total termite counts and caste composition for the 20 colonies found in the 3 plots
are reported in Table 1. These data, which indicated that colony size of these two dry
wood termite species did not exceed 3,500 individuals, are consistent with previous
reports for the Kalotermitidae (Nutting 1969, Lenz 1994).
Although the total volume of dead wood was approximately six times greater in
Plot 1 than in the other two plots, the number of termites per unit volume of dead
wood was very similar: 0.37 termites per cm3 in Plot 1, 0.45 termites per cm3 in Plot
2, and 0.42 termites per cm3 in Plot 3. The similarity of these data likely stems from
the fact that average colony size coupled with average available wood volume per col
ony was greater in Plot 1 than in the other two plots. In Plot 1, colonies averaged 1,198
termites, and dead wood volume per colony averaged 2,553 cm3; in Plot 2, colonies av
eraged 243 termites, and wood volume averaged 151 cm3; in Plot 3, colonies averaged
222 termites, and wood volume averaged 176 cm3. These calculations exclude colonies
117 and 118 in Plot 1, because some of these termites were extracted from live wood.
Linear regression indicates a positive relationship between the logl, of dead wood
volume and the logl, of colony size (y= 0.783x+ 0.351) (Fig. 3). With an R of 0.501,
linear regression explains more than half of the variation in the data. Thus, drywood
termite colonies on Mona Island attain larger sizes in larger pieces of dead wood. The
size of mature drywood termite colonies typically is positively correlated with food re
source size (Lenz 1994).
Colony density ranged from 0.6 colonies per m2 in Plots 2 and 3, to 1.2 colonies per
m2 in Plot 1. These high densities probably are representative of the termite popular
tion in the Leucaena stand but not of Mona Island as a whole. Observations during
our survey, however, indicate a relatively high density of drywood termite colonies in
other habitats on the island.
Termite densities in the dry forests of Mona Island were severalfold higher than
those observed in Cyrilla racemiflora L. in the Luquillo tropical montane forest of Pu
erto Rico (Torres 1994). This tends to support the hypothesis that termites are more
important as wood decomposers in dry forests, whereas fungi are more important in
wet forests (Chudnoff & Goytia 1972, Bultman & Southwell 1976). The dry forest life
zone is a significant component in the tropics and subtropics, comprising approxi
mately 42% of forest lands (Murphy & Lugo 1986).
Caste composition varied with colony size (Table 1). All colonies with >992 individ
uals (n = 5) contained eggs as well as alates. Small colonies with both these develop
mental stages could be found, but not consistently; alates and eggs were found in 2
small colonies, one with n = 126 termites and the other with n = 244 termites. The

Florida Entomologist 78(2)


3.5 --------------------------------c---------

O ] [
O p

o 0
1. -

1 D I I I
1 .5 --- - - - - - - - ----- - - - - - - - - - - - - -

1 1.5 2 2.5 3 3.5 4

Fig. 3. Relationship between the log10 of dead wood volume and the log10 of dry
wood termite colony size on Mona Island (y= 0.783x+ 0.351; =- 0.501), (n 18).

smallest colony with alates present was comprised of 126 termites. The 6 smallest col
onies (<71 termites) had neither eggs nor alates and may represent incipient colonies
or those on the decline. We suspect that these colonies were at the end of their colony
cycle because of the absence of primary reproductive and early instars. Only one of
the six smallest colonies contained primary reproductive.
Soldier percentages ranged from 0.7% to 20.5% for the 17 P corniceps colonies and
were 9.8% and 38.5% for the two colonies of I. nr. incisus. Soldier percentages were
more variable for small colonies than for large colonies.
In conclusion, this survey revealed four species of kalotermitids, including I nr.
bequaerti, which is a new record for Mona Island. These termites were found in a va
riety of tree species. Drywood termites apparently are important wood decomposers
in the subtropical dry forest life zone of Mona Island, with an average of approxi
mately three-quarters of dead wood branches showing signs of termite attack. Ap
proximately 0.4 termites per cm3 of dead wood were noted. Colony size of P corniceps
ranged from 11 to 3,359 individuals, with large variations in caste composition among
colonies. Data on wood volume together with total termite counts support the hypoth
esis that kalotermitid colony size is closely attuned to the size of the food resource.


