Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00043
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 1998
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: VID00043
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Full Text

Brailovsky:A Revision of the Genus Amblyomia


Institute de Biologia, UNAM, Departamento de Zoologia, Apdo Postal 70153 M6xico
04510 D.F. M6xico


The genusAmblyomia Stal is revised and two new species,A. foreroi andA. prome-
ceops from Colombia, are described. New host plant and distributional records of A.
bifasciata Stal are given; habitus illustrations and drawings of male and female gen-
italia are included as well as a key to the known species. The group feeds on bromeli-

Key Words: Insecta, Heteroptera, Coreidae, Leptoscelini,Amblyomia, Bromeliaceae


El g6neroAmblyomia Stal es revisado y dos nuevas especies,A. foreroi yA. prome-
ceops, recolectadas en Colombia, son descritas. Plantas hospederas y nuevas local-
idades paraA. bifasciata Stal son incluidas; se ofrece una clave para la separaci6n de
las species conocidas, las cuales son ilustradas incluyendo los genitales de ambos
sexos. Las preferencias tr6ficas del grupo estan orientadas hacia bromelias.

Palabras clave: Insecta, Heteroptera, Coreidae, Leptoscelini, Amblyomia, Bromeli-

The neotropical genus Amblyomia Stal was previously known from a single Mexi-
can species,A. bifasciata Stal 1870. In the present paper the genus is redefined to in-
clude two new species collected in Colombia. This genus apparently is restricted to
feeding on members of the Bromeliaceae, and specimens were collected on the heart
ofAnanas comosus andAechmea bracteata.
All measurements are in millimeters.


Amblyomia Stal 1870: 171.
Redescription. Head longer than wide, elongate, pentagonal, non-declivent, and
produced forward between bases of antennae; tylus blunt, forming rounded elevated
ridge, slightly projecting beyond juga;juga unarmed, thickened; mandibular plate un-
armed; antenniferous tubercles unarmed, widely separated, space between them
slightly more than two times the width of one tubercle; antennal segment I shorter
than head, thicker, slightly curving; segments II and III cylindrical, IV fusiform; seg-
ment IV longest, segment I shortest, II longer than III; preocellar pit deep, nearly cir-
cular; ocelli elevated; eyes hemispherical, prominent; area between eyes convex;
postocular tubercle low, almost absent; buccula short, unarmed, not extending beyond
antenniferous tubercles; rostrum reaching posterior border of abdominal sternite III;

Florida Entomologist 81(4)

December, 1998

rostral segment I longest, reaching base of head, segment III shortest, and II longer
than IV; neck short. Thorax. Pronotum. Wider than long, trapeziform, gradually de-
clivent; collar wide; anterolateral borders obliquely rounded and entire; frontal and
humeral angles rounded, not exposed; posterolateral and posterior borders straight,
entire; disc deeply punctate except for smooth callar region. Ventrally smooth, except
acetabula, anterior and posterior margin of propleura, posterior margin of me-
sopleura and metapleura, deeply punctate; prosternum with deep excavation; mesos-
ternum shallowly sulcate; metasternum flat; anterior lobe of metathoracic peritreme
elevated, reniform, posterior lobe small, acute. Legs. Femora ventrally armed with
two rows of spines, dorsally with scattered spines or low tubercles; hind femora in-
crassate, moderately in females, strongly in males; hind tibiae shorter than hind fem-
ora, sulcate, triquetrous in cross section, and armed on distal third with short spines
on low tubercles, conspicuous in males, hard to see in females. Scutellum. Triangular,
flat, wider than long; apex subacute; disc punctate. Hemelytra. Macropterous, extend-
ing far beyond apex of last abdominal segment; costal margin emarginate, apical mar-
gin weakly sinuate; clavus and corium deeply punctate. Abdomen. Connexival
segments higher than margin of hemelytron at rest; posterior angles of connexival
segment complete, not extending on a short spine; abdominal spiracle submarginal,
close to middle third; sternum smooth, without punctures. Male genitalia. Genital
capsule simple; posteroventral margin with a shallow median notch (Fig.1).
Parameres. Shaft robust; anterior lobe convex, posterior lobe elongate, slender, and
nearly perpendicular to shaft (Fig.2). Female genitalia. Spermatheca. Distal bulb
oval; sclerotized duct leading from bulb moderately coiled; proximal duct slightly wid-
ened near distal flange; distal duct membranous, narrowed (Fig.3).
This genus is related to Coribergia Casini (1984) and Dalmatomammurius
Brailovsky (1982), but differs in a number of characters: posttylar sulcus absent, an-
tennal segment I much shorter than head, humeral pronotal angles rounded, and hind
tibiae slightly expanded. In the other two genera the posttylar sulcus is present, an-
tennal segment I longer, humeral pronotal angles acute to subacute, and hind tibiae
cylindrical and sulcate. In Dalmatomammurius the antenniferous tubercles are exter-
nally lobulate; they are truncated and unarmed in Amblyomia Stal and Coribergia.
The suprageneric position of these three genera is complex. "Packauskas (personal
communication) believes that the genus Amblyomia deserves its own tribe. He also be-
lieves, based on aedeagal characters and lack of a posttylar sulcus, that its affinities may
be closer to members of the tribe Nematopodini or even the subfamily Meropachydinae.


1. Buccula and rostral segment I bright orange ....... promeceops new species
1'. Buccula black to reddish brown; rostral segment I black ................. .2
2. (l1).Acetabula black; pronotal disc with a wide orange transverse fascia; poste-
rior margin of pronotum black; pronotal collar white to yellow; corium with wide yel-
low to orange transverse fascia (Fig. 4) ....................... bifasciata Stal
2'. Acetabula orange; pronotal disc, including posterior margin yellow; pronotal col-
lar black; corium without orange or yellow transverse fascia ..... foreroi new species

Amblyomia bifasciata Stal
Figs. 1-4

Amblyomia bifasciata Stal 1870: 172
Redescription. Body including antennal segments (apex of IV pale brown to or-
ange), rostral segments, hemelytral membrane, connexival segments, abdominal seg-

Brailovsky:A Revision of the Genus Amblyomia

Figs. 1-3. Amblyomia bifasciata Stal. 1, Caudal view of male genital capsule. 2,
Paramere. 3, Spermatheca.

ments, and legs black; head with short orange stripe below eye and external to ocelli;
pronotum with white to yellow collar, and posterior portion of pronotal disc with a
wide orange transverse fascia; corium with yellow transverse fascia almost straight;
posterior margin of abdominal sterna IV to VI without or with orange irregular spots
lateral to middle third.

Florida Entomologist 81(4)

December, 1998

Measurements. 6 first, then 9. Head length 1.76, 2.04; width across eyes 1.52,
1.80; interocular space 0.84, 1.00; interocellar space 0.52, 0.48; preocellar distance
0.98, 1.14; length antennal segments: I, 1.00, 1.16; II, 1.80, 2.04; III, 1.24, 1.48; IV,
2.76, 3.04. Pronotum. Total length 1.84, 2.44; width across frontal angles 1.20, 1.40;
width across humeral angles 3.16, 3.88. Scutellar length 1.28, 1.56; width 1.44, 1.64.
Total body length 10.68, 12.97.
Biology. Nothing has been known of its biology. Numerous adults and late-instar
nymphs were taken on the heart of pineapple (Ananas comosus) (Bromeliaceae) in the
State of Chiapas (Municipio de Ocosingo, Santo Domingo, M6xico), and a few adults
on the heart ofAechmea bracteata (Bromeliaceae) in the State of Veracruz (Los Tuxt-
las, M6xico).
Distribution. This species was originally described from M6xico, without data.
New Records. M6xico: 2 9, Veracruz, Los Tuxtlas, 22-VII-1968 (Carlos Beutel-
spacher); 1 6, 2 9, same locality, 22-V-1969 (Carlos Beutelspacher). 1 9, Tabasco,
Cardenas, 4-VIII-1970 (R. Arias); 1 9, Chiapas, Bonampak, 20-V-1980 (Harry
Brailovsky); 1 6, 1 9, Chiapas, Simojovel, 25-XII-1968 (Carlos Beutelspacher); 2 6, 2
9, Chiapas, Rio Santo Domingo, 16-1-1983 (Ernesto Barrera); 10 6, 10 9, Chiapas,
Municipio de Ocosingo, Santo Domingo, 24-26- VIII-1893 (Vicente Hernandez); 1l,
Chiapas, Palenque, 25-XII-1990 (M.J. and C.A. Tauber).
It may be distinguished from the new species by possession of a white to yellow col-
lar, the orange transverse fascia crossing the posterior lobe of the pronotum, and by
the yellow transverse fascia of the corium.

Amblyomia foreroi Brailovsky, New Species
Fig. 5

Description. Head black with short orange stripe below eye and external to ocelli;
antennal segments I to III bright red brown, IV black with apex pale orange yellow;
pronotum yellow with collar, calli, and anterolateral margins (except humeral angles)
black; scutellum, clavus, corium, and hemelytral membrane black, with following ar-
eas dull orange: apex of scutellum, irregular stripe on clavus, and two small spots
near middle third of corium; connexival segments III to VI dark brown, with posterior
margin orange; segment VII dark brown; abdominal segments III to VI orange, VII
dark brown. Ventral coloration. Head black, with middle third dark hazel; rostral seg-
ments I to IV red brown; thorax black to red brown, with acetabula and posterior mar-
gin yellow; anterior and posterior lobe of metathoracic peritreme dark hazel;
evaporative area dull black; legs dark hazel; abdominal sterna orange hazel, with pos-
terior margin of sterna IV to VI yellow; genital capsule dark brown.
Measurements. Head length 1.88, width across eyes 1.62, interocular space 0.96,
interocellar space 0.47, preocular distance 1.12; length of antennal segments: I, 1.36,
II, 2.20, III, 1.64, IV, 3.08. Pronotum: Total length 2.32, width across frontal angles
1.40, width across humeral angles 3.64. Scutellar length 1.44, width 1.52. Total body
length 12.28.
Holotype: 6 Colombia: Municipio Risaralda Pueblo Rico, Santa Cecilia, II-1992 (F.
Fernandez). In Universidad Nacional de Colombia, Santa Fe de Bogota (Instituto de
Investigaciones de Recursos Biologicos Alexander von Humboldt).

Etymology: Named for Dimitri Forero.

Amblyomia foreroi is readily distinguishable because it is the only known species
in the genus with the pronotum yellow, except the collar, calli, and anterolateral mar-

Brailovsky:A Revision of the Genus Amblyomia

* :7

Fig 4. Amblyomia bifasciata Stal, dorsal view.

Florida Entomologist 81(4)

December, 1998

Fig. 5.Amblyomia foreroi Brailovsky, New Species, dorsal view.

gins (except humeral angles) black, and corium lacking yellow or orange transverse
fascia. Amblyomia bifasciata Stal, the most closely related species, has the pronotum
black, with the collar white to yellow, and a wide orange transverse fascia over the
pronotal disc, and the corium is black with the yellow transverse fascia almost



Brailovsky:A Revision of the Genus Amblyomia

as W


Fig. 6.Amblyomia promeceops Brailovsky, New Species, dorsal view.

Florida Entomologist 81(4)

December, 1998

Amblyomia promeceops Brailovsky, New Species
FIG. 6

Description. Dorsal coloration. Head black with short orange stripe below eye
and external to ocelli; antennal segments I to IV black (apex of IV pale orange); prono-
tum black with following areas orange: collar, humeral angles, posterior margin and
narrow arcuate transverse fascia over pronotal disc; scutellum, clavus, and hemely-
tral membrane black; corium black with narrow orange transverse fascia near middle
third; connexival segments III to VI black with posterior margin orange; segment VII
almost entirely black; abdominal segments III to VI pale brown, VII black. Ventral col-
oration. Head black with buccula orange; rostral segment I bright orange, II hazel
with basal joint orange, III hazel with posterior half orange, and IV hazel; thorax in-
cluding anterior and posterior lobe of metathoracic peritreme bright to dull black with
following areas orange: collar, acetabula, and posterior margin of propleura, me-
sopleura, and metapleura; legs red brown; coxae pale hazel; abdominal sterna and
genital capsule black, with posterior margin of sterna III to VII orange.
Measurements. Head length 1.88, width across eyes 1.60, interocular space 0.88,
interocellar space 0.46, preocular distance 1.04; length of antennal segments: I, 1.16,
II, 2.04, III, 1.52, IV, 2.80. Pronotum: Total length 2.20, width across frontal angles
1.48, width across humeral angles 3.52. Scutellar length 1.44, width 1.52. Total body
length 12.15.
Holotype: 6 Colombia: Rio Negro, Cundina-Maria, 1000m, 10-1-1965 (W.
Schmidt). In Forschungsinstitut und Naturmuseum Senckenberg, Frankfurth am
Main, Germany.

Etymology: From the Greek, promeces, elongate, and ops, face.

Amblyomia promeceops differs from all other members of the genus in having the
buccula and rostral segment I orange, the corium black with a narrow orange trans-
verse fascia, and the pronotum black with following areas orange: collar, humeral an-
gles, posterior margin, and narrow arcuate transverse fascia over pronotal disc. The
other two species differ in having the buccula and rostral segment I black, and the
pronotum and corium with other color pattern.


I thank the following colleagues and institutions for the loan of specimens: Wolf-
gang Naessig (Forschungsinstitut und Naturmuseum Senckenberg, Frankfurth am
Main, Germany) and Dimitri Forero and Fernando Fernandez (Universidad Nacional
de Colombia, Santa F6 de Bogota (Instituto de Investigaciones de Recursos Biologicos,
Alexander von Humboldt). I express special thanks to Ernesto Barrera (Instituto de
Biologia, Universidad Nacional Aut6noma de M6xico), Jesus Contreras, and Cristina
Urbina for the dorsal view illustrations and genital drawings.


BRAILOVSKY, H. 1982. Hemiptera-Heteroptera de M6xico XXIV. Nuevos registros de la
tribu Mictini y descripci6n de un nuevo g6nero y dos nuevas species (Core-
idae). An. Inst. Biol. Univ. Nal. Aut6n. M6xico, Ser. Zool. 52 (1981): 277-288.
CASINI, C. 1984. Descripci6n de un nuevo g6nero de la Familia Coreidae y algunas con-
sideraciones sobre el g6nero Mamurius (Hemiptera). Bol. Soc. Zool. del Uru-
guay (2da. 6poca) 2: 5-11.
STAL, C. 1870. Enumeratio Hemipterorum I. K. Svenska Vetensk-Akad. Handl. 9 (1):

2-16445/483-488/FES/ddj Page 483 Wednesday, January 6, 1999 9 06 AM
t t

Thomas:A New Chlorocoris 483


USDA-ARS Subtropical Agricultural Research Center
2301 S. International Blvd., Weslaco TX 78596

A new subgenus and new species of Chlorocoris is described from the Caribbean is-
land of Jamaica. This is the first species of this genus reported from the Antilles. The
species is remarkable for an unusually enlarged metatarsus in males and the angu-
late apex of the femora.

Key Words: Pentatomidae, stink bug, taxonomy, Jamaica

Un subg6nero nuevo, y una especie nueva, del g6nero Chlorocoris es descrito con
origen en la isla Jamaica del mar Caribe. La misma es el primer registro de este
g6nero para las Islas Antillas. La especie nueva es notable porque tiene el metatarso
alargado en los machos y el apice de la superficie superior del f6mur esta angulada.

When the genus Chlorocoris Spinola was last revised (Thomas 1985) no material
was available, nor had any species been reported, from the West Indies. In recent
years four specimens have come to light representing a new species from the island of
Jamaica. Based on the triangular form of the head (Fig. 1), the new species would be
assignable to the nominate subgenus. However, a characteristic feature of the nomi-
nate subgenus is the presence of a pair of elongated processes on the male proctiger.
This character is lacking in the new species. Another remarkable feature of the Ja-
maican species is a sexual dimorphism. The metatarsus of the male has an unusually
enlarged basal segment (Figs. 2-3). In addition, the dorsal apex of each femur is an-
gulate. In all other species the apex is rounded. In the keys to genera including Chlo-
rocoris and its relatives the presence or absence of a stout spine at the apex of the
femur is a diagnostic character (Eger 1978, Rolston & McDonald 1984). The new spe-
cies will key to Chlorocoris if the angulation is not confused with a true spine. Because
of the above stated differences, I am assigning the new Jamaican species to a new sub-
genus of Chlorocoris, described below.

The material available for study consisted of four specimens, two males and two fe-
males, from two localities in Jamaica. The Jamaican specimens were compared to ma-
terial in my reference collection, including paratypes from my earlier revision of
Chlorocoris. Type depositions are indicated by acronyms: United States National Mu-
seum [USNM], Florida State Collection ofArthropods [FSCA], and Donald B. Thomas
collection [DBTC]. The habitus drawing of the new species was prepared by Daniel
Schmidt of Schuyler, Nebraska, based on one paratype specimen. All other illustra-
tions were tracings from camera lucida with a Wild M-5 dissecting microscope at mag-
nification 25x and 50x. All measurements are from the male holotype unless

t t
-,T t t

2-16445/483-488/FES/ddj Page 484 Wednesday, January 6, 1999 9 06 AM

484 Florida Entomologist 81(4) December, 1998

Fig. 1. Chlorocoris tarsalis, new species.

- -
-'7 7 -' 7+-7

2-16445/483-488/FES/ddj Page 485 Wednesday, January 6, 1999 9 06 AM
t t

Thomas:A New Chlorocoris 485

1.0 mm


Fig. 2-3. Sexual dimorphism in metatarsi of C. tarsalis. 2. female, 3. male.

otherwise indicated. Measurements were made with a Zeiss SV8 dissecting micro-
scope with a 10x graduated ocular. Anatomical nomenclature follows Nichols (1989).

Arawacoris, New Subgenus

Type species: Chlorocoris tarsalis, New Species.
Description. Head triangular, lateral margins ofjuga straight, or nearly so, to be-
yond apex oftylus. Rostrum long, apex reaching third visible abdominal sternite in re-
pose. Superior apices of femora angulate. Dorsum of male proctiger inornate, lacking
processes. Inferior margin of posterior rim of pygophore with pair of articulated ap-
pendages, one each side of midline, each about same size as a paramere. Other char-
acters as in Chlorocoris (see Thomas 1985).
Etymology: a latinized combination of"Arawak," the indigenous people of Jamaica,
and greek "koris," meaning "bug."