We are grateful to the Commonwealth of Puerto Rico, Department of Natural Re
sources for providing transportation to and lodging on Mona Island and for permis
sion to collect termite specimens. We thank G. M. Hernandez and other Mona Island
personnel for providing transportation to field sites and assistance with tree identify

June, 1995

Jones et al.: Termites of Mona Island 313

cation and other logistics. We appreciate the assistance of R. H. Scheffrahn (Univ.
Florida) and J. Krecek (Univ. Florida) with termite species identification. We thank A.
Dvorak for translating the abstract. We also thank S. L. Buchmann, G. D. Hoffman,
and R. H. Scheffrahn for critically reviewing the manuscript.


BANKS, N. 1919. Antillean Isoptera. Bull. Mus. Comp. Zool. 62: 475-489.
BULTMAN, J. D., AND C. R. SOUTHWELL. 1976. Natural resistance of tropical woods to
terrestrial wood-destroying organisms. Biotropica 8: 71 95.
CHUDNOFF, M., AND E. GOYTIA. 1972. Preservative treatments and service life of fence
posts in Puerto Rico. USDA Forest Service Research Paper ITF-12, 28 pp.
CINTRON, G. 1991. Introduction to Mona Island. Acta Cient. 5: 69.
CINTRON, B., AND L. ROGERS. 1991. Plant communities of Mona Island. Acta Cient. 5:
JONES, S. C. 1991. Termites (Isoptera: Kalotermitidae) of Mona Island: A preliminary
report. Acta Cient. 5: 7375.
LENZ, M. 1994. Food resources, colony growth and caste development in wood-feeding
termites, pp. 159-209 in J. H. Hunt & C. A. Nalepa [eds.], Nourishment and
evolution in insect societies. Westview Press, Boulder, CO.
LITTLE, E. L., JR., AND F. H. WADSWORTH. 1964. Common trees of Puerto Rico and the
Virgin Islands. USDA Forest Service, Agric. Handbook 249, 556 pp.
LITTLE, E. L., JR., R. O. WOODBURY, AND F. H. WADSWORTH. 1974. Trees of Puerto
Rico and the Virgin Islands, Second Volume. USDA Forest Service, Agric.
Handbook 449, 1024 pp.
MURPHY, P. G., AND A. E. LUGO. 1986. Ecology of tropical dry forests. Ann. Rev. Ecol.
Syst. 17: 67-88.
NUTTING, W. L. 1969. Flight and colony foundation, pp. 233-282 in K. Krishna and F
M. Weesner [eds.], Biology of termites, Volume 1. Academic Press, New York.
RAMOS, J. A. 1946. The insects of Mona Island (West Indies). J. Agric. Univ. Puerto
Rico 30: 174.
1994. Termites (Isoptera: Kalotermitidae, Rhinotermitidae, Termitidae) of the
West Indies. Sociobiology 24: 213-238.
SCHEFFRAHN, R. H., N-Y. SU, AND B. DIEHL. 1990. Native, introduced, and structure
infesting termites of the Turks and Caicos Islands, B.W.I. (Isoptera: Kaloter
mitidae, Rhinotermitidae, Termitidae). Florida Entomol. 73: 622-627.
SNYDER, T E. 1956. Termites of the West Indies, the Bahamas and Bermuda
(Isoptera). J. Agric. Univ. Puerto Rico 40: 189-202.
TORRES, J. A. 1994. Wood decomposition of Cyrilla racemiflora in a tropical montane
forest. Biotropica 26: 124-140.
WEESNER, F. M. 1965. The termites of the United States, a handbook. Natl. Pest Con
trol Assoc., Elizabeth, NJ.

Florida Entomologist 78(2)


'Dept. Ecology and Evolutionary Biology
University of Connecticut
Storrs, Ct. 06269

2University of Florida
Institute of Food and Agricultural Sciences
Tropical Research and Education Center
Homestead, FL 33031


The discovery of a species of Ozophora not previously reported for Florida led to a
re-examination of the Ozophora atropicta complex. This examination showed that 0.
heydoni had been incorrectly synonymized and is resurrected from synonymy and re
ported from Florida for the first time. A new species, Ozophora atropictoides, is de
scribed from Trinidad.

Key Words: Ozophora, Florida, synonymy, atropictoides.


El descubrimiento de una especie de Ozophora no previamente reportada de la
Florida condujo a un reexamen del complejo de Ozophora atropicta. Este examen mos
tr6 que 0. heydoni habia sido incorrectamente sinonimizada y es resucitada de la si
nonimia y reportada de la Florida por primera vez. Una nueva especie, Ozophora
atropictoides, es descrita de Trinidad.

The discovery of a species of Ozophora not previously reported for Florida led to re
examination of the Ozophora atropicta complex. Recently the junior author and Mrs.
Holly Glenn, Biological Scientist II, Tropical Research and Education Center, col
elected a number of specimens of a species of Ozophora on the grounds of the Tropical
Research and Education Center (TREC) under a large Ficus tree. This Ozophora was
determined to be a species not known from Florida, either 0. atropicta or a closely re
lated species. A comparison of the male genital capsule, parameres and cuplike scler
ite (Schaefer, 1977) of 0. atropicta from the Dominican Republic with those of
specimens collected at TREC showed distinct differences.
Slater & Hassey (1981) discussed the status of 0. atropicta Barber in detail. They
pointed out that the type series was mixed. Material not conspecific with the holotype
was subsequently described as O. levis by Slater & Baranowski (1983). Slater & Has
sey ibidd) also synonymized 0. heydoni Barber & Ashlock with 0. atropicta. They dis
cussed variation in, and distribution of 0. atropicta, hypothesizing the origin and
dispersal of West Indian populations from a mainland source area.

June, 1995

Slater & Baranowski: Ozophora heydoni in Florida

It is now apparent that Slater & Hassey (1981) were incorrect in synonymizing 0.
heydoni and 0. atropicta and in believing that the genitalia do not differ significantly.
These taxa, which we now believe is a complex of species, are very similar externally,
differing externally chiefly by characters discussed by Slater & Hassey in their dis
cussion of "variation" between Bahamian and Greater Antillean populations. How
ever, we now find the shape of the genital capsule to be consistently different. This
structure has proven to be diagnostic for many closely related species of Ozophora.
Furthermore, mainland populations of what Slater & Hassey believed to be 0. atrop
icta also have a distinctly different genital capsule.

Three distinct species are thus involved:

1. Ozophora heydoni Barber & Ashlock NEW STATUS, occurs in extreme southern
Florida (Homestead, Fla. I 31-1991 R. M. Baranowski, light trap and under Ficus)
and probably throughout the Bahamas. The holotype is from New Providence,
paratypes from Abaco Cays, Great Abaco I. and Andros I. Slater & Hassey (1981) also
report it (as 0. atropicta) from Eleuthera Is. (Current Cut, Powell Pt.), Arthur's Town
Cat Is.
In 0. heydoni the posterior margin of the genital capsule has a slight mesal pro
trusion when viewed dorsally (Fig. 3) and in lateral view this protrusion can be seen
as a caudal projection of the capsule (Fig. 2). The arms of the cuplike sclerite are
widely separated and evenly conical with heavily sclerotized distal ends (Fig. 1).
The paramere of 0. heydoni (Fig. 11) has a large toothlike inner projection that ex
tends over the base of the outer projection. There is no minute secondary tooth.
Ozophora heydoni keys to couplet four in Slater & Baranowski (1983, 1990). It can
be separated from 0. levis by the pale apex of the membrane, and the variegated hem
elytra. In some specimens, the pale membrane apex is obscure. In such cases, 0. hey
doni may be distinguished from O. levis by having the lateral corial margins entirely
pale and in contact with the pale subapical macula, whereas in 0. heydoni the lateral
corial margins are separated from the subapical pale macula by a large dark trans
verse band. In O. levis, the posterior pronotal lobe is, for the most part, concolorous
with the anterior lobe, whereas in 0. heydonithe posterior lobe is sometimes pale and
contrasts markedly with the dark anterior lobe. Such specimens of 0. heydoni key to
couplet seven where they can be separated from 0. carol by lacking a yellow scalloped
posterior pronotal margin and from 0. floridana by having one or more dark bars on
the clavus and anterior one-half of the corium.
2. Ozophora atropicta Barber as here restricted appears to be confined to the
Greater Antilles. Slater & Hassey (1981) reported numerous records from Cuba (mis
spelled as Cubra in text) and the Dominican Republic. The holotype is from Puerto
In nominal 0. atropicta, the posterior margin of the genital capsule is conspicu
ously produced (Fig. 6), resulting from a strong, almost fingerlike projection backward
of the posterior margin of the capsule (Fig. 5). The arms of the cuplike sclerite are di
vergent, but almost in contact along the midline at their inner angles and are strongly
bent and heavily sclerotized distally (Fig. 4).
The paramere (Fig. 10) has a much smaller, less toothlike inner projection that
does not extend over the base of the outer projection, and also has a minute secondary
3. Ozophora atropictoides New Species. The restriction of 0. atropicta to the West
Indies and 0. heydonito the Bahamas and extreme southern Florida leaves mainland
populations that have previously been referred to 0. atropicta without a name; these
are described below as a new species. We have examined the genital capsules of spec