Chlorocoris tarsalis New Species
Figs. 1-6
Description. Dorsal color green fading to yellow. Form oval, depressed dorsoven-
trally, with angular head and humeri. Length 13.3 mm (female 16.1 mm), width
across pronotum 8.5 mm (female 9.6 mm).
Head. Dorsum flat, surface strigose, devoid of black punctations except at apices of
juga. Lateral margins ofjuga straight to apices. Apex of each jugum subacuminate,
exceeding apex of tylus and forming sinus before tylus, sinus twice as long as wide.
Cranial length (tip ofjugum to imaginary line connecting ocelli) 3.0 mm, width (across
anteocular angles) 2.2 mm. Antennal segment I immaculate, its apex attaining apex
ofjugum. Segments II-IV subequal (exact proportions vary among individual speci-
mens), each about twice length of I; V slightly shorter than IV. Posterior termination
of each buccula evanescent in profile. First rostral segment slightly longer than buc-
cula. Rostrum in repose attaining third visible abdominal sternite.
Thorax. Anterolateral pronotal margins rectilinear in dorsal view, serrate. Humeri
angular, prominent, acute. Dorsum of pronotum devoid of black punctations except in

t t
-T t t

2-16445/483-488/FES/ddj Page 486 Wednesday, January 6, 1999 9 06 AM
t t

486 Florida Entomologist 81(4) December, 1998

area immediately adjacent to humeral angles. Length of pronotum at midline 3.1 mm.
Scutellum and hemelytral coria devoid of markings except for pale pustule on disc of
latter. Length of scutellum 4.9 mm. Posterior margin of corium sinuate. Hemelytral
membrane clear, transparent, with scattered green flecks. Auricle of metathoracic
scent gland orifice short, reaching about one-fifth distance to metapleural margin.
Femora and tibiae immaculate. Superior surface of apex of each femur angulate but
not spinous. Basal tarsal segment of metathoracic legs of male fusiform, thickened,
and notably elongate (Fig. 3) compared to mesotarsi, or metatarsi of female (Fig. 2).
Abdomen. Midline of first three abdominal sternites sulcate for reception of ros-
trum. Spiracular margins and apices of sternites concolorous with disc of sternum.
Connexivum without spots or stripes. Greatest width of abdomen, 8.3 mm.
Genitalia. Male pygophore (Fig. 4) broadly open posteriorly and dorsally with ven-
troposterior rim deflexed. Surface of pygophore at lateral angles dense with short
bristles and with a smaller patch of dense bristles on either side of midline just ectal
to inferior margin. Inferior margin bearing, on either side of midline, an articulated,
sclerotized L-shaped "pseudoclasper" (hypandrium?), projecting into lumen of procti-
ger. Basal portion of each pseudoclasper bearing a porrect, angular tooth. Erect arm
of pseudoclasper subfoliate. Proctiger inornate. Parameres large with broad com-
pressed base expanding into two angular projections: a dorsal, flat, rectangular pro-
jection, and a ventral, flat, acutely angled, rhomboidal projection (Fig. 5).
First gonocoxites of female thickened, posterior margin strongly sinuate (Fig. 6).
Ninth paratergites elongate, apices acuminate, exceeding posterior margin of eighth
paratergite. Apex of eighth paratergite subspinose with spiracles present but dis-
placed to notch of basal angle.
Holotype. Male. verbatim label data: JAMAICA: Green Hills. 13-20-XI-66. A. B.
V- Gurney [USNM]. Allotype: Female, with same label data as holotype [USNM].
Paratypes: Female, with same label data as holotype [DBTC]. Male labeled: (a) JA-
MAICA: Parish of St. Andrew, 4,000 ft. Holywell Forest Camp. Blacklight. (b) R. E.
Woodruff, 16-VI-75, Blacklight Trap [FSCA].


The genus Chlorocoris is superficially similar and probably closely related to a
group of genera that includes Chloropepla Stal, Loxa Amyot & Serville, Fecelia Stal,
and Mayrinia Horvath. All of these genera, except Chlorocoris, have the superior apex
of the femora terminating in a minute spine. All known species of Chlorocoris, with
the exception of the new Jamaican species, have the apex of the femora rounded. The
new species has an intermediate condition with the apex of the femora angulate. In-
asmuch as the condition of the femoral apex is important in distinguishing genera,
and considering the unusual sexual dimorphism of the metatarsus, one might plausi-
bly erect a new genus for this species. I, therefore, reviewed the characteristics of each
genus and searched for trends that might support this position.
Within this generic complex, the most striking morphological variation is found in
the male genital apparatus. In fact, the terminalia are elaborate to a degree that
would almost seem to impede rather than enhance coition. One structure in particu-
lar, referred to in the description of the new species as a "pseudoclasper," is especially
enigmatic. It is an appendage situated at the middle of the posterior rim of the pygo-
phore. Its function is unknown. In most species of Chlorocoris, and its ally Mayrinia,
the appendage is a singular, erect structure, fused to the rim of the pygophore, the size
and conformation of which vary greatly among the individual species. It is referred to
as the "hypandrium" in the revision of Mayrinia by Grazia-Vieira (1972). In Chlo-

t T
I-t t t

t 2-16445/483-488/FES/ddj Page 487 Wednesday, January 6, 1999 9 06 AM

Thomas:A New Chlorocoris


1 0LO mm

Figs. 4-6. Genitalia of C. tarsalis. 4. pygophore, posterior view. 5. left paramere, en-
tal view. 6. Female genitalia. Gx, = first gonocoxite, Inf.M. = inferior margin, Pm =
paramere, Pr = proctiger, PsC = pseudoclasper, Pt, = eighth paratergite, Pt, = ninth





2-16445/483-488/FES/ddj Page 488 Wednesday, January 6, 1999 9 06 AM
t t

488 Florida Entomologist 81(4) December, 1998

ropepla, Fecelia, and Loxa, there is a pair of articulated structures in the same posi-
tion, variously called "hypandria" (Grazia 1968, 1976) or "pygophoral appendages"
(McDonald 1966). I can only presume, as did Grazia, that the paired, articulated
structures found in some genera are homologous to one another and to the erect struc-
ture fused to the rim of the pygophore in others. Within the genus Chloropepla, C. vi-
gens Stal has a pair of appendages, while C. aurea Grazia has none. Likewise, among
the species ofLoxa they may be present or absent (Eger 1978). In most species ofChlo-
rocoris, as in Mayrinia, the appendage is completely fused to the rim of the pygophore.
But in C. rufispinus Dallas and C. rufopictus Walker, the appendage is articulated, or
at least, there is a perceptible line of attachment that is not evident in the other spe-
cies. The subgenus Arawacoris has a pair of articulated appendages and thus is ex-
ceptional in this regard from all other Chlorocoris. Yet, it seems hazardous to attach
much significance to the difference. The degree of homoplasy in the structure of the
pygophore is mirrored in the ornamentation of the proctiger. The nominate subgenus
of Chlorocoris has a pair of elongate processes that overlie, and are parallel to, the
dorsum of the proctiger. In the subgenus Monochrocerus the processes are short and
are oriented horizontal to the length of the proctiger. But some species, including C.
tarsalis, have no processes at all. Similarly, in the genus Loxa, some species have
paired spines on the proctiger and others none.
The pattern of divergence and convergence in the form of the genitalia in this com-
plex presents a challenge to those who are firm in the belief that phylogeny can be ob-
jectively derived from the cladistic nesting of character-states. Having failed in my
own feeble efforts to find clear evolutionary trends among the recognized genera with
respect to these genital characters, I am reluctant to erect a new genus. Rather, I
opine that this somewhat aberrant Jamaican species is best classified as an isolated
subgenus of Chlorocoris.


I am indebted to my colleagues Joseph E. Eger of DowElanco Corporation and
David A. Rider of North Dakota State University for bringing these specimens to my
attention and for critical reviews of the manuscript.


EGER, J. E. 1978. Revision of the genus Loxa (Hemiptera: Pentatomidae). J. New York
Entomol. Soc. 86: 224-259.
GRAZIA, J. 1968. S6bre o g6nero "Chloropepla" Stal, 1867, com a descricao de uma
nova esp6cie (Hemiptera, Pentatomidae, Pentatominae). Rev. Bras. Biol. 28:
GRAZIA, J. 1976. Revisao do g6nero Fecelia Stal, 1872 (Heteroptera, Pentatomidae,
Pentatomini). Rev. Bras. Biol. 36: 229-237.
GRAZIA-VIEIRA, J. 1972. O genero Mayrinia Horvath, 1925 (Heteroptera, Pentatomi-
dae, Pentatomini). Rev. Peruana Entomol. 15: 117-124.
MCDONALD, F. J. D. 1966. The genitalia of North American Pentatomoidea (Hemi-
ptera: Heteroptera). Quaest. Entomol. 2: 7-150.
NICHOLS, S. W. (compiler). 1989. The Torre-Bueno Glossary of Entomology. New York
Entomological Society, New York, NY.
ROLSTON, L. H., AND F. J. D. MCDONALD. 1984. A conspectus of Pentatomini of the
Western Hemisphere, Part 3 (Hemiptera: Pentatomidae). J. New York Entomol.
Soc. 92: 69-96.
THOMAS, D. B. 1985. Revision of the genus Chlorocoris Spinola (Hemiptera: Pentato-
midae). Ann. Entomol. Soc. Am. 78: 674-690.

-It t t

Mayfield: Chapmannia floridana Pollination


Reed College 3203 SE Woodstock, Portland, OR 97202-8199


The visitation rates of two major insect visitors, Bombus impatiens Cresson and
Augochloropsis sp., of the Florida endemic plant species Chapmannia floridana Tor-
rey & A. Gray, (Fabaceae); are determined for different microhabitats at the Archbold
Biological Station in the summer of 1995. Significant differences in the total number
of visits to each site were observed. Each pollinator species was found to visit different
vegetation densities and sites with different frequencies. Relationships were also
found between visitation rates and temperature and flower size. Variation in visita-
tion rates did not significantly affect seed set in C. floridana. Disturbance did not
seem to play a major role in determining visitation rates of either of the pollinators.
Vegetation composition and flower density appear to be the best indicators of visita-
tion rates to populations of this plant species.

Key Words: Bombus impatiens; Augochloropsis; Florida scrub; disturbance; pollina-
tion; foraging patterns; Chapmannia floridana


Las tasas de visitaci6n de los dos insects visitantes principles, Bombus impa-
tiens Cresson yAugochloropsis sp., de una plant end6mica de Florida, Chapmannia
floridana Torrey y & A. Gray (Fabaceae); fueron determinadas para los micro-habitats
diferentes en la Estaci6n Biol6gica Archbold en el verano de 1995. Se observaron
diferencias apreciables entire los numerous de visitaciones totales en cada terreno.
Tambi6n habia relaciones entire las tasas de visitaci6n y la temperature, y el tamano
de la floor. La variaci6n en las tasas de visitaci6n no mostr6 efecto apreciable en el
numero de semillas de Chapmannia floridana. El disturbio no fue important para
determinar las tasas de visitaci6n de ningun polinizador. La composici6n de la veget-
aci6n y la densidad de las flores parecen ser los mejores indicadores de las tasas de
visitaci6n a poblaciones de esta especie de plant.

Many studies have shown that disturbance, both natural and human induced, can
cause significant changes in the structure and dynamics of ecosystems. Some of the
most obvious changes caused by habitat destruction or damage are changes in biodi-
versity, niche occupation, and animal feeding behavior (Armesto & Pickett 1985,
Glitzenstein et al. 1986, Coffin & Lauenroth 1988, Foster & Zebryk 1993, Aizen &
Feinsinger 1994). When disturbance is natural, occurring because of fire, storm dam-
age, or landslides, biodiversity in an area tends to increase and niche occupations and
animal feeding preferences remain unchanged (Armesto & Pickett 1985, Feinsinger
et al. 1987). However, when areas are damaged by human activities such as grazing,
agricultural development, and logging, drastic changes in species composition and
foraging behavior are often observed (Brian 1959, Halpern 1988).
Despite the growing interest in the effects of human disturbance on ecosystems
and on specific rare plants and animals, there are still many gaps in our understand-

Florida Entomologist 81(4)

December, 1998

ing of how human disturbance affects the life history of these organisms and their in-
teractions with each other. In this study, I examine the apparent success of the plant
species Chapmannia floridana Torrey & A. Gray; (Fabaceae) in relation to the visita-
tion rates of its major pollinators. I selected C. floridana for this study because of the
differences in its population structure in disturbed and undisturbed areas. C. flori-
dana is a very unusual Florida endemic scrub plant because of its success in disturbed
areas. Most endemic plant species in central Florida are disappearing as central Flor-
ida scrub disappears. Undoubtedly, there are many factors that contribute to the suc-
cess of C. floridana in disturbed areas. I was interested in determining if the
differences in disturbed and undisturbed habitat and in population structures affect
pollination services to this plant, possibly contributing to its success.
In this study, variation in pollinator composition and visitation rates to C. floridana
were determined for six different sites varying in vegetation composition and distur-
bance. The effects of flower size and temperature on the pollinator composition visiting
this plant were also determined (Cruzan et al. 1988). Information on the foraging behav-
ior of this plant's pollinators can provide valuable information about how human and
natural disturbance affects the relationship between C floridana and its pollinators.


Species description of Chapmannia floridana:

Chapmannia floridana is an endemic Florida herb ranging from Clay to Collier
Counties in central Florida (Gunn, 1980). It can be found growing in open scrub, san-
dhills, and disturbed areas such as pastures and road sides. Little attention has been
paid to C. floridana since its original description by Torrey and Gray in 1838. The in-
formation available about this species is minimal and often inconsistent (Gunn 1980).
My greenhouse studies on this species suggested that selling does not occur, but re-
sults were inconclusive and more study is needed. C. floridana is not thought to be a
nitrogen fixer.
C. floridana blooms between mid-April to early September. It is commonly found
in large numbers in highly disturbed roadsides and pastures and in smaller numbers
in undisturbed or burned habitat. Flowering individuals range in height from 20-101
cm and usually consist of 1 to 10 flowering stalks. Flowers are yellow-orange and
range in size from 2.0-4.0 cm from top to the bottom and 0.9-4.0 cm horizontally across
the petals. Flowers open at about 8:00 am early in the season and about 6:00 am later
in the season. Flowers remain open from three to five hours for a single day, depend-
ing on air temperature and light conditions. I never observed flowers open after 10:30
am. Each flower produces one to four seeds in a legume covered with small viscid
hairs (Gunn 1980).

Site description

I chose six sites for this study at the Archbold Biological Station in Highlands
County, Florida. All sites were located in different micro-habitats (Table 1). Sites A,
B, D, and E were 288 m2(12 m x 24 m) and sites C and F were 240 m2 (12 m x 20 m).
Each site was defined by its disturbance level and the density of C. floridana (Table
1). In this study, human disturbance was any area that had been grazed, plowed, bull-
dozed, or driven over extensively within the past 15 years. Burned sites were not con-
sidered disturbed since fire is a natural part of C. floridana's environment. I was not
able to find any high density-undisturbed sites or low density-disturbed sites.

Mayfield: Chapmannia floridana Pollination


Mean # Density
Site Flowers Rating Disturbance Habitat Type Major Vegetation

A 10 Medium Disturbed Damaged Scrub Scrub oak, small
herbaceous species
B 43 High Disturbed Former pasture Legumes, weedy
herbs, oaks, palmet-
tos, grasses
C 175 High Disturbed Road side near Surrounded by oaks,
a citrus orchard weedy herbs, grass
D 3 Low Undisturbed Scrubby Flat- Oaks, palmettos, and
woods other Scrubby flat-
E 5 Low Undisturbed Southern Ridge Turkey oak, hickory,
Sandhill and pine, scrub oaks,
palmettos, native
F 13 Medium Undisturbed Rosemary Bald Scrub rosemary, oaks,
palmettos, and native

Field Experiments

I observed pollination events in each of the six sites five days a week for all weeks
between June 26 and July 28, 1995. Each morning I observed one of the six sites for
2 h. The exact time observations began and ended depended on weather conditions
and when flowers began to close. This 2 h period encompassed the majority of the time
these flowers were open and the sun was up.
Within a selected site, individual flowering C. floridana were observed for 10
minute intervals. In each observation period the number, type, and the time spent by
insect visitors at flowers were recorded. I watched between 1 and 8 flowers at a time,
depending on the orientation and distance of the plants from my observation point
(~ 1 m from the nearest plant). All flowers on a selected plant were observed together.
In low density sites, all flowering C. floridana were observed at the same time and in
high density sites plants were selected arbitrarily. Different plants at high and me-
dium density sites were selected throughout the summer of 1995.
Prior to each ten minute observation period, I recorded the air temperature. After
observations were completed each day, the flower dimensions of all observed flowers
were measured (Kearns & Inouye 1993), the number of flowers blooming during the
observation period were counted, and the approximate vegetation cover in a 1 m ra-
dius around each C. floridana plant was determined. For vegetation cover I used five
arbitrary categories: 0%-15% (mostly open sand), 15%-30%, 30%-60%, 60%-80% and
80%-100% (the most dense vegetation). Seed set was determined several weeks after
visitation observations.
Insect identifications were based on the insect collection at Archbold Biological
Station. One time visitors were not always identified.

Florida Entomologist 81(4)

December, 1998

Statistical Analysis

Splus (MathSoft Inc., Seattle, WA), SAS (SAS Institute Inc., Cary, NC), and Stat-
view (Abacus Concepts Inc., Berkley, CA) were used to analyze the data from this
study. Because of the low number of visits of both species, visits by each species were
considered independent of one another for all statistical tests. The low number of vis-
its also made it necessary to use a Poisson regression (A Poisson regression is a re-
gression based on a Poisson rather than normal distribution), an analysis of deviance
based on a Poisson distribution and analysis of variance (ANOVA) were used to look
at differences in pollinator visitation frequency (visitation rate of each species per 10
minute interval) (Snedecor & Cochran 1989). Visits by minor visitors were analyzed
with a one-group t-test, and the average time spent on a flower by a visitor was ana-
lyzed with ANOVA.



In this study, C. floridana flowers were commonly visited only by two types of
bees: bumble bees, Bombus impatiens (Hymenoptera: Apidae; bumble bees), and sev-
eral species of metallic green solitary bees in the genus Augochloropsis (Hy-
menoptera: Halictidae). These bees were usually observed ripping holes in the sides
of the keel petals of the flowers and vibrating their wings to get pollen out of the floral
tube. In a past study on C. floridana pollination, Dialictus pilosus (Smith) (Hy-
menoptera: Halictidae), a small solitary bee, was the only major insect visitor to this
plant (Gunn et al. 1980). In my entire study, a single Dialictus nymphalis (Smith) was
observed visiting a C. floridana flower and no D. pilosus were seen. The observed Di-
alictus sp. did not vibrate its wings to get pollen from the flower and it was too small
to touch the stigma of these flowers.
Several minor insect visitors were observed foraging C. floridana flowers within
my plots, including: D. nymphalis, Geron vitripennis Loew (Diptera: Bombyliidae) and
Copestylum barei Loew (Diptera: Syrphidae), and several other unidentified bees and
flies. In total, 15 of the 262 observed insect visits were by insects other than B. impa-
tiens andAugochloropsis spp. Ten of these visits occurred at Site D, four at Site B, and
one was observed visiting a single flower at Site F. No ants or beetles were observed
visiting flowers, although several beetles were seen eating petals in the afternoon af-
ter the flowers were closed. Honey bees were seen foraging at other flowering species
in all sites, but were never observed visiting C. floridana.