Florida Entomologist 78(2)



i ''I

Figures 1-9. Male genital capsule showing cup-like sclerite. Posterior view. Fig. 1
0. heydoni; Fig. 4 0. atropicta; Fig. 7 0. atropictoides; Male genital capsule lateral
view. Fig. 2 0. heydoni; Fig. 5 0. atropicta; Fig. 8 0. atropictoides; Male genital cap
sule dorsal view. Fig. 3 0. heydoni; Fig. 6 0. atropicta; Fig. 9 0. atropictoides.
imens from Mexico, Central America, and Trinidad and all show characteristics de
scribed below.
In 0. atropictoides, the posterior margin of the genital capsule lacks a caudal pro
section (Figs. 7, 9) and, in lateral view, the capsule can be seen to slope evenly poste
riorly from the ventral to dorsal margins and thus has a less noticeable backward
protrusion dorsally (Fig. 8). The arms of the cuplike sclerite are strongly divergent
and not heavily sclerotized at their distal ends (Fig. 7).
Slater & Hassey's (1981) discussion of the dispersal vs. vicariance hypothesis for
the populations of these species still has relevance despite their mistaken view that a
single species was involved. It suggests, however, that a longer time period was in
volved and makes the absence of the complex from Jamaica less puzzling.

Ozophora atropictoides Slater & Baranowski, New Species
Head and anterior pronotal lobe dark red-brown. Anterior and lateral pronotal
margins and entire posterior pronotal lobe strongly and contrastingly pale yellow.
Scutellum chocolate brown with two divergent yellow vittae on distal one-third, these
not attaining base of scutellum. Hemelytra chiefly pale yellow, with large dark mac
ula present laterad of radial vein with anterior margin at level of distal end of claval

June, 1995




Fr' I-

:;j Lg

Slater & Baranowski: Ozophora heydoni in Florida

10 1

Figures 10-11. Parameres lateral view: Fig. 10 0. atropicta: Fig. 11 0. heydoni.
commissure, second large macula at apex of corium. Membrane fumose. Legs, first
and second antennal segments pale yellow. Third antennal segment also yellow, but
darkened near distal end. Fourth segment fuscous with conspicuous large white sub
basal annulus on proximal third. Punctures dark brown, small and well separated
from one another. Body nearly glabrous above (few very short inconspicuous hairs
present when viewed laterally).
Head slightly declivent anteriorly, reaching over basal one-third of first antenna
segment. Eyes very large, sessile, occupying most of lateral head surface. Length head
0.90, width 1.00, interocular space 0.30. Lateral pronotal margins carinate; calli gran
ulose, well separated from one another mesally, very sparsely punctate. Length
pronotum 1.02, width 1.56. Length scutellum 1.04, width 0.88. Lateral corial margins
evenly but shallowly concave. Length claval commissure 0.80. Midline distance apex
clavus-apex corium 1.32. Midline distance apex corium-apex abdomen 0.92. Metatho
racic scent gland auricle short, bent slightly caudolaterally; evaporative area occupy
ing inner 2/3 of anterior lobe of metapleuron, outer margin straight. Forefemora
moderately incrassate, armed below with three major spines followed proximally by
4-5 hair spines. Labium extending posteriorly well between metacoxae, first segment
attaining or slightly exceeding base of head. Length labial segments I 0.90, II 0.94, III
0.64, IV 0.40. Antennae elongate, terete, fourth segment narrowly fusiform. Length
antenna segments I 0.66, II 1.80, III 1.40, IV 1.52. Total body length 5.88.