Results of Statistical Analysis

An analysis of deviance test (Snedecor & Cochran 1989) on the total number of in-
sect visits to C. floridana flowers showed that the total number of insect visits was sig-
nificantly different between site, bee type, temperature, percentage vegetation cover,
and flower size as well as all interactions with bee type (all p-values < 0.001 except for
flower size which had a p-value of 0.0153). Specific effects were determined using a
Poisson regression (Table 2).
A single group t-test indicated that there was no significant difference between the
number of minor visitors visiting in each site or between disturbed and undisturbed
sites. No significant differences (based on an ANOVA test) in the time spent by visi-
tors foraging in 10 minute intervals were found for any variable. There were also no
significant differences in seed set for any of the variables listed above.

Mayfield: Chapmannia floridana Pollination


Variable Coefficient Error Z-value P-value

Site A (intercept) 4.65 0.856 5.44 < 0.01
Site B -0.667 0.287 -2.33 < 0.05
Site C 0.658 0.308 2.13 < 0.05
Site D -1.43 0.473 -3.02 < 0.01
Site E -0.561 0.359 -1.56 NS
Site F -1.56 0.329 -4.75 < 0.01
Bee Type -6.17 1.57 -3.94 < 0.01
Temperature -0.237 0.0314 -7.53 < 0.01
60% Cover -0.895 0.210 -4.27 < 0.01
30% Cover 0.963 0.216 4.45 < 0.01
15% Cover 0.0126 0.393 0.0321 NS
0% Cover 0.477 0.281 1.70 NS
Flower Area (cm2) 0.614 0.0255 2.41 < 0.05
Site B vs. Bee Type -1.12 0.724 -1.55 NS
Site C vs. Bee Type -2.061 0.582 -3.54 < 0.01
Site D vs. Bee Type 1.103 0.767 1.44 NS
Site E vs. Bee Type 0.0650 0.551 0.118 NS
Site F vs. Bee Type 0.573 0.494 1.16 NS
Temp vs. Bee Type 0.242 0.0588 4.12 < 0.01
60% Cover vs. Bee Type 0.483 0.478 1.010 NS
30% Cover vs. Bee Type -1.71 0.437 -3.920 < 0.01
15% Cover vs. Bee Type -0.0444 0.894 -0.0496 NS
0% Cover vs. Bee Type -0.264 0.445 -0.594 NS

1Null Deviance: 1304.3 on 1289 degrees of freedom
Residual Deviance: 894.9 on 1267 degrees of freedom

For the Poisson regression shown in Table 2, the dependent variable is the number
of bees visiting during a 10 min observation period. All main effects and interaction
effects are compared to the values for Site A at 0C (the assumed intercept for this
model), 80-100% vegetation cover, and with a flower area of 1 cm2. For all of the main
effects, negative Z-values indicate that there were fewer total visits than expected.
The significant negative Z-value for bee type indicates that there were significantly
fewer Augochloropsis than expected. The negative Z-value for the temperature indi-
cates that the number of total insect visits decreased as the temperature increased.
The significant positive Z-value for flower area indicates that flowers with larger
flower areas had significantly more insect visits than smaller flowers. For interaction
effects, negative values indicate fewer Augochloropsis than expected. Positive values
indicate fewer B. impatiens than expected. The significant negative interaction effect

Florida Entomologist 81(4)

December, 1998

of bee type and temperature shows that as the temperature increased B. impatiens
visitation decreased significantly, while the number of Augochloropsis visitation did
not differ significantly according to temperature. The significance of the main temper-
ature effect is due mainly to the change in the number ofB. impatiens visiting flowers
at higher temperatures.
The results of the Poisson regression also indicate that sites B, D, and F had sig-
nificantly fewer total pollinator visits than expected. Site C had significantly more to-
tal visits than Site A but significantly fewer Augochloropsis visitors than expected
(Table 2). The fitted means in Table 3 show differences between the number of visits
to each site byAugochloropsis and B. impatiens. The mean number of Augochloropsis
visits are lower than the mean number of B. impatiens visits at all sites except site D,
a low density undisturbed site. The mean number of B. impatiens visiting is highest
in the three disturbed sites A, B, and C.
Both bee types showed no vegetation preferences, although each type of bee was
observed visiting significantly more frequently in certain vegetation covers. The Pois-
son analysis indicates a significantly lower number of visits by both species to areas
of 60-80% vegetation cover and significantly more total visits to 30-60% vegetation
cover but significantly fewerAugochloropsis visits indicating that the total increase in
visitation in 30-60% vegetation cover was due to high visitation rates byB. impatiens
(Table 4). An analysis of variance, type three sums of squares, indicates that there is
a significant interaction effect of site and percentage vegetation on the number of to-
tal bee visits (p-value = 0.0056) and with the number of B. impatiens visits (p-value =
0.0001). The frequencies of visits by each bee type within each vegetation coverage are
shown in Table 5.


Both B. impatiens and Augochloropsis sp. exhibited complex foraging behavior
while visiting C. floridana flowers. Statistical results indicate that both pollinators
preferred larger flowers. This preference could indicate that the pollinators are using
flower size to identify flowers with larger pollen or nectar rewards. The results also in-
dicate that the number of bees visiting decreased significantly as temperature in-


Bombus Impatiens Augochloropsis
Site Mean Mean Std Dev H Mean Std Dev H
Flower #

A 10 0.70 1.14 33 0.47 1.14 22
B 43 0.53 0.90 37 0.06 0.24 8
C 175 0.56 0.96 62 0.06 0.26 11
D 3 0.16 0.62 7 0.21 0.47 9
E 5 1.40 0.70 20 0.18 0.44 9
F 13 0.19 0.49 13 0.15 0.47 16

Mayfield: Chapmannia floridana Pollination


Bombus Impatiens Augochloropsis Total

Veg Std Std Std
Cover Mean Dev N Mean Dev N Mean Dev N

0-15% 0.49 0.95 15 0.20 0.41 20 0.69 0.99 35
15-30% 0.25 0.50 6 0.07 0.26 11 0.69 1.05 17
30-60% 0.66 1.04 84 0.09 0.33 13 0.31 0.55 97
60-80% 0.20 0.38 15 0.07 0.49 5 0.75 1.12 20
80-100% 0.57 0.10 43 1.22 2.26 26 0.87 1.40 69

creased. It is likely that the bees' visitation rates decreased as flowers began to close
later in the morning when temperatures were warmer and it is possible that floral re-
wards were depleted later in the morning.
Although the total number of insects visiting each site varied significantly for all
but one site, only site C had a significantly different number of Augochloropsis visit-
ing. Site C had the greatest number of flowering C. floridana and was a highly dis-
turbed site on the edge of a citrus orchard. The high number of B. impatiens visiting
this site could have been related to the larger number of flowers at site C (Table 3),
proximity to other flower sources, or the presence of near by nests. The low number of
Augochloropsis forging at this site could have been caused by overwhelming competi-
tion from the abundant B. impatiens at this site.
The amount of vegetation cover most often visited by each type of pollinator did
not follow a pattern except when considered in connection with the sites. Vegetation
preferences of Augochloropsis sp. were not related to sites dominated by a particular
vegetation coverage while B. impatiens visits were (Tables 4 and 5).
There was no clear indication from this study that the foraging patterns of C. flori-
dana's pollinators were directly affected by the level of human disturbance in the
area. Seed set levels did not differ significantly between sites, indicating that despite
the significant difference in pollinator visitation between sites, all populations ob-
served in this study were receiving similar pollination services. Bombus impatiens
were observed visiting sites A, B, and C more frequently than in the undisturbed sites,
but this was not a significant difference. Augochloropsis was found visiting most fre-
quently in disturbed site A and least frequently at the other two disturbed sites B and


Vegetation Cover Site A Site B Site C Site D Site E Site F

0-15% 9 4 30 59 12 0
15-30% 1 0 170 30 0 0
30-60% 0 39 49 42 26 39
60-80% 0 0 0 0 0 76
80-100% 33 0 0 21 12 20

Florida Entomologist 81(4)

December, 1998

C (Table 3). Sites B and C had the highest flowering densities and high visitation
rates by B. impatiens indicating that competition may play a role in the visitation
rates of Augochloropsis to this plant. Site A was bordered on three sides by undis-
turbed habitat which may have contributed to the visitation rates of Augochloropsis.
The visitation rates of these pollinators were undoubtedly affected by factors not
determined in this study. From the data collected, the factors that appear to play the
greatest role in determining visitation rates to C. floridana are temperature, flower
size, and flowering density. Although disturbance does not seem to strongly affect the
pollinator visitation rates or resulting seed set in this plant, the higher numbers of
flowers in disturbed areas clearly attract more pollinators than the low density undis-
turbed populations do. This relationship between flowering density and visitation
rate may be important to the reproductive success of this plant species in the long
term or at least in years with fewer pollinators.


Special thanks to Pedro Quintana-Asencio and Eric Menges for providing me with
the opportunity to perform this study and for much help with my experimental design
and analysis; Albyn Jones and Dr. Kenneth Koehler for hours of statistical help; Fritz
Davis, Margaret Evans, Deborah Graves, Vishnu Manteuffel, Joyce Voneman, Alex
Wild, and Becky Yahr for helpful suggestions and field assistance; and Nick Waser,
Paul Aigner, Sandy DeSimone, Bob Kaplan, Keith Karoly, and John Mayfield for help-
ful suggestions and proofreading of this report. Also, thanks to The Archbold Biologi-
cal Station for letting me use their property and for providing me with the internship
to perform this study.


AIZEN, M. A., AND P. FEINSINGER. 1994. Forest fragmentation, pollination, plant re-
production in a chaco dry forest, Argentina. Ecol. 75: 330-351
ARMESTO, J. J., AND S. T. A. PICKETT. 1985. Experiments on disturbance in old-field
plant communities: impacts on species richness and abundance. Ecol. 66: 230-240.
BRIAN, A. D. 1959. Differences in the flowers visited by four species of bumble-bees
and their causes. J. Animal Ecol. 26: 71-98.
COFFIN, D. P., AND W. K. LAUENROTH. 1988. The effects of disturbance size and fre-
quency on a shortgrass plant community. Ecol. 69: 1609-1617.
CRUZAN, M. B., P. R. NEAL, AND M. F. WILSON. 1988. Floral display in phyla-incisa
consequences for male and female reproductive success. Evolution 42: 505-515.
turbance, pollinator predictability, and pollination success among Costa Rican
cloud forest plants. Ecol. 68: 1297-1305.
FOSTER, D. R., AND T. M. ZEBRYK. 1993. Long-term vegetation dynamics and distur-
bance history of a Tsuga-dominated forest in New England. Ecol. 74: 982-998.
GLITZENSTEIN, J. S., P. A. HARCOMBE, AND D. R. STRONG. 1986. Disturbance, succes-
sion, and maintenance of species diversity in an east Texas forest. Ecological
Monographs 56: 243-258.
GUNN, C. R., E. M. NORMAN, AND J. S. LASSETTER. 1980. Chapmannia floridana Tor-
rey and Gray (Fabaceae). Brittonia 32: 178-185.
HALPERN, C. B. 1988. Early successional pathways and the resistance of forest com-
munities. Ecology, 69: 1703-1715.
KEARNS, C. A., AND D. W. INOUYE. 1993. Techniques for pollination biology. The Uni-
versity Press of Colorado. Niwot, CO.
SNEDECOR, G. W., AND W. G. COCHRAN. 1989. Statisitcal Methods: Eighth edition.
Iowa State University Press, Ames.

Robacker et al.: Bacteria Attractive to Mexican Fruit Fly 497


'Crop Quality and Fruit Insects Research, USDA, Agricultural Research Service
2301 South International Blvd., Weslaco, TX 78596

2Mission Biological Control Center, USDA, Animal and Plant Health Inspection
Service, Box 2140, Mission, TX 78573

3National Center for Agricultural Utilization Research, Bioactive Agents Research,
USDA, Agricultural Research Service, 1815 N. University Street, Peoria, IL 61604


Filtrates of 11 bacteria representing 4 higher taxonomic categories were attractive
to Mexican fruit flies, Anastrepha ludens (Loew) (Diptera: Tephritidae) in laboratory
bioassays. All bacterial filtrates were more attractive at pH 9 than at pH 5 although
filtrates at pH 5 were more attractive than water controls. The effects of pH on attrac-
tiveness of filtrates were consistent with an hypothesis that attractive principals of
bacterial filtrates were various nitrogen-containing compounds and carboxylic acids
that became more volatile at specific pH's resulting in increased attractiveness. Vola-
tiles produced by the bacteria were sampled by solid-phase microextraction and iden-
tified by GC and GC-MS. Attractive principals identified were ammonia, aliphatic
amines, pyrazines, imines, and acetic acid. Relative amounts of most of the chemicals
were not closely tied to bacteria taxonomy.

Key Words: Anastrepha ludens, attractants, bacteria, amines, acetic acid, solid phase
microextraction (SPME)


Los filtrados de 11 bacteria que representan a 4 categories altas taxon6micas
atrayeron a las mosca de la fruta mexicana, Anastrepha ludens (Loew) (Diptera: Te-
phritidae), en bioensayos de laboratorio. Todos los filtrados bacteriales fueron mas
atrayentes al pH 9 que al pH 5, aunque los filtrados de pH 5 fueron mas atrayentes
que el testigo. Los efectos del pH sobre la atracci6n de los filtrados fueron consistentes
con la hip6tesis de que los quimicos atrayentes de los filtrados bacteriales eran various
compuestos de nitr6geno y acido carboxilico que se hacen mas volatiles a pH especifi-
cos resultando en un aumento en su atracci6n. Volatiles producidos por las bacteria
fueron colectados usando micro-extracci6n de fase s6lida y fueron identificados por
cromatografia de gas y espectr6metro de masa. Los quimicos atrayentes fueron iden-
tificados como amoniaco, aminos alifaticos, pirazines, imines y acido ac6tico. Las con-
centraciones de various quimicos no estuvieron muy cercanamente relacionadas a la
taxonomia de las bacteria.

Volatile chemicals from bacterial fermentations attractive to fruit flies have come
under increased scrutiny during the last 10 years as possible attractants. Several pa-
pers have reported identification of ammonia from bacteria cultures (Gow 1954, Drew

Florida Entomologist 81(4)

December, 1998

& Fay 1988, Robacker & Flath 1995). Ammonia has long been known as a powerful at-
tractant for fruit flies (Jarvis 1931). Other studies have resulted in identification of
additional volatile chemicals from cultures of bacteria (Hayward et al. 1977, Lee et al.
1995, Robacker & Flath 1995, DeMilo et al. 1996, Robacker & Bartelt 1997).
Attractiveness of chemicals (other than ammonia) identified from bacterial odors
has been demonstrated in only a few studies. Drew (1987) demonstrated that the bac-
teria-produced chemicals 2-butanone and 1-butanol were attractive to Bactrocera try-
oni (Froggatt), purportedly because of their structural similarity to the
parapheromone cuelure. Robacker & Flath (1995) and Robacker & Bartelt (1997)
identified and demonstrated attractiveness for ammonia, several amines, imines,
pyrazines and acetic acid from three species of bacteria.
In this research, principals attractive to the Mexican fruit fly, Anastrepha ludens
(Loew) were identified from 11 strains of bacteria that had not been investigated be-
fore. A 3-step procedure was used. First, attractiveness of each bacterium was veri-
fied. Second, the effect of fermentation pH on attractiveness was determined to
characterize the classes of chemicals involved in the attraction response. Third, chem-
icals that fit the attractive-principal profile as determined in the pH tests were iden-
tified and quantified.
The purposes of the work were to determine if similar bacteria produce similar at-
tractive chemicals and conversely if dissimilar bacteria produce different, perhaps
novel, chemicals. Novel chemicals, along with knowledge gained in this work of gen-
eral patterns of volatiles produced by attractive bacteria, could be used in develop-
ment of new lures for fruit flies.


Insects and Test Conditions

Flies used to test attractiveness of bacterial preparations were from a culture that
originated from yellow chapote, Sargentia greggii Coult., (Rutaceae), fruit, a native
host of the fly, collected in Nuevo Leon, Mexico, in 1987. Fly handling and laboratory
maintenance were as described in Robacker & Flath (1995). Flies were sugar-fed and
protein-starved (since eclosion) because previous work indicated this physiological
state maximizes attraction to bacterial odor (Robacker & Garcia 1993). Flies were
used when 6-10 days old.

Bacterial Preparations

Bacteria species used in this work were: Enterobacter cloacae (Jordan);Alcaligenes
faecalis faecalis Castellani & Chalmers; Micrococcus luteus Schroeter; Bacillus
sphaericus Meyer & Neide; B. subtilis Ehrenberg; B. megaterium de Bary; B. popilliae
Dutky; and B. thuringiensis Berliner subspecies shandongiensis, coreanensis,
konkukian, and darmstadiensis. Strains obtained from the American Type Culture
Collection (ATCC) (Rockville, MD) were: E. cloacae (ATCC strain 961);A. f faecalis
(ATCC strain 8750); M. luteus (ATCC strain 23259); B. sphaericus (ATCC strain
4525); B. subtilis (ATCC strain 6051); B. megaterium (ATCC strain 14581); and B.
popilliae (ATCC strain 14706). Strains obtained from the Institut Pasteur (Paris,
France) were:B. t. shandongiensis (strain 22001);B. t. coreanensis (strain 25001); and
B. t. konkukian (strain 34001). B. t. darmstadiensis (strain GUAT1) was obtained
from a soil sample from Guatemala (Martinez et al. 1997).
These taxa were chosen to survey volatile chemicals attractive to the Mexican fruit
fly produced by bacteria over both broad and narrow levels of classification. The four

Robacker et al.: Bacteria Attractive to Mexican Fruit Fly 499

genera represent four distinct higher taxonomic categories: Enterobacter, faculta-
tively anaerobic, gram-negative rods; Alcaligenes, aerobic gram-negative rods and
cocci; Micrococcus, gram-positive cocci; and Bacillus, endospore-forming, gram-posi-
tive rods (Holt 1984). The five species of Bacillus and the 4 subspecies of B. thuring-
iensis allow an analysis of volatiles produced by more closely related strains.
All ATCC strains were fermented in trypticase soy broth (BBL, Baltimore, MD) in
a shaker for 5 days at 30C. B. t. shandongiensis, B. t. konkukian, and B. t. darmsta-
diensis were fermented in Bacto nutrient broth (DIFCO Laboratories, Detroit, MI) in
a shaker for 3, 6, and 3 days, respectively, at 30C. B. t. coreanensis was fermented in
a shaker in a growth medium (medium B) developed by Dulmage et al. (1970) for 5
days at 30C. Fermentation times and media were based on preliminary bioassays
showing maximum attractiveness for these times and media. Bacterial cultures were
centrifuged and the resulting supernatants were filtered to remove bacterial cells as
described previously (Robacker & Flath 1995). Martinez et al. (1994) demonstrated
that filtered and unfiltered cultures of several bacteria species were equally attractive
indicating that the attractants were dissolved in the filtrate. Because attractive
chemicals were retained in filtrate, there was no concern that bacterial cultures may
have been too old to contain actively growing cells at the time they were harvested.
Five fermentations of each bacterium were conducted.

Evaluation of Attractiveness of Bacterial Filtrates

Attractiveness of each bacterial filtrate was evaluated using cage-top bioassays as
described in Robacker & Flath (1995) with water as the control. The three culturing
media were also tested against water. Briefly, the bioassay was conducted by placing
two filter paper triangles containing 10 il of bacterial or growth-medium filtrate and
two papers containing 10 pl of water on the top of an insect cage. The filter papers
were raised 5 mm above the cage top using plastic rings. Each bioassay cage contained
180-200 flies. The number of flies beneath each filter paper was counted once each
minute for 10 min following application of the test materials to the papers. The 11 bac-
terial filtrates and the three growth-medium filtrates were tested in random order.
Two bioassay replications were conducted for each fermentation.
To analyze bioassay results, total flies counted at water control papers were sub-
tracted from total flies counted at treatment papers for each cage-top bioassay to ob-
tain a bioassay count difference. The two bioassay count differences per fermentation
were averaged. The resulting fermentation-level means were then used as data points
in one-way analysis of variance (ANOVA) using SuperANOVA (Abacus Concepts
1989) to compare attractiveness of the various bacteria. Means separations were con-
ducted by Fisher's protected least significant difference method (LSD).

Effects of pH on Attractiveness of Bacterial Filtrates

For one of the five fermentations, the pH of bacterial filtrates and growth-medium
filtrates was adjusted to 5, 7, and 9 with 85% phosphoric acid (Fisher Scientific, Fair
Lawn, NJ) or saturated sodium hydroxide (Fisher). Attractiveness of each pH treatment
was tested against water controls using cage-top bioassays. The purpose was to deter-
mine if attractive principals of the bacterial filtrates were nonionizing chemicals that
would not be affected by pH or chemicals that ionize into relatively nonvolatile forms
and therefore contribute little to attractiveness at certain pH's. Thus, carboxylic acids
(pKa's 4-5) would contribute little to attractiveness at pH 9; ammonia (pKa 9.2) and
amines aliphaticc amines, pKa's 10-11) would contribute little to attractiveness at pH 5;

Florida Entomologist 81(4)

December, 1998

imines (1-pyrroline, pKa 6.7) would be most attractive at pH > 7; and pyrazines (pyra-
zine, pKa 0.6) and nonionizing compounds would be volatile and attractive throughout
the pH range tested (pKa's from March 1968, Amoore et al. 1975, Weast 1976).
Each replication of the experiment consisted of one cage-top bioassay for each of
the three pH treatments of all filtrates, tested in random order. Ten replications of the
experiment were conducted. Each bacterial filtrate or growth medium filtrate was an-
alyzed separately by one-way ANOVA to compare pH effects. Bioassay count differ-
ences (described above) were used as data and pH means were separated by Fisher's
protected LSD.

Volatiles Sampling

Chemicals were sampled in the headspace above filtrates of bacteria and uninoc-
ulated growth media by solid phase microextraction (SPME) with a 100 pm polydim-
ethylsiloxane-coated fiber (Supelco, Inc., Bellefonte, PA). The fiber was inserted
through a septum into the headspace above 1 ml of filtrate in a 4 ml vial for 30 min
at 21-23C before analysis by GC or at 25-27C before analysis by GC-MS.

Chemical Identifications

Two methods were used to identify chemicals. For bacteria that had volatiles pro-
files similar to those of three species of bacteria studied previously (Robacker & Flath
1995, Robacker & Bartelt 1997), chemicals were identified by matching GC retention
times and detector response ratios with those of standards. The gas chromatograph
was a Shimadzu GC-17A (Shimadzu Scientific Instruments, Inc., Columbia, MD) with
flame ionization (FID) and flame thermionic (FTD) (Model FTD-17) detectors. A DB-
1 capillary column (J & W Scientific, Folsom, CA) with a 5 pm film was used. FTD/FID
response ratios were obtained to establish the presence of C-N bonds. A detailed de-
scription of the GC method can be found in Robacker & Bartelt (1997).
For bacteria that had nitrogen-containing peaks not observed in previous studies,
chemicals were identified by GC-MS. GC-MS data were acquired using a Hewlett
Packard 5890 GC with a HP 5970 mass selective detector (electron energy = 70 eV).
GC-MS identifications were based on computer matching of unknown spectra with
those in the Wiley 138K Mass Spectral Database (John Wiley and Sons, New York).
Identifications were authenticated by comparing spectra with those of standards for
most chemicals. A DB-1 column with a 5 pm film also was used. A detailed description
of the GC-MS method can be found in Robacker & Bartelt (1997). Chemicals were
sampled from headspace above unaltered filtrates, and filtrates to which sodium hy-
droxide was added to enhance volatilization of basic compounds.

GC Analysis of Headspace Volatiles

Relative amounts of eight attractive chemicals in the headspace of bacterial fil-
trates (at unaltered pH of filtrates) and the 3 uninoculated growth media filtrates
were measured. This analysis was conducted to compare amounts of the various
chemicals produced by different bacteria taxa.
Analyses of seven nitrogen-containing chemicals were conducted using the Shi-
madzu GC-17A with FTD as described above. GC peak heights were measured using
Millennium 2010 Chromatography Manager software (Waters Corporation, Milford,
MA). Two headspace analyses were conducted for each of five fermentations of the 11
bacteria strains and the three growth media. One headspace analysis per fermenta-
tion was also done using FID for acetic acid.

Robacker et al.: Bacteria Attractive to Mexican Fruit Fly 501

Peak heights from the 2 GC-FTD analyses were averaged to give a fermentation-
level mean. A one-way ANOVA was conducted for each chemical to assess amounts in
headspace above the various filtrates, using the fermentation-level means as data
points. ANOVA was also conducted to analyze individual GC-FID peak heights (not
means) of acetic acid in the various filtrates. Means were separated by Fisher's pro-
tected LSD.

GC Standards

Ammonium carbonate, 2-methylpropanamine, 2-methylbutanamine, 3-methylbu-
tanamine, 2-phenylethanamine, pyrazine, methylpyrazine, 2,3-dimethylpyrazine, 2,5-
dimethylpyrazine and trimethylpyrazine were obtained from Aldrich Chemical Com-
pany, Inc. (Milwaukee, WI). Methylamine HC1, trimethylamine HC1, and cyclohexy-
lamine were obtained from Sigma Chemical Company (St. Louis, MO). Acetic acid was
obtained from Fisher Scientific (Pittsburgh, PA) and 2-methylpropanoic acid and 3-me-
thylbutanoic acid were obtained from Eastman Chemical Products, Inc. (Kingsport, TN).
Five imines were synthesized. 1-pyrroline was synthesized by acid hydrolysis of 4-
aminobutyraldehyde diethyl acetal (Aldrich) according to methods of Schopf & Oechler
(1936). 2, 3, 4, 5-tetrahydropyridine was synthesized by reaction of N-chlorosuccinim-
ide (Aldrich) with piperidine (Matheson, Coleman & Bell, Norwood, OH, 98%) to form
N-chloropiperidine, followed by elimination of HC1 from N-chloropiperidine with KOH
(Fisher) (Quick & Oterson 1976). Other imines were prepared by the general method
of addition of aldehydes to primary amines (March 1968). N-isopentylidene-3-meth-
ylbutanamine was prepared by addition of 3-methylbutanal (Aldrich) to 3-methylbu-
tanamine in methylene chloride (Fisher) at room temperature. Anhydrous sodium
sulfate (EM Science, Cherry Hill, NJ) was then added to clear turbidity due to water
formed as a reaction byproduct. Likewise, N-phenylmethylene-2-methylpropanamine
and N-phenylmethylene-3-methylbutanamine were prepared by addition of benzalde-
hyde (Aldrich) to 2-methylpropanamine and 3-methylbutanamine, respectively.


Attractiveness of Bacterial Filtrates

All bacterial filtrates were significantly more attractive than uninoculated media
(F = 7.8; df = 13,56; P < 0.0001) (Fig. 1). No major differences in attractiveness oc-
curred among the bacteria strains except that the B. thuringiensis group was gener-
ally less attractive than the others. In this work and in previous studies (Robacker &
Flath 1995, Robacker & Bartelt 1997), cultures of many species, genera, and higher
taxa of bacteria have been demonstrated attractive to Mexican fruit flies. We conclude
that Mexican fruit fly attraction to bacteria cultured in aqueous laboratory media is
a general phenomenon.

Effects of pH on Attractiveness of Bacterial Filtrates

The pH's of bacterial filtrates, before manipulation with phosphoric acid or sodium
hydroxide, generally were between 7.8 and 9.2. These solutions contained more equiv-
alents of bases than acids. Exceptions were M. luteus and B. t. coreanensis that had
pH's of 7.1 and 5.3, respectively. The B. t. coreanensis filtrate contained more acids
than bases. Uninoculated growth media had pH's between 6.7 and 7.0.

Florida Entomologist 81(4)

December, 1998

Sde de de
e cde bcde
7 bde
0 6- bcd

S 5-


, 2-

% I I I I i I I I I

Fig. 1. Attractiveness of bacterial filtrates toA. ludens in cage-top bioassays. Rel-
ative attractiveness = mean counts at papers containing filtrates, divided by mean
counts at papers containing water. Bars with the same letter are not significantly dif-
ferent from each other by Fisher's protected LSD (P < 0.05).

Attractiveness of all filtrates was greatly affected by changing filtrate pH (small-
est F = 9.2; df = 2,27; P < 0.001 for nutrient broth) (Fig. 2). Most filtrates at pH 7 and
all at pH 9 were more attractive than filtrates at pH 5. This is a critical result because
it indicated that the most important attractive principals are compounds containing
protonizable nitrogen with pKa's of 7 or above because these chemical classes would
be largely ionized and nonvolatile at pH 5.
All filtrates at pH 5 except E. cloacae and A. f faecalis were significantly more at-
tractive than water controls, although attractiveness of most was not high compared
with attractiveness at pH 7-9. However, B. popillae and Bt.t. coreanensis filtrates at
pH 5 were much more attractive than water controls (paired t-test for B. popillae, t =
7.5, df = 9, P < 0.001; for B. t. coreanensis, t = 6.6, df = 9, P < 0.001). Chemicals that
could account for the attractiveness of filtrates at pH 5 include carboxylic acids, pyra-
zines, and various nonionizing chemicals such as hydrocarbons, alcohols, aldehydes,
ketones, esters, etc., that would exist largely in nonionized, volatile forms at pH 5.

Chemical Identifications

Because all filtrates were most attractive at pH 9, chemical identifications were fo-
cused on chemicals containing protonizable nitrogen. The relatively low attractive-
ness of most filtrates at pH 5 indicated that nonionizing chemicals probably did not
play major roles in attractiveness and were not identified in this work. However, the
moderate attractiveness of the naturally acidic B. t. coreanensis filtrate also led us to
identify carboxylic acids from the bacterial volatiles.
Filtrates of A. f faecalis, B. popillae, and B. t. coreanensis were analyzed by GC-
MS. Nitrogen-containing chemicals and carboxylic acids that were identified are

Robacker et al.: Bacteria Attractive to Mexican Fruit Fly 503

12- A -- trpticase soy broth

10 E. cloacaeX
-t- A.f. faecali
8- -*- M. luteus

6- -0- B. sphericusNX
-0- B. subtilis"'
S 4 B. megaterium
2- B. popillae'

5 7 9
S 12 B
-- nutrient broth'
8- shandongiensis
o" 8-
S8 --A- B. t. konkukian'
6- B.t
4- -o- medium B

2- 0- B. t. coreanensisV

5 7 9

pH of Filtrate

Fig. 2. Attractiveness of pH altered bacterial filtrates toA. ludens in cage-top bio-
assays. Relative attractiveness = mean counts at papers containing filtrates, divided
by mean counts at papers containing water. k indicates significant difference in at-
tractiveness between pH 5 and 7; i indicates significant difference in attractiveness
between pH 7 and 9; P < 0.05 by Fisher's protected LSD.

shown in Table 1. Of these chemicals, trimethylamine, 2-methylpropanamine, 3-me-
thylbutanamine, 2-methylbutanamine, methylpyrazine, 2,5-dimethylpyrazine and
trimethylpyrazine had been reported from one or more of the bacteria Staphylococcus
aureus, Klebsiella pneumoniae, and Citrobacter freundii (Robacker & Flath 1995, Ro-
backer & Bartelt 1997). Lee et al. (1995) and DeMilo et al. (1996) identified numerous
pyrazines from headspace ofK. pneumoniae and C. freundii including most of those in
Table 1, and others. Acetic acid had been reported from headspace of S. aureus (Ro-
backer & Flath 1995) and 2-methylpropanoic acid and 3-methylbutanoic acid from

504 Florida Entomologist 81(4) December, 1998


Aff+ Bp+ Btc+
Aff NaOH Bp NaOH Btc NaOH




ethyldimethylpyrazine isomer2
diethylmethylpyrazine isomer2
acetic acid
2-methylpropanoic acid
3-methylbutanoic acid

- +
+ +
+ ++

++ +++
++ +++

+ +

+ ++


+ +4
+ + + + +
++ ++ ++ ++ ++ +4
- - + +
+ + + + ++ ++

+4 -
- - ++

Aff = A. f faecalis, Bp = B popillae, Btc = B t. coreanensis, = not detected above baseline, + = trace (< 500
area counts), ++ = minor (500 5000 area counts), +++ = major (> 5000 area counts).
'Good hbrary match, but not verified with standard.

headspace of K. pneumoniae (Lee et al. 1995, Robacker & Bartelt, 1997). Cyclohexy-
lamine, 2-phenylethanamine, N-isopentylidene-3-methylbutanamine, N-phenylme-
thylene-2-methylpropanamine, N-phenylmethylene-3-methylbutanamine, and 2,4,5-
trimethyl-3-oxazoline had not been reported from any bacteria, to our knowledge. N-
isopentylidene-3-methylbutanamine has been found in volatiles of NuLure, a protein
bait for fruit flies (Flath et al. 1989).
Additional chemicals were identified by GC-FID and GC-FTD. These were ammo-
nia, pyrazine, 1-pyrroline and 2,3,4,5-tetrahydropyridine. The latter 3 chemicals were
identified by the highly sensitive GC-FTD technique. They had been identified previ-
ously by GC-MS in headspace of other bacteria (Robacker & Flath 1995, Robacker &
Bartelt 1997). Ammonia was also identified from these other bacteria by GC-FID and
GC-FTD. It was not identified by GC-MS because of its low molecular weight. All four
chemicals were verified by GC analyses of standards.

Robacker et al.: Bacteria Attractive to Mexican Fruit Fly 505

Attractiveness of Chemicals Identified by GC-MS/GC Analyses

Many of the chemicals have been evaluated for attractiveness to Mexican fruit
flies (Robacker & Warfield 1993, Robacker & Flath 1995, Robacker et al. 1996, Ro-
backer & Bartelt 1997, Robacker et al. 1997). Ammonia, 2-methylpropanamine, 3-me-
thylbutanamine, 2-methylbutanamine and acetic acid were 2-3 times more attractive
than water controls. Trimethylamine, 1-pyrroline, 2,3,4,5-tetrahydropyridine, pyra-
zine, 2,5-dimethylpyrazine, and trimethylpyrazine were 1.1 to 1.5 times more attrac-
tive than water. Methylpyrazine, 2-phenylethananime, and 3-methylbutanoic acid
were not attractive. N-isopentylidene-3-methylbutanamine was not attractive to four
species of fruit flies that did not include the Mexican fruit fly in olfactometer tests
(Flath et al. 1989). The other chemicals in Table 1 have not been tested for attractive-
ness to fruit flies.

Comparison of Attractants Produced by Bacteria Taxa

Results of the GC-FTD analyses of 8 attractive components of bacterial volatiles
are shown in Tables 2 and 3. Ammonia was produced in about the same amounts by
all of the bacteria. Emission of most other chemicals varied greatly from strain to
strain. Some generalizations can be observed in the tables regarding production of
some chemicals by closely related taxa. For example, the only two bacteria that pro-
duced large amounts of 2-methylpropanamine and 3-methylbutanamine were in the
genus Bacillus. However, the other three species of Bacillus produced very little of
these two chemicals. Thus, chemicals were not produced in similar amounts by re-
lated taxa in many cases. In other cases, chemicals were produced by distantly related
taxa but not by closely related ones. An example of this is trimethylpyrazine that was
produced in relatively high amounts by E. cloacae and B. t. coreanensis but in lower
amounts by other Bacillus and even other strains ofB. thuringiensis.
Some of the differences in volatiles production may be attributable to differences
in culturing media, but in other cases, bacteria grown on different media produced the
same chemicals. For example, highest amounts of 2, 5-dimethylpyrazine were pro-
duced by bacteria cultured on trypticase soy broth and highest amounts of trimethy-
lamine were produced by the B. thuringiensis strains cultured on nutrient broth. On
the other hand, the trimethylpyrazine example discussed above is a case in which two
bacteria grown on different media produced about the same amount of a chemical.
The discussion of similarities and differences in volatiles profiles can be expanded
by including results of previous analyses of bacteria volatiles. Profiles of K. pneumo-
niae and C. freundii (Robacker & Bartelt 1997), members of the family Enterobacte-
riaceae along with E. cloacae, differed from the profile of E.oacae (Tables 2 and 3)
in amounts of 3-methylbutanamine, 2, 5-dimethylpyrazine and trimethylpyrazine
but were similar with regard to several other chemicals. Also, amounts of trimethy-
lamine, 3-methylbutanamine and acetic acid produced by S. aureus (Robacker &
Flath 1995), a member of the Micrococcaceae along with M. luteus, differed dramati-
cally from amounts produced by M. luteus (Tables 2 and 3). Conversely, acetic acid
production by S. aureus was high as in B. t. coreanensis, a species that is not in the Mi-
crococcaceae. These examples suggest a great diversity of metabolic pathways in bac-
teria that do not tie closely to currently held views of taxonomic relatedness.
Note that peak sizes do not reflect absolute amounts of different chemicals. For ex-
ample, the small ammonia peaks indicate filtrate concentrations in the 100 Pg/ml to
1 mg/ml range while the large 2-methylpropanamine peaks indicate concentrations
only in the 1 to 10 pg/ml range (Robacker & Bartelt, 1997).