Male genital capsule as in figures 7-9.
TYPES. Holotype. Male. TRINIDAD: Simla, Arima-Blanchisseuse Rd. 600 ft. VII
20-1975 (J. Price) blacklightt trap). In American Museum of Natural History.
Paratypes: TRINIDAD: 15 males, 15 females same data as holotype. 1 male same ex

Florida Entomologist 78(2)

cept VII-14-1975. 1 female Simla Arima Valley II-4-1965 (J. A. Slater & N. T Davis. 1
male, 1 female St. Augustine VI-14-1973 (R. Baranowski, F. O'Rourke, V. Picchi, J.
Slater) (light trap). In National Museum of Natural History (USNM), R. M. Bara
nowski and J. A. Slater collections.
Although the posterior lobe of the pronotum in the holotype is entirely pale, and
the hemelytra chiefly so, this is not true of most of the type series. This species ap
pears to be sexually dimorphic. Females, in addition to being larger, are usually very
dark with the hemelytra predominately dark chocolate brown. Many males also are
much darker than the holotype. Specimens of both sexes usually have a dark stripe on
the meson of the posterior pronotal lobe. They frequently have additional dark strip
ing on the posterior pronotal lobe and have the anterior third of the corium with a
dark macula. The darker coloration is found in most Central American specimens as
Specimens listed by Slater & Hassey (1981) from Mexico, Honduras, Costa Rica,
Panama and Venezuela have been reexamined and appear to be 0. atropictoides. The
Brazilian material listed by Slater & Hassey also appears to represent this species
with the exception of the male and female from "Corupa (Hans Humbolt) S. Cat. XI
1944" which we believe represents neither 0. atropictoides nor 0. atropicta.
ETYMOLOGY. Referring to a similarity to 0. atropicta.


Florida Agricultural Experiment Station Journal Series No. R-03607.


SCHAEFER, C. W. 1977. Genital capsule of the Trichophora male (Hemiptera: Het
eroptera: Geocorisae). Int. Jour. of Morph. and Embryol. 6: 277-301, Table 1.
SLATER, J. A., AND R. M. BARANOWSKI. 1983. The genus Ozophora in Florida (Hemi
ptera:Heteroptera:Lygaeidae). Florida Entomol. 66: 416-440.
SLATER, J. A., AND R. M. BARANOWSKI. 1990. Lygaeidae of Florida. (Hemiptera: Het
eroptera). Arthropods of Florida and Neighboring Land Areas. 14:1-211. Flor
ida Dept. of Agr. and Consumer Serv. Div. of Plant Ind. Gainesville.
SLATER, J. A., AND M. HASSEY. 1981. The distribution and systematics of Ozophora at
ropicta Barber, with the description of a new species from the Neotropics. Flor
ida Entomol. 64: 246249.

June, 1995

Sanborn et al.: Endothermic Cicadas


1Barry University, School of Natural and Health Sciences,
11300 N.E. Second Avenue,
Miami Shores, FL 33161, U.S.A.

2Department of Physiology, University of Illinois,
524 Burrill Hall, 407 South Goodwin Avenue,
Urbana, IL 61801, U.S.A.

3Department of Biochemistry, University of Arizona,
Biol. Sci. W440,
Tucson, AZ 85721, U.S.A.


Proarna bergi (Distant) and Proarna insignis Distant use metabolic heat to raise
body temperature (Tb) for activity when ambient conditions would prevent activity in
ectothermic animals. Both species were observed singing during overcast or rainy
conditions and at dusk. TbS in the field exceeded ambient by as much as 7.4'C when
solar radiation was unavailable to the insects. In the laboratory voluntary metabolic
heat production raised Tb as much as 12.3'C and 10.7'C above ambient in P bergi and
P insignis respectively. Estimates of metabolic rate from heating and cooling curves
were 0.118 ml 02 per min for P bergi and 0.126 ml 02 per min for P insignis. Fine
shiver-like movements of the thoracic musculature produced the heat. The Tb at
which endogenous warm-up voluntarily stopped in the laboratory was similar to the
Tbs measured in active animals in the field. Thermal responses measured in the lab
oratory also illustrate these animals are thermoregulating with endogenous heat. En
dogenous heat production uncoupled reproductive behavior from environmental