Florida Entomologist 81(4)

December, 1998


trimethyl- 2-methyl- 3-methyl-
ammonia amine propanamine butanamine

trypticase soy broth 0.14 abc 0.1 a 0.2 a 0.1 a
E. cloacae 0.30 cde 0.6 a 0.4 a 0.2 a
A. f. faecalis 0.30 cde 0.4 a 0.9 a 0.6 a
M. luteus 0.29 cde 0.6 a 1.0 a 0.7 a
B. sphericus 0.32 de 0.5 a 126.1 b 61.7 b
B. subtilis 0.35 de 0.8 a 3.2 a 1.4 a
B. megaterium 0.28 cde 0.7 a 0.8 a 0.4 a
B. popilliae 0.25 bcd 0.6 a 107.4 b 53.6 b
nutrient broth 0.05 a 1.7 a 0.4 a 0.2 a
B. t. shandongiensis 0.25 bcd 14.6 b 1.5 a 0.8 a
B. t. konkukian 0.39 de 29.7 c 0.4 a 0.3 a
B. t. darmstadiensis 0.44 e 34.2 c 0.5 a 0.3 a
medium B 0.09 ab 0.4 a 0.3 a 0.2 a
B. t. coreanensis 0.34 de 2.0 a 0.4 a 0.2 a

iFlame thermionic detection. For a given chemical, mean peak heights followed by the same letter were not
significantly different from each other by Fisher's protected LSD (P < 0.05, n = 5 fermentations.

Attractive Principals vs. pH of Filtrates

Experiments with filtrate pH indicated the importance of chemicals containing
protonizable nitrogen to the attractiveness of the bacterial filtrates (Fig. 2). Ammonia,
amines, imines, and pyrazines, all chemicals previously demonstrated attractive to
Mexican fruit flies, were then identified from the filtrates (Tables 1-3). Also, these
chemicals had been identified previously as the attractive principals of three other
bacteria (Robacker & Flath 1995, Robacker & Bartelt 1997) and nonionizing chemi-
cals identified from odor of two of those bacteria were not attractive to flies primed for
response to bacterial odor. We conclude that the attractive principals of the naturally
basic bacterial filtrates tested in this work (pH 7.8 to 9.2), as well as all filtrates ad-
justed to pH 9, were ammonia, amines, imines and pyrazines.
Two bacteria, B. t. coreanensis and B. popillae, were also moderately attractive
at pH 5 (Fig. 2). The chemical most responsible for the moderate attractiveness of
these two filtrates at pH 5 probably was acetic acid. These two filtrates had the larg-
est peak heights for acetic acid (Table 3). As discussed above, acetic acid is very at-
tractive to Mexican fruit flies. Because the B. t. coreanensis filtrate was pH 5.3
before manipulation with phosphoric acid, we conclude that acetic acid played a ma-
jor role in attractiveness of this filtrate at its natural pH. Pyrazines may also con-
tribute to attractiveness of these and all filtrates at pH 5 because of their low pKa's.
Possible minor roles of nonionizing chemicals at all pH levels have not been deter-

Robacker et al.: Bacteria Attractive to Mexican Fruit Fly 507


2,5-dimethyl- trimethyl- acetic
pyrazine pyrazine pyrazine acid

trypticase soy broth 3.4 b 69.9 bcd 3.0 a 0.1 a
E. cloacae 5.1 bc 152.5 f 20.9 b 0.0 a
A. f.faecalis 4.0 b 112.1 def 4.4 a 0.1 a
M. luteus 5.4 bc 91.8 cde 6.2 a 0.0 a
B. sphericus 4.8 bc 160.2 f 6.3 a 0.2 a
B. subtilis 6.4 cd 161.5 f 7.1 a 0.2 a
B. megaterium 5.4 bc 128.4 ef 5.7 a 0.0 a
B. popilliae 6.2 cd 127.4 ef 7.7 a 0.6 a
nutrient broth 0.0 a 4.1 a 0.5 a 0.0 a
B. t. shandongiensis 0.9 a 48.2 abc 7.7 a 0.2 a
B. t. konkukian 3.9 b 37.5 ab 3.2 a 0.1 a
B. t. darmstadiensis 1.2 a 44.4 abc 3.4 a 0.1 a
medium B 8.1 de 7.9 a 0.7 a 0.0 a
B. t. coreanensis 8.6 e 67.7 bcd 24.5 b 3.0 b

iPyrazlnes determined by flame thermlonic detection; acetic acid determined by flame lomzation detection.
For a given chemical, mean peak heights followed by the same letter were not significantly different from each
other by Fisher's protected LSD (P < 0.05, n = 5 fermentations.

Novel Attractants from Bacterial Odors

Several novel chemicals fitting the attractive-principal profile were identified from
the four major bacteria taxa that were investigated. However, the principal differ-
ences among the bacteria were quantitative rather than qualitative in that they pro-
duced mostly the same chemicals but in widely different amounts, at least when
grown on laboratory media. Thus, an exhaustive investigation of bacteria species for
potential new attractants would likely result in relatively few candidate compounds.


We thank Maura Rodriguez and Cyndi Rodriguez for technical assistance. Use of
a product brand in this work does not constitute an endorsement by the USDA.


ABACUS CONCEPTS. 1989. SuperANOVA. Abacus Concepts, Inc., Berkeley, CA.
AMOORE, J. E., L, J. FORRESTER, AND R. G. BUTTERY. 1975. Specific anosmia to 1-pyr-
roline: The spermous primary odor. J. Chem. Ecol. 1: 299-310.
DEMILO, A. B., C. -J. LEE, D. S. MORENO, AND A. J. MARTINEZ. 1996. Identification of
volatiles derived from Citrobacter freundii fermentation of a trypticase soy
broth. J. Agric. Food Chem. 44: 607-612.

Florida Entomologist 81(4)

December, 1998

DREW, R. A. I. 1987. Behavioural strategies of fruit flies of the genus Dacus (Diptera:
Tephritidae) significant in mating and host-plant relationships. Bull. Entomol.
Res. 77: 73-81.
DREW, R. A. I., AND H. A. C. FAY. 1988. Comparison of the roles of ammonia and bac-
teria in the attraction of Dacus tryoni (Froggatt) (Queensland fruit fly) to pro-
teinaceous suspensions. J. Plant Prot. Tropics. 5: 127-130.
DULMAGE, H. T., J. A. CORREA, AND A. J. MARTINEZ. 1970. Coprecipitation with lac-
tose as a means of recovering the spore-crystal complex of Bacillus thuringien-
sis. J. Invertebr. Pathol. 15: 15-20.
1989. Effect of pH on the volatiles of hydrolyzed protein insect baits. J. Agric.
Food Chem. 37: 814-819.
Gow, P. L. 1954. Proteinaceous bait for the oriental fruit fly J. Econ. Entomol. 47:
Methyl mercaptan and dimethyl disulfide production from methionine by Pro-
teus species detected by head-space gas-liquid chromatography. J. Clin. Micro-
biol. 6: 187-194.
HOLT, J. G. 1984. Bergey's Manual of Systematic Bacteriology, Vol. I & II, 1599 pp.,
Williams & Wilkins, Baltimore, MD.
JARVIS, H. 1931. Experiments with a new fruit fly lure. Queensland Agric. J. 36:485-491.
LEE, C. -J., A. B. DEMILO, D. S. MORENO, AND A. J. MARTINEZ. 1995. Analysis of the
volatile components of a bacterial fermentation that is attractive to the Mexi-
can fruit fly,Anastrepha ludens. J. Agric. Food Chem. 43: 1348-1351.
MARCH, J. 1968. Advanced organic chemistry: reactions, mechanisms, and structure,
1098 pp., McGraw-Hill, New York.
MARTINEZ, A. J., D. C. ROBACKER, AND J. A. GARCIA. 1997. Toxicity of an isolate of Ba-
cillus thuringiensis subspecies darmstadiensis to adults of the Mexican fruit fly
(Diptera: Tephritidae) in the laboratory. J. Econ. Entomol. 90: 130-134.
MARTINEZ, A. J., D. C. ROBACKER, J. A. GARCIA, AND K. L. ESAU. 1994. Laboratory and
field olfactory attraction of the Mexican fruit fly (Diptera: Tephritidae) to me-
tabolites of bacterial species. Florida Entomol. 77: 117-126.
QUICK, J., AND R. OTERSON. 1976. A convenient synthesis of pelletierine (2-piperidyl-
propanone). Synthesis 1976: 745-746.
ROBACKER, D. C., AND R. J. BARTELT. 1997. Chemicals attractive to the Mexican fruit
fly from Klebsiella pneumoniae and Citrobacter freundii cultures sampled by
solid-phase microextraction. J. Chem. Ecol. 23: 2897-2915.
ROBACKER, D. C., AND R. A. FLATH. 1995. Attractants from Staphylococcus aureus cul-
tures for the Mexican fruit fly,Anastrepha ludens. J. Chem. Ecol. 21: 1861-1874.
ROBACKER, D. C., AND J. A. GARCIA. 1993. Effects of age, time of day, feeding history,
and gamma irradiation on attraction of Mexican fruit flies (Diptera: Tephritidae),
to bacterial odor in laboratory experiments. Environ. Entomol. 22: 1367-1374.
ROBACKER, D. C., A. B. DEMILO, AND D. J. VOADEN. 1997. Mexican fruit fly attracta-
nts: Effects of 1-pyrroline and other amines on attractiveness of a mixture of
ammonia, methylamine, and putrescine. J. Chem. Ecol. 23: 1263-1280.
ROBACKER, D. C., D. S. MORENO, AND A. B. DEMILO. 1996. Attractiveness to Mexican
fruit flies of combinations of acetic acid with ammonium/amino attractants
with emphasis on effects of hunger. J. Chem. Ecol. 22: 499-511.
ROBACKER, D. C., W. C. WARFIELD, AND R. F. ALBACH. 1993. Partial characterization
and HPLC isolation of bacteria-produced attractants for the Mexican fruit fly,
Anastrepha ludens. J. Chem. Ecol. 19: 543-557.
SCHOPF, C., AND F. OECHLER. 1936. Die Synthese des desoxyvasicins unter physiolo-
gischen bedingungen. Annalen der Chemie. 523: 1-29.
WEAST, R. C. 1976. Handbook of chemistry and physics, 57th ed., 2390 pp., CRC
Press, Cleveland, OH.

Taylor & McPherson: Voltinism in Merragata brunnea 509


1Center for Biodiversity, Illinois Natural History Survey, 607 East Peabody Drive
Champaign, Illinois 61820

2Department of Zoology, Southern Illinois University at Carbondale
Carbondale, Illinois 62901


Voltinism in Merragata brunnea Drake was studied in southern Illinois during
1989 and 1990. This species overwintered as adults, which became active in early
March. First instars were found from mid-May through mid-September, second in-
stars from mid-May through mid-October, third instars from early June through mid-
October, fourth instars from late May through early November, fifth instars from late
May through late October, and adults from early March through mid-November. The
sequences of peaks of nymphal instars and adults indicate that this species is bi- or
trivoltine in southern Illinois.

Key Words: Merragata brunnea, voltinism, southern Illinois, life history


Se estudi6 el voltinismo en Merragata brunnea Drake en el sur de Illinois durante
1989 y 1990. Merragata brunnea invernaron como adults, los que se activaron a prin-
cipios de marzo. Se encontraron instares del primer estadio desde mediados de mayo
a mediados de septiembre, del segundo estadio de mediados de mayo a mediados de
octubre, del tercer estadio de principios de junio a mediados de octubre, del cuarto es-
tadio de finales de mayo a principios de noviembre, del quinto estadio de finales de
mayo a finales de octubre, y adults de principios de marzo a mediados de noviembre.
La secuencia de crestas de instares ninfales y de adults indica que esta especie es bi-
o trivoltina en el sur de Illinois.

The velvet water bug Merragata brunnea Drake, based on scattered records, oc-
curs from New Jersey south to Florida, and west to Minnesota, Nebraska, and Texas;
it also occurs in "S. Canada" (Polhemus and Polhemus 1988). In Illinois, it has been
found only in the southernmost counties (i.e., Alexander, Jackson, Johnson, Massac,
and Union) (Taylor 1996).
Little has been reported on this insect's life history. It occurs in a variety of habi-
tats (e.g., lakes, ponds, swamps, roadside ditches, and rivers) and often is associated
with floating vegetation (Taylor 1996). It has been collected in various months from
April to November in New Jersey (Chapman 1959), Minnesota (Bennett and Cook
1981), Illinois (Taylor 1996), Missouri (Froeschner 1949), and Mississippi (Wilson
1958); and during most of the year from January to December in Florida (Chapman
1958). Porter (1950) noted that of 127 adults he examined from across much of the
range of the species, 5 were collected in April, 4 in May, 84 in July, and 34 in August.

Florida Entomologist 81(4)

December, 1998

Wilson (1958) reported that of 65 specimens he examined from Mississippi, 1 had
been collected in May, 2 in July September, 60 in September, and 2 in October.
Both brachypterous and macropterous adults have been reported. Adults prima-
rily are brachypterous in Florida (Chapman 1958), Minnesota (97.5% of 81 adults;
Bennett and Cook 1981), and Wisconsin (97% of 301 adults; Hilsenhoff 1986).
Porter (1950) reared this species in the laboratory and briefly described the imma-
ture stages. He reported the incubation period and stadia for the egg and first through
fifth instars as 8-12, 3-6, 3-4, 3-4, 5-6, and 5-6 days, respectively.
During 1989 and 1990, we studied voltinism in a population of this species at Pres-
ident's Pond on the campus of Southern Illinois University at Carbondale, Jackson
County, Illinois (see Taylor [1996] for detailed description of pond). President's Pond
is a roughly triangular 0.29 hectare (0.71 acre) pond. It is connected at the northern
end to the adjacent Lake on the Campus by a narrow channel (approximately 2-5 m
wide, 2 m deep). Along the eastern shore (where the present study was conducted),
water depth increased sharply between 1 and 2 m from shore and commonly exceeded
2 m at 2.5 m from shore.
Floating, emergent, and shoreline vegetation associated with the pond was diverse
(Taylor 1996). The western margin was bordered by a dense, but narrow, band of cat-
tails (Typha angustifolia L.). The southern border was comprised of a riprap dam cov-
ered with soil and crossed by a paved road. The eastern margin was bordered by
overhanging trees and other vegetation. During the summer, the pond filled with a
dense growth of aquatic vascular plants and filamentous algae. Near the shoreline,
and wherever the aquatic plants reached the water surface, duckweeds built up into
dense mats. Air currents tended to move the duckweeds (i.e., Lemna minor L.,
Spirodela polyrhiza (L.) Scheiden, and Wolffia papulifera Thompson) around the pond
unless the plants were partially anchored in the underlying aquatic vegetation.
This paper presents information on voltinism in M. brunnea, including times of oc-
currence of the adults and nymphal instars.


Samples were collected weekly from 18 March to 25 November 1989, and biweekly
from 11 February to 2 December 1990, along the eastern shore. Sampling was con-
fined to this area because (1) the cattails along the western shoreline prevented use
of the quadrat sampler (see below); (2) the riprap shoreline of the southern border was
unnatural and, often, disturbed by fishermen; and (3) the water surface along the
eastern shore, which was a mosaic of open water, duckweeds, and emergent stems,
supported a diverse gerromorphan fauna.

15 1st 2n4th 5th


0.100 0.125 0.150 0.175 0.200 0.225 0.250 0.275 0.300 0.325 0.350
mesotibial length (mm)

Fig. 1. Approximate instars of M. brunnea (n = 340), as delineated by mesotibial
length. Specimens collected in 1989 and 1990 from President's Pond, Southern Illinois
University at Carbondale campus, Jackson County.

Taylor & McPherson: Voltinism in Merragata brunnea 511

u) 25-
C 5 (N=26)
u 50-
H 25-
z 3RD
2 (N=34)



Fig. 2. Percent of individuals in each stage per sample of M. brunnea collected at
President's Pond, Southern Illinois University at Carbondale campus, Jackson
County, during 1989. Beginning and end points of each shaded area represent sample
dates preceding and following collection of specimens, respectively.

Four transects, 60 m in length, were made parallel to a relatively uniform section of
the eastern margin at 0, 0.5, 1.0, and 1.5 m from the shoreline. Each sample was col-
lected with a floating quadrat sampler (0.25 x 0.25 x 0.05 m), with four replicates placed
randomly along each transect; the resulting 16 quadrat samples, which provided a
broad sampling of the habitat, were then pooled. Prior to each sample, the collector
(SJT) stood for approximately three minutes to allow the insects to acclimate to the dis-
turbance; then, the sampler was placed on the surface of the water. Specimens were re-
moved with a fine mesh nylon net, preserved in alcohol, and sorted in the laboratory.
Adults could be distinguished from nymphs by their well-developed external gen-
italia and the presence of wings, even in the brachypterous form. Nymphal instars
were difficult to separate because they are small and show little progressive change

512 Florida Entomologist 81(4) December, 1998

50- -- -- -- --- -- ----------- -- -- ---




2 (N=34)

F 25-
z 3RD


S (N-=1)

Fig. 3. Percent in each sample of total individuals of same stage ofM. brunnea col-
lected at President's Pond, Southern Illinois University at Carbondale campus, Jack-
son County, during 1989. Beginning and end points of each shaded area represent
sample dates preceding and following collection of specimens, respectively.

in external characters during development.However, we found that mesotibial length
was a useful character for distinguishing instars, although separation between in-
stars was not complete (Fig. 1).


In southern Illinois, this species overwintered as adults, which were active from
early March through mid-November (Figs. 2-5). Some adults were collected in mid-
February in 1990, indicating that the species can be active during warm spells in win-
ter. No eggs were collected. First instars were found from mid-May through mid-Sep-
tember, second instars from mid-May through mid-October, third instars from early

__ _ _

Taylor & McPherson: Voltinism in Merragata brunnea 513



25- N=2)

W 25-

Lwi (N=18)

25 (N=28)

Fig. 4. Percent of individuals in each stage per sample of M. brunnea collected at
President's Pond, Southern Illinois University at Carbondale campus, Jackson
County, during 1990. Beginning and end points of each shaded area represent sample
dates preceding and following collection of specimens, respectively.

June through mid-October, fourth instars from late May through early November, and
fifth instars from late May through late October.
Merragata brunnea is bi- or trivoltine in southern Illinois. Most fifth instars of the
first generation became adults in June, and first instars of the second generation were
found in July. The second generation apparently reached adults in late July and Au-
gust. The second and third generations were not readily distinguishable but a third
generation apparently occurred, with fifth instars found in September and October
and the resulting adults appearing shortly thereafter. It cannot be determined
whether these generations corresponded to the apparent numerical peaks in July and
September reported by Porter (1950) and Wilson (1958), respectively.


Florida Entomologist 81(4)







Fig. 5. Percent in each sample of total individuals of same stage of M. brunnea col-
lected at President's Pond, Southern Illinois University at Carbondale campus, Jack-
son County, during 1990. Beginning and end points of each shaded area represent
sample dates preceding and following collection of specimens, respectively.

Of the 850 adults collected during this study, 474 were males and 376 were fe-
males; of these, 844 were micropterous (6 6, 99.6%, n=472; 2 9, 98.9%, n=372), and
six were macropterous, thus corroborating the findings of Bennett and Cook (1981),
Chapman (1958), and Hilsenhoff (1986). The six macropterous adults were collected
in April (1 Y), July (2 Y Y), and August (2 6 6, 1 2).

We thank the following for their critical reviews of this manuscript: R. A. Brandon,
J. A. Beatty, B. M. Burr, Department of Zoology; D. Ugent, Department of Plant Biol-
ogy, Southern Illinois University at Carbondale; and D.W. Webb and M.J. Wetzel, Cen-
ter for Biodiversity, Illinois Natural History Survey.


-L- _1 I I~]

December, 1998





Coss-Romero & Pefa: Broad Mite and Injury Levels in Pepper 515


BENNETT, D. V., AND E. F. COOK. 1981. The semiaquatic Hemiptera of Minnesota
(Hemiptera: Heteroptera). Minnesota Agric. Exp. Stn. Tech. Bull. 332: 1-59.
CHAPMAN, H. C. 1958. Notes on the identity, habitat and distribution of some semi-
aquatic Hemiptera of Florida. Florida Entomol. 41: 117-124.
CHAPMAN, H. C. 1959. Distributional and ecological records for some aquatic and
semi-aquatic Heteroptera of New Jersey. Bull. Brooklyn Entomol. Soc. 54: 8-12.
FROESCHNER, R. C. 1949. Contributions to a synopsis of the Hemiptera of Missouri,
Pt. IV. Hebridae, Mesoveliidae, Cimicidae, Anthocoridae, Cryptostemmatidae,
Isometopidae, Meridae (sic). Am. Midland Nat. 42: 123- 188.
HILSENHOFF, W. L. 1986. Semiaquatic Hemiptera of Wisconsin. Great Lakes Ento-
mol. 19: 7-19.
POLHEMUS, J. T., AND D. A. POLHEMUS. 1988. Family Hebridae Amyot and Serville,
1843, pp. 152-155. In T. J. Henry and R. C. Froeschner [eds.], Catalog of the
Heteroptera, or true bugs, of Canada and the continental United States. E. J.
Brill, New York. 958 pp.
PORTER, T. W. 1950. Taxonomy of the American Hebridae and the natural history of
selected species. Ph.D. Dissertation, University of Kansas, Lawrence. 185 pp. +
12 plates.
TAYLOR, S. J. 1996. Habitat preferences, species assemblages, and resource partition-
ing by Gerromorpha (Insecta: Heteroptera) in southern Illinois, with a faunal
list and keys to species of the state. Ph.D. Dissertation, Southern Illinois Uni-
versity at Carbondale, Carbondale. xviii + 345 pp.
WILSON, C. A. 1958. Aquatic and semiaquatic Hemiptera of Mississippi. Tulane Stud.
Zool. 6: 115-170.


Coss-Romero & Pefa: Broad Mite and Injury Levels in Pepper 515


'Universidad Autonoma de Chiapas, P.O. Box 34, Tapachula, Chiapas, Mexico

2University of Florida, Tropical Research & Education Center, 18905 S.W. 280th
Street, Homestead, Florida 33031


The responses of broad mite, Polyphagotarsonemus latus Banks (Acarina: Tarson-
emidae), were studied on four phenological stages of pepper plants: vegetative (V),
blossoming (B), early fruiting (EF) and late fruiting (LF) stages. All stages of the mite
preferred the undersides of the leaves to the uppersides. Plants in V, B, and EF stages
had higher numbers of mites per cm2 of foliage than plants in the late fruiting stage.
A damage index scale (0-6) was developed to assess broad mite injury to pepper
plants. Eight to nine cumulative mite days/cm2 were needed to reach a damage index
equal to 3 for V, B and EF plant stages. The damage index was also used to relate

Florida Entomologist 81(4)

December, 1998

broad mite injury to leaf area, height, water content, number of leaves, flowers, buds,
fruits and fruit weight of plants infested at four different phenological stages. Plants
infested when 14 weeks old (late fruiting stage), had less damage, significantly higher
number of fruits and fruit weight than plants infested at earlier plant stages, i.e., veg-
etative, flowering or early fruiting. The relationship between the damage rating (x)
and fruit numbers per plant (y,) and fruit weight in grams (y,) was given by y, = 2.83-
0.45x and y, = 232.5-37.234x, respectively.

Key Words: Polyphagotarsonemus latus, Capsicum annum, injury levels, green pep-


Se estudi6 la respuesta de cuatro estados fenol6gicos vegetativeo (V), floraci6n (F),
fruta pequena (EF) y fruta madura(LF)) de piment6n verde a el ataque del acaro
blanco, Polyphagotarsonemus latus Banks (Acarina: Tarsonemidae). Los acaros prefi-
rieron el env6s a el haz de las hojas. Aquellas plants en estado V, F y EF mantuvieron
un numero mas alto de acaros por cm que aquellas plants en estado LF. Se establecio
una escala de dano del 0 al 6 para evaluar la acci6n del acaro blanco en las plants de
piment6n. Un rango de 8 a 9 dias cumulativos de acaros son necesarios para causar
un nivel de dano igual a 3 en plants en estado V, F y EF. Las plants en estado LF
necesitan 6 dias-acaros /cm2para alcanzar un nivel de dano igual a 1. La escala de
dano fu6 tambien utilizada para relacionar el dano causado por los acaros y el area fo-
liar, altura de la plant, contenido liquid, numero de hojas, flores, yemas, frutos, y
peso de frutos por plants en los 4 estados fenol6gicos. Aquellas plants infestadas
cuando tienen 14 semanas, presentaron menor dano y un numero significativamente
mayor de frutas y peso de fruta, que plants en estado V, B y EF. La relaci6n entire la
escala de dano (x) y el numero de frutas (y,) y el peso de fruta en gramos (y2) esta des-
crita por la ecuaci6n: y = 2.83-0.45x y y = 232.5-37.234x, respectivamente.

Outbreaks of plant feeding tarsonemid mites often occur in vegetative, blossoming
or early fruiting stages in the host plant (Jeppson et al. 1975). Because of the tarson-
emid's short generation time (approx. 5 days), high fecundity, small size and protected
habitat, the injury it produces is often confused with diseases and phytotoxicity Jepp-
son et al., 1975, Aubert et al. 1981, Cross and Bassett 1982). The impact of the broad
mite, Polyphagotarsonemus latus (Banks) feeding has been qualitatively described for
cotton, cucumber, potatoes, tomatoes, gerberas, beans, papaya, and pepper (Aubert et
al. 1981, Bassett, 1981, Beattie & Gellatley 1983, Cross & Basset 1982, Hooper 1957,
Laffi 1982, Lo & Chao 1972, Pena & Bullock 1994, Schoonhoven et al. 1978, Jeppson
et al. 1975, Ochoa et al. 1994). While these observations suggest a causal relation to
host phenology, quantitative assessment of actual impact of feeding by broad mite on
growth, leaf area and yield is apparently not well correlated with levels of visible in-
jury and with broad mite densities (Dhoria and Bindra, 1977, Jones & Brown 1983,
Pena, 1990). We observed in commercial pepper, Capsicum annuum L., that rapid in-
creases of broad mite numbers coincided with early stages of the plant. However, un-
der field conditions it is difficult to determine whether enlarged broad mite
populations on early vegetative or reproductive host plant stages resulted from an en-
hanced mite growth rate compounded over time, or from immigration from outside

Coss-Romero & Pefa: Broad Mite and Injury Levels in Pepper 517

Our study was designed to (1) determine under greenhouse conditions the re-
sponse of mite populations to host phenology in pepper and (2) to measure the impact
of mite density on total yield, fruit number, number of leaves and flowers of different
developmental stages of pepper plants.


Pepper "Early Calwonder"was grown to the stages desired for testing in 3.781 plas-
tic pots. Plants were fertilized with 20-20-20 NPK, plus micronutrients. Treatments
consisted of 15 plants at the vegetative stage(V) (ca. 5 weeks old), 15 plants at the
blossoming stage (B) (ca. 7 weeks-old), 15 plants at the early fruiting stage (EF) (ca.
10 weeks old) and 15 plants at the late fruiting stage (LF) (ca. 14 weeks old). Treat-
ments and the untreated controls were replicated 4 times and arranged in a random-
ized complete block design in a greenhouse maintained at 26 + 2C; 75-89% RH.
Twelve adult broad mite females from a colony maintained on "Podsquad" garden
bean plants, were placed on 2 apical leaves of each treated plant. One leaf was col-
lected 4 days after exposure and thereafter every 4th day until 50 days after exposure
from each pepper plant and the number of mites per cm2 determined under a micro-
scope. Levels of mite populations were measured in cumulative mite-days per cm2
with 1 mite-d defined as one mite (any motile stage) per leaf for 1 d. Broad mite days
were calculated as the sum of the two successive counts (mean number of mites/cm2)
divided by two and multiplied by the number of days between evaluations, and then
summed over the evaluation period. It was assumed that the amount of physical in-
jury (removal of cell contents) increased with the number of broad mite days; thus, the
term is used interchangeably with injury. Leaf area, fresh weight, dry weight of
leaves, were measured weekly during the growing season. Leaf area was determined
with a leaf area meter (LI-COR, Lambda Instruments Corporation, Lincoln, NE) and
water content was determined by subtracting leaf dry weight from leaf fresh weight.
Amount of vegetative growth was determined by dividing the dry leaf weight by the
total leaf area.

Damage Index Among Different Plant Ages:

A second experiment consisted of 30 broad mite-infested plants and 30 uninfested
plants at four phenological ages (V, B, EF, LF), where the number of leaves, buds, flow-
ers, fruit, fruit weight and mite days and damage index per plant was assessed
weekly. To establish a damage index per plant, plants were separated into 6 categories
of damage. The damage categories were defined as follows: Category 0.5: apical leaves
have begun to curl; the mid vein has become sinuous and the color of the leaves has
changed from shiny green to opaque green. Category 1: the mesophyl of the leaf un-
dersides sunken; the basal portion of the apical leaves showed a light green color and
the apical leaves curled down. Category 2: a bronze color is present in the apical
leaves. If the apical leaves were large, bronzing was a characteristic of the leaf base.
If the leaves were small, the leaves were completely bronzed and their tips necrosed.
Floral buds were necrosed; leaf area was reduced and damage was observed in axil-
lary leaves. Category 3: petioles of apical leaves have elongated and are thicker than
uninfested ones and when stems were necrosed and bronzed, necrosis and/or hyper-
trophy were observed also in floral buds. Category 4: apical and lateral floral buds
have proliferated but are deformed and hypertrophied and have failed to develop.
Category 5: apical leaves have become necrotic and necrotic floral buds have aborted.
A damage pattern similar to categories 1 to 4 is observed in lateral leaves and floral

Florida Entomologist 81(4)

December, 1998

buds. Category 6:apical and lateral leaves show lignification, floral buds and flowers
are hypertrophied and new leaves are necrotic. New leaves are observed but they
show symptoms similar to categories 1 and 4. No fruits are observed or if present, they
are deformed. Differences in these categories for infested and uninfested plants were
determined by student-t-test (SAS 1987) and analysis of variance (ANOVA) was used
to determine differences among plant stages.

Damage Index and Yield Reduction:

Fruit yields were determined by harvesting and weighing all peppers grown from
each plant and calculating the total weight of the fruit per plant. Regression analysis
was used to determine if there was a relationship between broad mite days (x) and
damage index (y). Data were combined over four plant stages. The fruit weight (de-
pendent variable) was also regressed on the damage index (independent variable). To
establish injury levels, a linear regression model was used i.e.,y = a + bx, where y is
the percentage of fruit weight reduction per plant, x is the damage index per plant.


Mite Dynamics Related to the Abaxial and Adaxial Leaf Surfaces:

Mites were first observed on the underside basal portion of the leaf near the mid
and lateral veins. In general, number of broad mites were significantly higher 4 to 14
days after infestation; thereafter, densities remained below 93 broad mites per leaf (F
= 2.78; P = 0.0073; df = 11, 47; N = 59). Oviposition was first observed near the mid
and secondary leaf veins but later continued at random on the leaf. During the first 8
days following infestation, higher P. latus densities were observed on the leaf under-
side. Maximum oviposition rate and maximum number of immature and mature
stages were observed on the eighth day after infestation (Fig. 1, 2). When the number
of eggs was > 200/cm2 on the leaf underside, the number of females increased on the
leaf upperside (Fig. 1A). This population trend might suggest that females search for
a new habitat after the carrying capacity of the preferred habitat on the leaf under-
side has been reached. However, from the 19th through the 73rd day of the evaluation
period, the proportion of broad mites on the leaf underside was always higher than on
the leaf upperside (Fig 1, 2). This pattern may be due to the propensity of the mites
to avoid sunlight or to avoid parts of the plant with low humidity. High light intensity,
low humidity and high temperature combinations are unfavorable for P. latus (Jepp-
son et al. 1975, Jones & Brown 1983).

Mite Dynamics Related to Plant Age:

The number of mites feeding varied for different plant stages (F = 8.54;P = 0.0001;
df = 3, 226; N = 230)(Table 1). Even though the same number of females were allocated
per plant, higher numbers of eggs deposited per cm2 were observed on vegetative(V),
blossoming (B) and early fruiting (EF) plant stages than on late fruiting stages (LF)
(Table 1). The male to female ratio varied from 5: 1 for 5 week-old plants (V) to 3: 1 for
the 7-14 week old plants (B, EF, LF). Mite populations on plants in V to EF stages in-
creased significantly more than populations on LF stages of pepper (Fig. 3). These re-
sults strongly indicate that P. latus responds to some physiological change in the late

Coss-Romero & Perfa: Broad Mite and Injury Levels in Pepper 519


7 Upper
E Lower


S Upper

35- A
W 25

m 15 -


z 4



4 8 14 19 22 27 33

T ~ I-

40 46 54 62
40 46 54 62

Days After Treatment
Fig. 1. Mean number of P. latus female(A) and male (B) on the upperside and un-
derside of green pepper leaves recorded for 73 days after infestation.

fruiting stage of the plant. Peaks of mite abundance were observed during the first 14,
20, 7 and 20 days following infestation of vegetative, blossoming, early fruiting and
late fruiting plants, respectively (Fig. 2).
Early fruiting stage plants had the highest numbers of mites/cm2 and late fruiting
plants had the lowest number of mites/cm2. There were no significant differences in
mites/cm2 for V and B plant stages (Table 1). Apparently, tarsonemid mouthpart ap-
pendages are unsuitable for effective penetration of renitent tissues (Jeppson et al.
1975). Thus P. latus may not be able to puncture the more lignified tissues found in 14
week-old plants as opposed to those tissues in 5-10 week old plants. These data may


Plant Age Plant Mites

(weeks) Stage /cm2 Eggs Nymphs Female Male

5 V 23.87b 42.68a 10.79b 7.92ab 2.05b
7 B 24.19b 39.84a 11.22b 7.36ab 2.64b
10 EF 45.13a 52.27a 23.38a 13.29a 4.88a
14 LF 9.60c 4.57b 6.49b 1.34c 0.51c

Numbers within the same column followed by the same letter were not significantly different (P > 0.05).

PIrt~p! s


520 Florida Entomologist 81(4) December, 1998

300 A T
250 7 E0 Upper
200- 1 1 Lower
150 -i

50 B
50-' B T ---
40 I Upper
0 RM Lower
,- 30- |-
Z 0

5O I -

4 8 14 19 22 27 33 40 46 54 62 73
M 200 C_ Upper
150 Lower
100 5 7
50 -3

4 8 14 19 22 27 33 40 46 54 62 73
Days After Treatment

Fig. 2. Mean number of P. latus prelarvae (A), nymphs(B) and eggs (C) on the up-
perside and underside of green pepper leaves recorded for 73 days after infestation.

be of value in programs for evaluating resistance of peppers to P. latus. Thus, assess-
ments of plant resistance to P. latus made at early growth stages of peppers would be
particularly effective for identifying highly resistant plants.

Relationship between damage index and mite-day/cm2:

The greenhouse experiment indicated that 9.24, 8.24 and 9.24 cumulative mite
days/cm2 are needed for 5, 7 and 10 day-old plants (V, B, EF stages) to reach an aver-
age damage index equal to 3.59, 2.56 and 3.02, respectively, whereas 6.31 mite days/
cm2 are necessary to reach a damage index equal to 0.98 for 14 week-old plants (LF
stage) (Table 3).
Broad mites significantly reduced the increment in leaf sizes of infested plants
compared to the control (Table 4). Fresh leaf weight was reduced for all plant stages,
but significant reductions in dry weight were observed only for V and B plant stages.
The levels of significance associated with soluble solids were reduced for V, B and EF
plant stages. (Table 4). Table 5 shows that the numbers of leaves per plant, plant
heights and numbers of fruit per plant were also affected by mite injury during all
plant stages. The data suggest that broad mites reduce height in the infested plants,
and that they induce lateral shoot growth (Table 5). The number of flowers and buds
in V, B and EF plant stages was significantly reduced compared with the uninfested
check plants, but corresponding reductions were not observed on LF plants (Table 5).

__ _ _


Cumulative Number of Mite Days

2 7 14 20 26 33 46 50 DAI

Average Damage


5 V 1.40 5.08 7.42 8.12 8.14 8.99 9.13 9.24 3.59
7 B 1.33 3.52 5.17 6.77 7.96 8.21 8.22 8.24 2.56
10 EF 1.36 4.96 8.96 9.08 9.20 9.21 9.21 9.21 3.02
14 LF 1.12 2.24 6.24 6.28 6.29 6.31 6.31 6.31 0.98

DAI = Days after infestation.

Plant Age

Plant Stage


Plant Age Plant Damage Leaf Area/ leaf Leaf Fresh weight Leaf Dry weight Soluble Solids
(weeks) Stage Rating (cm2) (g) (g) (g)

Infested Control Infested Control Infested Control Infested Control Infested Control

5 V 3.59a 0.Ob 1.76b 5.20a 0.12b 0.16a 0.04b 0.05a 0.08b 0.1la ,
7 B 2.56a 0.Ob 2.12b 5.04a 0.08b 0.09a 0.02b 0.01a 0.07b 0.08a a
10 EF 3.02a O.Ob 5.22b 9.94a 0.18b 0.25a 0.03a 0.02a 0.16b 0.23a
14 LF 098a 0.Ob 5.10b 7.23a 0.18b 0.20a 0.05a 0.04a 0.14a 0.16a 0

Means for each parameter within rows for each parameter, followed by the same letter are not significantly differently (t-test; P = 0.05)



Plant Plant
Age Stage

5 V
7 B
10 EF
14 LF

Means for each paral

Leaves Plant Height No. Buds No. Flowers No. Fruits Fruit weight/plant 3i
/plant (cms.) /plants /plant /plant (g)

Infested Control Infested Control Infested Control Infested Control Infested Control Infested Control

17.68b 29.62a 11.97b 20.63a 2.17b 6.79a 0.12b 0.80a 0.08b 1.88a 10.67b 270.51a
14.44b 19.33a 11.76b 16.11a 1.15b 5.31a 0.01b 0.50a 0.02b 0.37a 3.90b 50.78a -
18.03b 30.49a 14.50b 22.54a 2.58b 10.83a 0.09b 1.31a 0.34b 0.98a 16.87b 55.62a
37.52b 41.39a 30.82b 32.54a 4.41a 5.28a 0.83a 0.93a 2.48b 3.10a 333.17a 346.07a Q-

neter within rows followed by the same letter are not significantly different (t-test, P=0.05)


Florida Entomologist 81(4)

December, 1998

100 ...... A

A Plant Age

1iE il 10
I 14
40- i 1

5 100
*5 100 --
B I Plant Age
S80o (Weeks)

60i 7
I I 10
40J r i14
8 -

2 7 14 20 26 33 46 50
Days After Infestation

Fig. 3. Densities ofP. latus motile and prelarva stages (A) and eggs (B) in 5, 7, 10
and 14 week-old pepper plants recorded for 50 days after infestation.