Key Words: Endothermy, thermoregulation, temperature, shivering, cicadas, Cica
didae, Proarna


Proarna bergi (Distant) y Proarna insignis Distant utilizan calor metab6lico para
elevar la temperature del cuerpo (Tb) y ser activos cuando las condiciones ambientales
impiden la actividad en otros animals ectot6rmicos. Ambas species fueron observa
das cantando bajo condiciones de cielo completamente nublado, bajo la lluvia y en la
obscuridad. Las Tb en el campo excedieron las del ambiente en 7.4'C cuando la radia
ci6n solar no era disponible a los insects. En el laboratorio la producci6n voluntaria
de calor aument6 la Tb en 12.3'C y 10.7'C por encima del ambiente en P bergi y P in-
signis respectivamente. Los estimados de la tasa metab6lica de las curvas de calenta
miento y enfriamiento fueron de 0.118 ml de 02 por minute para P bergi y 0.126 ml
de 02 por minute para P insignis. Finos movimientos en forma de temblor de la mus
culatura toracica produce el calor. La Tb a la cual el calentamiento end6geno volun
tariamente ces6 en el laboratorio fue similar a la Tb media en animals activos en el
campo. Las respuestas t6rmicas medidas en el laboratorio ademas ilustran que estos
animals son termorreguladores con calor end6geno. La producci6n de calor end6geno

Florida Entomologist 78(2)

posibilit6 el desarrollo de la conduct reproductive independientemente de las condi
ciones ambientales.

Cicadas must maintain their body temperature within a small range to coordinate
reproductive activity (Heath 1967; Heath 1972). Thermoregulation is generally ac
complished by altering the uptake of solar radiation (Heath 1967; Heath & Wilkin
1970; Heath et al. 1972; Sanborn et al. 1992). Endothermy was first described in cica
das by Bartholomew & Barnhart (1984) and has recently been described as a mecha
nism of thermoregulation in cicadas (Sanborn et al., in press).
This paper describes and quantifies an additional mechanism of endogenous heat
production and its utilization by two grass dwelling cicadas in northern Argentina.



Proarna bergi (Distant) and Proarna insignis Distant are medium-sized cicadas.
Live weight averages 342 12 (n=41) and 434 + 14 (n=21) milligrams respectively. Al
though the data was collected mainly from males, both sexes were used in field and
laboratory measurements. Animals captured for laboratory experimentation were
placed in a cardboard container along with plant samples and a wet paper towel. The
containers were kept on ice until the experiments could be performed during the af
ternoon or evening of the day of capture. Live weights were measured on a
Cent-O-Gram triple beam balance (Model CG 311, Ohaus Scale Corporation) accurate
to 5 mg. All statistics are presented as mean + standard error. Voucher specimens
are deposited in the collection of the Museo Nacional de La Plata, La Plata, Argentina.

Temperature Measurements

Equipment. Body temperature (Tb) of the cicadas was measured with a Sensortek
Model BAT 12 digital thermocouple thermometer with a type MT 29/1 copper/con
stantan hypodermic microprobe (accurate to 0.1 C) which had been calibrated with
a National Bureau of Standards mercury thermometer. The probes were inserted dor
sally midway into the mesothorax of each cicada to measure Tb. All TbS were mea
sured within five seconds of the insect being captured or the insect performing the
behavior under study.
Laboratory Measurements. Laboratory experiments were performed to determine
the range of temperature in which the cicadas can be fully active. The procedures used
to determine the thermal responses were described in previous cicada studies (Heath
1967; Heath & Wilkin 1970).
Cooling curves of P bergi and P insignis were measured from tethered cicadas. A
copper/constantan thermocouple wire (30 gauge) was implanted into and secured to
the dorsal mesothorax of the cicadas to measure changes in Tb. The tethered animal
was permitted to fly to increase Tb above ambient. The animal was placed into a sty
rofoam box immediately after the flight terminated. The box served as a controlled ra
diant environment and prevented forced convective heat loss. Tb was recorded with
the BAT 12 every 15 sec as the animal cooled until Tb remained relatively constant (10

June, 1995

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