Mean fruit weight on LF plants did not differ significantly between infested and con-
trol plants. However, mean fruit weights ofV, B, EF plants exposed to broad mite were
significantly lower than those of control plants. Fruit weight was consistently lower
from V, B and EF broad mite-stressed plants compared to uninfested plants at the
same growth stage. Lower fruit numbers were recorded from mite-stressed plants
compared to the untreated check.

Relationship between damage index and plant yield:

The damage caused by broad mites appears to be dependent on the stage of devel-
opment of the pepper plant. Plants infested when 14 weeks-old (LF) had significantly
more fruits than plants infested at an earlier plant stage. This experiment indicated
an intermediate (y = 2.83 0.45x; P= 0.46; P= 0.0001) relationship between damage
index (x) and the number of fruits per plant (y). However, the relationship between
fruit weight per plant (y) in grams and damage index (x) was less than intermediate
(y = 232. 50 37.23x; P= 0.38; P= 0.0001). Nevertheless, these relationships may be
used to predict yield loss for P. latus infested pepper plants. For example, using the in-
tercepts of the above equations, and the damage index is 0, the yield of undamaged
plants would be 2.83 fruits or 232.50 grams per plant. However, if the damage index
(x) is 5, the yield will be reduced by 80%.
High levels of stress induced by P. latus feeding resulted in reductions in quantity
and quality of fruit, reduction in vegetative growth and flower development responds
to some anatomical, physiological or biochemical differences between vegetative and
reproductive stage plants. This reductions were due to chronic feeding on plants with
younger leaf tissue, which appear to be more susceptible than plants with greater

Coss-Romero & Pefa: Broad Mite and Injury Levels in Pepper 525

numbers of mature leaves. This effect has been shown to vary with the phenological
development of hedera (Nemestothy et al. 1982). Plants with younger hirsute leaves
suffered the strongest damage compared to older plants with leaves with less hairs
and where cell differentiation has already occurred. These results are in agreement
with the reports of Smith (1935) who stated that the broad mite cannot survive long
on the tough, mature leaves of most plants.
Regardless of the causative factors, our results help to explain why outbreak popu-
lations of P. latus are observed only in vegetative and early reproductive stages of the
crop. The response of P. latus to pepper phenology appears to be an important compo-
nent of the broad mites pest potential in the pepper ecosystem. In pepper, flowering
and fruit formation induce a significant increase in the growth rate of P. latus popula-
tions. This rapid increase in mite density, together with the production of new lateral
growth, may stimulate mite movement onto new lateral leaves. However, when the
leaves are mature (LF), the plants seem to be unable to support broad mite popula-
tions. Thus, mites invading plants younger than 14 weeks encounter a potential host
suitable for colonization and favorable rapid growth. The sequence of motile mites ob-
served every 8 days, explains why plants at these early ages have the potential for in-
ducing damaging mite outbreaks. This potential is often realized under the
exacerbating effects of hot humid weather and certain pesticide programs. The knowl-
edge that the potential of damage arises from mite responses to the phenological stage
of the crop can enhance the efficiency and value of broad mite monitoring programs
and control strategies by focusing attention on the critical periods prior to flowering
and fruiting in pepper. However, yield responses to broad mite damage under field con-
ditions may differ from those observed under controlled conditions in the greenhouse.


We thank A. Zaragoza and R. Duncan for technical help and D. Schuster,
G. Nuessly and W. Klassen for helpful review of an earlier draft of this manuscript.
This research was partially supported by the Organization of American States
through the award of a scholarship to the senior author. This is Florida Agricultural
Experiment Station Journal Series No. HOM-03505


en evidence de degats causes par Polyphagotarsonemus latus (Banks) sur pa-
payer a l'ile de la Reunion. Fruits. 36: 9-24.
BASSETT. P. 1981. Observations on broad mite (Polyphagotarsonemus latus) (Acarina:
Tarsonemidae) attacking cucumber. Proc. 1981 Br Crop Prot Conf. Pests and
Diseases 1: 99-103.
BEATTIE, G. A. C., AND GELLATLEY, J. G. 1983. Mite pests of citrus. Agfacts H2, AE 3,
Dept. Agriculture, New South Wales.
CROSS, J. V., AND P. BASSETT. 1982. Damage to tomato and aubergine by broad mite,
Polyphagotarsonemus latus (Banks). Plant Pathology. 31: 391-393.
DHOORIA, M. S., AND O. S. BINDRA. 1977. Polyphagotarsonemus latus (Banks), a mite
pest of chili and potato in Punjab. Acarology Newsletter. 4: 7-9.
HOOPER, G. H. S. 1957. The potato broad mite. Queensland Agr. J. 83: 56-58.
JEPSSON, L. R., H. H. KEIFFER, AND E. W. BAKER. 1975. Mites injurious to economic
plants Univ. California Press, Los Angeles.
JONES, F. P., AND R. D. BROWN. 1983. Reproductive responses of the broad mite,
Polyphagotarsonemus latus (Acari: Tarsonemidae), to constant temperature
humidity regimes. Ann. Entomol. Soc. Am. 76: 466-469.

526 Florida Entomologist 81(4) December, 1998

LAFFI, F. 1982. Occurrence of Polyphagotarsonemus latus (Banks) on capsicum sed-
bedsin North Italy. Informatore Fitopatologico. 32: 55-57.
Lo, P. K., AND S. R. CHAO. 1972. Tea broad mite (Hemitarsonemus latus (Banks))(Ac-
arina:Tarsonemidae) injure the string bean plants in greenhouses. J. Taiwan
Agric. Research 21: 59-61.
NEMESTHOTY, K., E. VOLCSANSKY, AND N. SIMON. 1982. Influence of damage of the
mites Tarsonemus pallidus and Polyphagotarsonemus latus Banks (Acari: Tar-
sonemidae) on the morphological properties of fashedera and hedera leaves.
Novenyvedelem 10: 437-442
OCHOA, R., H. AGUILAR, AND C. VARGAS. 1994. Phytophagous mites of Central Amer-
ica: An illustrated guide. CATIE, Turrialba 234 p.
PENA, J. E. 1990. Relationships of broad mite (Acari: Tarsonemidae) density to lime
damage. J. Econ. Entomol. 83: 2008-2015.
PENA, J. E., AND R. C. BULLOCK. 1994. Effects of broad mite, Polyphagotersonemus la-
tus feeding on vegetative plant growth. Florida Entomol. 77: 180-184.
Schoonhoven, A. V., J. Piedrahita, R. Valderrama, and G. Galvez. 1978. Biologia, dano
y control del acaro tropical Polyphagotarsonemus latus (Banks) (Acarina: Tar-
sonemidae) en frijol. Turrialba. 28: 77-80.
SAS INSTITUTE. 1987. SAS user's guide: statistics. SAS Institute, Cary NC.
SMITH, F. F. 1935. Control experiments on certain Tarsonemus mites on ornamentals.
J. Econ. Entomol. 28: 91-98.


Florida Entomologist 81(4)

December, 1998



Center for Medical, Agricultural and Veterinary Entomology, U. S. Department of
Agriculture, Agricultural Research Service, P.O. Box 14565,
Gainesville, FL 32604, USA

2Department of Entomology and Nematology, University of Florida,
Gainesville, FL 32611- 0620.


Two species of parasitoids, Cotesia plutellae (Hymenoptera: Braconidae) and Dia-
degma insulare (Hymenoptera: Ichneumonidae), of diamondback moth, Plutella xy-
lostella, (Lepidoptera: Plutellidae), were colonized in cages in cabbage fields west of
Bunnell, Florida, from November 1996 to February 1997. Two kinds of cages were
used: large-screened cages and screened laundry hampers. Both parasitoids attacked
their host during the winter, completed development within the host, and increased
in numbers within field cages. Parasitism of diamondback moth larvae by C. plutellae
was 36-42% in laundry hampers, and 35-65% in large screened cages. The sex ratio of

Hu et al: Parasitoid Field Production

emerging C. plutellae was 1:1-1.2 (Y:6) in laundry hampers and 1:0.8-1.3 in large
screened cages. Parasitism of diamondback moth larvae by D. insulare was 55-90%,
parasitoid adults emerged from 89% of the cocoons, and the sex ratio was 1:1.4-2.1 (9
:6) in large screened cages. The results showed that it is possible to rear these para-
sitoids in field nursery cages to provide parasitoid sources for release to control dia-
mondback moth in cabbage in Florida.

Key Words: Plutella xylostella, Cotesia plutellae, Diadegma insulare, biological con-
trol, parasitism, cabbage


Dos species de parasitoides, Cotesia plutellae (Hymenoptera: Braconidae) y Dia-
degma insulare (Hymenoptera: Ichneumonidae), de la palomilla dorso de diamante,
Plutella xylostella (Lepidoptera: Plutellidae), fueron colonizadas dentro de jaulas en
campos de col al oeste de Bunnell, Florida, de noviembre de 1996 a febrero de 1997.
Dos tipos de jaulas fueron usadas: jaulas grandes con tela de malla y canastas para
la ropa con tela de malla. Los dos parasitoides atacaron a sus hospederos durante el
invierno, completaron su desarrollo dentro de sus hospederos, y aumentaron en sus
numerous dentro de las jaulas en el campo. El nivel de parasitismo de larvas de la pa-
lomilla por C. plutellae fue de 36-42% en las canastas de ropa y de 35-65% en lasjaulas
grandes. El coeficiente sexual de los C. plutellae que emergieron fue de 1:1-1.2 (Y:6 )
en los canastos de ropa y de 1:0.8-1.3 en las jaulas grandes. El nivel de parasitismo de
larvas de la palomilla por D. insulare fue de 55-90%, el 89% de los parasitoides adults
emergieron de los capullos, y el coeficiente sexual fue de 1:1.4-2.1 (Y:6) en las jaulas
grandes con malla. Los resultados demostraron que es possible criar estos parasitoides
dentro de jaulas el en campo para proveer a los parasitoides con recursos para su li-
beraci6n para controlar la palomilla dorso de diamante en la col en Florida.

The diamondback moth, Plutella xylostella (L.), is the most destructive pest of cab-
bage and other crucifers throughout the world. The annual cost for control of this pest
is estimated to be U.S. $1 billion (Talekar & Shelton 1993). In Florida, the annual av-
erage production of cabbage is 4,555 hectares with an average total value of $34.4 mil-
lion, and diamondback moth is one of the major pests of this crop (Leibee 1996). The
diamondback moth typically has been controlled using pesticides (Shelton et al.
1993). To prevent damage to cabbage by diamondback moth, Florida growers typi-
cally have relied on one or two applications of insecticide per week; however, this has
led to problems from insecticide resistance (Leibee 1996).
Integrated pest management (IPM) programs provide the most viable alternative
to reliance on pesticides. An IPM approach to control lepidopterous pests in cabbage
using multiple tactics is described by Biever et al. (1994). In Florida, an IPM program
has been under trial for control of diamondback moth in cabbage fields. This pilot pro-
gram contains a combination of strategies such as biological control (releases of par-
asitoids; Mitchell et al. 1997a, 1998), trap crops (Mitchell et al. 1997b), pheromone for
disrupting mating (McLaughlin et al. 1994, Mitchell et al. 1997c) and Bt pesticides.
Of the parasitoids attacking diamondback moth, Cotesia plutellae Kurdjumov and
Diadegma insulare (Cresson) show the most promise (Ooi & Lim 1989, Ooi 1990,
Tabashnik et al. 1990). Diadegma insulare is the most important parasitoid of dia-

Florida Entomologist 81(4)

December, 1998

mondback moth in North America (Latheef & Irwin 1983, Pimentel 1961, Oatman &
Platner 1969, Harcourt 1960, 1963, 1986, Losata & Kok 1986, Horn 1987), and has re-
sulted in greater than 90% parasitism in untreated fields (Muckenfuss et al. 1990). It
is the most abundant parasitoid of diamondback moth in Florida (Mitchell et al. 1997b).
The female wasps primarily attack 2nd and 3'd instars of diamondback moth (Hu, et. al.:
unpublished data). After completing larval development, this parasitoid makes its co-
coon within the host cocoon. In cabbage fields in northeast Florida, however, Diadegma
insulare populations typically do not increase until late March (Hu et al. 1997).
An imported diamondback moth parasitoid, C. plutellae, has been released in cab-
bage in Florida. This parasitoid primarily attacks early instars of diamondback moth
(Hu et al.: unpublished data). After completing development, the parasitoid larva mi-
grates outside the host larva to form its own cocoon. The results from field releases
have shown that C. plutellae competes with D. insulare for increasing parasitism fol-
lowing inundative releases. Unfortunately, C. plutellae has not yet been found to es-
tablish in cabbage production areas in Florida (Mitchell et al. 1997a). Moreover,
purchasing this parasitoid for release is costly (Mitchell et al. 1998). The objective of
this study was to determine the feasibility of colonizing these two parasitoids in field
cages to provide local sources for releases in cabbage for control of diamondback moth.


Experimental Location

This study was conducted in an area of commercial cabbage production, west of
Bunnell, Flagler County, Florida. Approximately 800 hectares of cabbage grow in the
winter-spring and fall-winter crop in this area. The fall-winter crop lasts from October
to February and the winter-spring season lasts from January to April. Temperature
and humidity were recorded in that area while the study was carried out (November
1996 February 1997). Average daily high temperature was 24C, ranging from 9.4 to
32.2C. Average daily low temperature was 9.6C, ranging from -4.4 to 21.1C. Rela-
tive humidity was 22-100%, with an average of 61.8%.


Two types of cages were used in this study: screened Sterilite" laundry hampers
(Sterilite Corporation, Townsend, MA) and large screened cages. Laundry hampers
(Fig. 1, left) were trapezoidal and 61 cm high: the bottom was 38-cm long x 27-cm wide
and the top (opening) was 46.4-cm long x 33.7-cm wide. Air vents on the sides were
covered using a fine Saran" screen of eight meshes per cm. The large screened cages
(Fig. 1, right) were constructed of a wire frame covered with two layers of Saran"
screens: the outer layer was fine with 16 meshes per cm and the inner layer was
coarse with six meshes per cm. The cage was semicircular in section, 2.5-m long, 2.2-
m wide (bottom), and 1.0-m high. Edges of the screens were buried into soil to prevent
the insects within the cage from escaping and the insects outside from invading the
cage. A fire ant bait, Amdro" (American Cyanamid, Wayne, NJ), was applied onto the
ground inside and around the cages once a week to help control Solenopsis invicta Bu-
ren (Hymenoptera: Formicidae), which attacked immature diamondback moth and its
parasitoids in the cages.

Parasitoid Source

Cotesia plutellae used in this study were purchased from Biofac, Inc., Mathis,
Texas. Cocoons established on paper towels (about 1,000 each) were placed in plastic

Hu et al: Parasitoid Field Production

Fig. 1. Screened laundry hampers (left) used for rearing Cotesia plutellae and large
screened cages (right) used for rearing C. plutellae and Diadegma insulare in cabbage

bags, wrapped with old newsprint, inserted into styrofoam containers, and shipped to
Gainesville, Florida. Diadegma insulare (collected from Bunnell, Florida, May 1996)
were reared on diamondback moth larvae reared on wheat germ-based artificial diet
(Shelton et al. 1991). Both species of parasitoids were fed honey and water, and main-
tained under a 12:12 h L:D cycle at 25C and 50% RH.

Parasitoid Rearing in Large-Screened Cages

To initiate rearing, 10 to 20 individually potted collard plants, infested with 100-
200 diamondback moth larvae (mixed instars) each, were introduced into each of the
two cages. A cumulative total of 75 pairs of D. insulare was introduced into one cage
from 25 November to 10 December, 1996. On 28 November, 1996, 200 pairs of C. plu-
tellae were introduced into the other large cage. No more parasitoids were added to
the cages.
Follow-up visits to the cages were made once a week. Each visit included watering
the collard plants, removing dead plants and adding new infested plants when
needed. Every 2-3 weeks, samples of diamondback moth larvae and parasitoid co-
coons were collected from the plants in the cages and brought to our Gainesville lab-
oratory, where the larvae were dissected to determine parasitism. Parasitoid cocoons
were allowed to emerge as adults to obtain sex ratio data.

Parasitoid Rearing in Laundry-Hampers

Because C. plutellae is an exotic species, no data were available on the possibility
for survival in our experimental location. To start tests, two fully-grown potted collard
plants were infested with mixed instars of 100-200 diamondback moth larvae, and
were covered by an upside down screened laundry hamper. Next, 50 pairs of 3-d old
C. plutellae were introduced into the hamper. Four pieces of metal wire were used to

Florida Entomologist 81(4)

December, 1998

anchor each corner of the cage edges to the soil. This test was replicated three times
and each replicate included three cages. After the third wk, the collard plants along
with the laundry hampers were brought to the Gainesville lab. The diamondback
moth larvae collected from the plants were dissected for parasitism, and the cocoons
collected from the plants and the laundry hampers were reared to adults to obtain sex
ratio data.


Cotesia plutellae

A total of 225 diamondback moth larvae (mixed instars) collected from the laundry
hampers was dissected to determine parasitism by C. plutellae. Parasitism caused by
this parasitoid was 36-42% from November 1996 to February 1997. A total of 310 co-
coons of C. plutellae was collected, from which 256 adults emerged. Emergence suc-
cess ranged from 80.8 to 85.3%, with an average of 82.8% per trial. Sex ratios of the
emerging adults were 1: 1.0-1.2 (Y:6), with an average of 1: 1.13 (Table 1).
Eighty diamondback moth larvae collected over the four sampling dates from the
large screened cage also were dissected for eggs or larvae of C. plutellae (Table 1). Par-
asitism caused by this parasitoid ranged from 35 to 65%, with an average of 55%. Sex
ratios of emerging adults were 1: 0.8-1.3, with an average of 1: 1.1 (Y:6). Adults C. plu-
tellae were observed to fly around within the large screened cage during the entire
rearing period, but remained on plants when ambient temperature dropped below 5C.

Diadegma insulare

In the large screened cage, 75 diamondback moth larvae collected from the four
sampling dates were dissected to determine parasitism by D. insulare (Table 2). Par-
asitism caused by D. insulare ranged from 55 to 90%, with an average of 75%. Sex ra-
tios of emerging adults were 1:1.4-2.1, with an average of 1:1.8 (9 :d), which is similar
to that reported to occur in field populations (Idris & Grafius 1993, Mitchell et al.
1997b) and in our rearing facility. Adult D. insulare were observed to fly around
within the large screened cage. They were seen to hover a few cm to 20 cm above col-
lard plants or weeds. When ambient temperature dropped below 5C, however, they
stayed on the plants.
The results showed that C. plutellae and D. insulare attacked diamondback moth
larvae, completed their development within the host larvae, and reproduced continu-
ously within the nursery cages during the winter months in east-central Florida. An
estimate of three generations completed in cages. At the end of the rearing period
(late-February of 1997), the numbers ofD. insulare and C. plutellae increased greatly
but we did not attempt to quantify populations of the caged parasitoids.
The unparasitized larvae of diamondback moth used for the rearing developed into
cocoons, and adults emerged from the cocoons continuously throughout the rearing
period within the cages where C. plutellae and D. insulare were maintained. The adult
parasitoids were observed to stay on collard plants or in weeds, occasionally flying be-
tween plants. However, even though diamondback moth adults were continuously
present, larvae had to be introduced continuously into the cages with collard plants
because larval numbers were very low during the winter and the plant quality de-
creased over time.
Following lifting the fine (outer layer) screen from the large cage, C. plutellae col-
onized in the large-screened cage were observed to fly out through meshes of the


Diamondback Moth Parasitoid Emergence
Dates Released

(Pairs per
Larvae Checked Parasitism Total No. Cocoons Sex Ratio
Start End Cage) per Cage (%) Collected % Emergence (Y:6)

07/11/96 29/11/96 50 25 41.3 + 5.0 130 80.8 1:1.2
25/11/96 31/12/96 50 25 42.0 + 4.9 85 82.4 1:1.2
21/01/97 12/02/97 50 25 36.7 + 6.7 95 85.3 1:1.0

Florida Entomologist 81(4)

December, 1998


No. Larvae Adults Sex Ratio
Dates Dissected % Parasitism Emerging (: 6)

D. insulare
Jan.7 20 55 39 1:1.8
Jan. 19 20 90 48 1:1.8
Feb. 4 15 80 34 1:2.1
Feb. 16 20 75 22 1:1.4
Total 75 75 143 1:1.8

C. plutellae
Jan. 16 20 35 52 1:1.3
Jan. 28 20 65 34 1:0.8
Feb. 14 20 55 27 1:1.1
Feb. 26 20 65 36 1:1.3
Total 80L 55 149 1:1.1

coarse (inner layer) screen and spread into the cabbage in nearby fields. The adults of
diamondback moth, however, could not escape through the screen due to their larger
size, eliminating the spread of this pest from the nursery cages.
Unfortunately, Diadegma insulare could not be released in this way because its
size is nearly the size of diamondback moth adults and they could not migrate
through the coarse layer screen. Therefore, cocoons of this parasitoid were collected
from the cage and placed into adjacent cabbage fields. Unfortunately, released para-
sitoids could not be evaluated for establishment due to heavy chemical pesticide ap-
plications by growers in the fields near the rearing cages.
Diamondback moth larvae sterilized by gamma radiation are just as suitable hosts
for the parasitoid (C. plutellae) as are unsterilized larvae (Okine et al. 1998) and may
be used in the future as the host for rearing D. insulare in field cages. The adults of
D. insulare and those of sterile diamondback moth can then be released simulta-
neously from the cages into the fields by lifting the cover screens without infesting the
field with diamondback moth.
Augmentation of parasitoids to increase their effectiveness involves their direct
manipulation, either by mass production and periodic colonization, or by some type of
planned genetic improvement, or by employing chemical cues that affect their behav-
ior (Debach & Rosen 1991). Our results showed that D. insulare and C. plutellae were
successfully colonized in field cages. Both parasitoids survived, completed their devel-
opment within the host, and increased in numbers. Therefore, colonizing these two
species of parasitoids in field cages may provide a good source of large numbers of C.
plutellae and D. insulare for control of diamondback moth in cabbage. The parasitoids
were ready for release into the cabbage fields either as adults or cocoons. This proce-
dure can be easily adopted by growers of commercial cabbage. Moreover, the rearing
facility was not costly and required minimal labor, both important considerations for
implementation of biological control.

Hu et al: Parasitoid Field Production


We thank J. Coleman, W. Copeland, R. Furlong, K. Gallagher, J. Gillett, and
J. Leach (USDA-ARS, Gainesville, FL) for technical assistance; and T. Turner and
B. Hawkins (Flagler Co., FL) for use of their land and crops for this research.


1. This article reports the results of research only. Mention of a proprietary prod-
uct does not constitute an endorsement or the recommendation for its use by USDA.


BIEVER, K. D., D. L. HOSTETTER, AND J. R. KERN. 1994. Evolution and implementation
of a biological control-IPM system for crucifers: a 24-year case study. American
Entomol. 40: 103-108.
DEBACH, P., AND D. ROSEN. 1991. Biological control by natural enemies. 2nd ed. Cam-
bridge University, New York.
HARCOURT, D. G. 1960. Biology of the diamondback moth, Plutella maculipennis
(Curt) (Lepdoptera: Plutellidae) in eastern Ontario III. Natural enemies. Ca-
nadian Entomol. 92: 419-428.
HARCOURT, D. G. 1963. Major mortality factors in population dynamics of the dia-
mondback moth, Plutella maculipennis (Cur.) (Lepidoptera: Plutellidae). Mem.
Canadian Entomol. Soc. 32: 55-56.
HARCOURT, D. G. 1986. Population dynamics of the diamondback moth in southern
Ontario, pp. 1-15 in N. S. Talekar and T. D. Griggs [eds.], Diamondback moth
management. Proc. 1" Inter. Workshop at Asian Veg. Res. and Dev. Center, 11-
15 March 1985, Shanhua, Taiwan.
HORN, D. J. 1987. Vegetational background and parasitism of larval diamondback
moth on collards. Entomol. Exp. Appl. 43: 300-303.
Hu, G. Y., E. R. MITCHELL, AND J. S. OKINE. 1997. Effect of habitat types on popula-
tion densities and parasitism of the diamondback moth (Lepidoptera: Plutel-
lidae) larvae in cabbage fields. J. Entomol. Sci. 32:56-71.
IDRIS, A. B., AND E. GRAFIUS. 1993. Field studies on the impact of pesticides on the di-
amondback moth, Plutella xylostella (L.) (Lepidoptera: Plutellidae) and para-
sitism by Diadegma insulare (Cresson) (Hymenoptera: Ichneumonidae). J.
Econ. Entomol. 86: 1196-1202.
LASOTA, J. A., AND L. T. KOK. 1986. Diadegma insulare (Hymenoptera: Ichneu-
monidae) parasitism of the diamondback moth (Lepidoptera: Plutellidae) in
southwest Virginia. J. Entomol. Sci. 21: 237-242.
LATHEEF, M. A., AND R. D. IRWIN. 1983. Seasonal abundance and parasitism of lepi-
dopterous larvae on Brassica greens in Virginia. J. Georgia Entomol. Soc. 18:
LEIBEE, G. L. 1996. Insect pest management in crucifers in Florida, pp. 413-422 in
Rosen D., F.D. Bennett and J. L. Capinera [eds.], Pest Management in the Sub-
tropics: Integrated Pest Management a Florida Perspective. Intercept Lim-
ited, Andover, United Kingdom.
MCLAUGHLIN, J. R., E. R. MITCHELL, AND P. KIRSCH. 1994. Mating disruption of dia-
mondback moth (Lepidoptera: Plutellidae) in cabbage: reduction of mating and
suppression of larval populations. J. Econ. Entomol. 87: 1198-1204.
mondback moth (Lepidoptera: Plutellidae) parasitism by Cotesia plutellae (Hy-
menoptera: Braconidae) in cabbage. Florida Entomol. 80: 477-484.

Florida Entomologist 81(4)

December, 1998

MITCHELL, E. R., G. Y. HU, AND J. S. OKINE. 1997b. Diamondback moth (Lepidoptera:
Plutellidae) infestation and parasitism by Diadegma insulare (Hymenoptera:
Ichneumonidae) in collards and adjacent fields. Florida Entomol. 80: 54-62.
MITCHELL, E. R., G. Y. HU, J. S. OKINE, AND J. MCLAUGHLIN. 1997c. Mating disrup-
tion of diamondback moth (Lepidoptera: Plutellidae) and cabbage looper (Lep-
idoptera: Noctuidae) in cabbage using a blend of pheromones emitted from the
same dispenser. J. Entomol. Sci. 32: 120-137.
MITCHELL, E. R., G. Y. HU, J. S. OKINE, AND J. E. CARPENTER. 1998. Parasitism of di-
amondback moth (Lepidoptera: Plutellidae) larvae by Cotesia plutellae (Hy-
menoptera: Braconidae) and Diadegma insulare (Hymenoptera:
Ichneumonidae) in cabbage fields after inundative releases of C. plutellae. J.
Entomol. Sci. 4: (December).
MUCKENFUSS, A. E., B. M. SHEPPARD, AND E. R. FERRER. 1990. Natural mortality of
diamondback moth in coastal South Carolina, pp. 27-36 in Talekar, N. S. [Ed.],
Diamondback Moth and Other Crucifer Pests: Proceedings of the Second Inter-
national Workshop, Tainan, Taiwan, 10-14 December 1990. Asian Vegetable
Research and Development Center, AVRDC Publication No. 92-368, 603 pp.
OATMAN, E. R., AND G. R. PLATNER. 1969. An ecological study of insect populations on
cabbage in southern California. Hilgardia 40: 1-40.
OKINE, J. S., E. R. MITCHELL, J. CARPENTER, AND G. Y. HU. 1998. Oviposition re-
sponse of Cotesia plutellae (Hymenoptera: Braconidae) to sterile and untreated
diamondback moth (Lepidoptera: Plutellidae) larvae and spatial dispersion of
larvae on collard plants. Environ. Entomol. 27(6): December.
Ooi, P. A. C. 1990. Management of diamondback moth in Cameron Highlands. Newlt.
Malaysia Plant Prot. Soc. 13, 55-56.
OoI, P. A. C., AND G. S. LIM. 1989. Introduction of exotic parasitoids to control the di-
amondback moth in Malaysia. J. Plant Prot. Tropics 6: 103-111.
PIMENTEL, D. 1961. Natural control of caterpillar populations in cole crops. J. Econ.
Entomol. 54: 889-892.
BRODE. 1991. Comparative analysis of two rearing procedures for diamondback
moth (Lepidoptera: Plutellidae). J. Entomol. Sci. 26: 17-26.
MAHR, AND S. D. EIGENBRODE. 1993. Insecticide resistance of diamondback
moth (Lepidoptera: Plutellidae) in North America. J. Econ. Entomol. 86: 11-19.
TABASHNIK, B. E., N. L. CUSHING, N. FINSON, AND M. W. JOHNSON. 1990. Field devel-
opment of resistance to Bacillus thuringensis in diamondback moth (Lepi-
doptera: Plutellidae). J. Econ. Entomol. 83: 1671-1676.
TALEKAR, N. S., AND A. M. SHELTON. 1993. Biology, ecology, and management of the
diamondback moth (Lepidoptera: Plutellidae). Annu. Rev. Entomol. 38: 275-

Hatle & Spring:AKHs in juvenile Romalea guttata


University of Southwestern Louisiana, Department of Biology, PO Box 42451
Lafayette, LA 70504


Romalea guttata Houttuyn (= R. microptera Beavois) is flightless, lethargic, apose-
matic, and chemically defended. R. guttata stores large quantities of two adipokinetic
hormone (AKH) family peptides in its corpora cardiac. In adults, these peptides
(Rom-CC-I and Grb-AKH) activate fat body glycogen phosphorylase but are not hy-
pertrehalosemic. Because juvenile R. guttata contain sufficient peptide to be bioac-
tive, we sought to determine whether these peptides are hypertrehalosemic,
phosphorylase activating, or hyperlipemic in juveniles. Late fifth (= last) instar and
adult R. guttata activated phosphorylase in response to Rom-CC-I injections. These
same individuals showed no hypertrehalosemia in response to Rom-CC-I. We hypoth-
esize that the glycogenolysis pathway is not started by activation of glycogen phos-
phorylase in response to Rom-CC-I. From fourth instar through third week adult, R.
guttata showed a slight, statistically insignificant hypolipemia, but clearly no hyper-
lipemia. R. guttata differs from Locusta migratoria in that it appears to show neither
hypertrehalosemia nor hyperlipemia at any developmental stage.

Key Words: lubber grasshopper; chemical defense; glycogenolysis; adipokinetic hormone


Romalea guttata Houttuyn (= R. microptera Beavois) es una especie que no vuela,
es letargica, es aposematica, y se defiende por medios quimicos. R. guttata guard
grandes cantidades de dos p6ptidos de la familiar de las hormones adipokineticas
(AKH) en la glandula corpora cardiac. En los adults, estos p6ptidos (Rom-CC- I and
Grb-AKH) activan fosforilasas de glic6geno del cuerpo graso pero no son hipertreha-
los6micos. Como los R. guttata juveniles contienen suficientes p6ptidos para que sean
bioactivos, hemos tratado de determinar si estos p6ptidos son hipertrehalosemicos,
activadores de fosforilasas, o hiperlipemicos en los juveniles. El quinto (iltimo) instar
y el adulto de R. guttata activaron la fosforilasa como respuesta a injecciones de Rom-
CC-I. Estos mismos individuos no mostraron ninguna hipertrehalosemia como res-
puesta a Rom-CC-I. Nosotros suponemos que el process de glicogenolisis no es ini-
ciado con la activaci6n de la fosforilasa del glic6geno como respuesta a Rom-CC-I. Del
cuarto estadio al adulto de tres semanas, los R. guttata mostraron una hipolipemia es-
tadisticamente insignificant, pero claramente ninguna hyperlipemia. R. guttata di-
fiere de Locusta migratoria en que parece no mostrar ni hipertrehalosemia ni
hyperlipemia en ninguna fase de su desarrollo.

The Eastern Lubber Grasshopper, Romalea guttata Houttuyn (= R. microptera
Beavois), is seasonally common in the Southeastern US, flightless, lethargic, apose-
matic, and chemically defended (Whitman et al. 1990). Adult R. guttata store large

Florida Entomologist 81(4)

December, 1998

quantities of two small peptides in their corpus cardiacum (CC; Gade and Spring
1986). These peptides (Rom-CC-I and Grb-AKH) are members of the adipokinetic /
red-pigment concentrating hormone (AKH/RPCH) family by both bioactivity in ap-
propriate test species (Gade and Spring 1986) and their primary structures (Gade et
al. 1988). Further, the peptides are released under in vitro physiological conditions
(Spring and Gade 1991). In adults, the endogenous functions of both Rom-CC-I and
Grb-AKH appear to be control of fat body glycogen phosphorylase (GP), converting in-
active GP to the active form. In adult R. guttata, separate injections of Rom-CC-I and
Grb-AKH activated GP in a dose dependent manner (Gade and Spring 1989). As is
true with the well-studied migratory locust (Locusta migratoria L.), the activation of
GP was not concurrent with hypertrehalosemia (Goldsworthy 1994). Also, the pep-
tides elicited no hyperlipemic response in adult R. guttata (Spring and Gade 1987). In
sum, injections of endogenous AKHs into adult R. guttata do not appear to affect
hemolymph metabolite levels.
Juvenile R. guttata contain sufficient Rom-CC-I and Grb-AKH to be biologically
active (Spring and Gade 1991). For example, early fourth instar R. guttata contain
about 230 pmol Rom-CC-I and about 25 pmol Grb-AKH in their CC (Spring and Gade
1991). These quantities are much greater than the minimum dosage (1 pmol) needed
for maximal activation of GP in adults. We hypothesized that because Rom-CC-I and
Grb-AKH are not known to play a biologically important role in adult R. guttata, they
may function to control hemolymph metabolites in juveniles. Alternatively, endoge-
nous AKHs in R. guttata may not serve any metabolite control functions that endog-
enous AKHs serve in other grasshoppers, such as L. migratoria.
We asked four primary questions. First, are juvenile R. guttata significantly hy-
pertrehalosemic in response to synthetic Rom-CC-I (sRom-CC-I)? Second, do these
same grasshoppers activate GP in response to sRom-CC-I? Third, are decapitated R.
guttata significantly hypertrehalosemic in response to sRom-CC-I? Fourth, are juve-
nile R. guttata hyperlipemic in response to injections of CC homogenates?


Experiment 1-Responses of Hemolymph Carbohydrates to sRom-CC-I Injections
through Development.

Experimental animals. In April, 1996, we collected first instar R. guttata near Ly-
dia, LA, USA. We brought the insects to the laboratory and raised them as described
by Whitman (1986). Briefly, we kept R. guttata at 30 + 2C on a 14L:10D photoperiod.
R. guttata were fed oatmeal and Purina Cricket Chowad libitum, Romaine lettuce
daily, and green beans, green onions, carrot tops, and apple occasionally.
Determination of hemolymph carbohydrates. We measured changes in hemolymph
carbohydrates in response to sRom-CC-I injections at six stages of development: in-
star 4 day 4 (= L4-d4); L5-d2; L5-d6; L5-d10; adult days 3 and 4 (= Ad-d3); and Ad-d9.
To determine the hemolymph carbohydrate concentrations, we collected 2 Pl
hemolymph samples and measured total carbohydrates as anthrone positive material
with trehalose standards (Spik and Montreuil 1964). We then injected each grasshop-
per with either 5 pl deionized HO2 or 20 pmol sRom-CC-I (Peninsula Laboratories
Inc.; San Carlos, CA) in 5 il deionized HO0 and measured hemolymph carbohydrates
again 90 min later.
Statistics. We tested the changes in carbohydrates for statistical differences be-
tween treatments with ANOVA and Tukey's post-tests.

Hatle & Spring:AKHs in juvenile Romalea guttata

Experiment 2-Responses ofActive GP to sRom-CC-I Injections through Development.

Experimental animals. Following the determination of carbohydrates, we offered
each R. guttata Romaine lettuce and kept them isolated at 25 + 2C overnight until
the GP experiments the next day. Half of the grasshoppers injected with deionized
H20 for carbohydrate determination were also injected with deionized HO0 for active
GP determination. The remainder of the grasshoppers injected with deionized HO for
carbohydrate determination were injected with sRom-CC-I for active GP determina-
tion. This treatment control influenced our data only once, for L5-d6 grasshoppers. At
this developmental stage, insects injected with sRom-CC-I the previous day did not
activate GP, but insects injected with deionized HO the previous day did activate GP.
Because this was the sole influence of our treatment control, we combined the data for
clarity of presentation.
Determination of active GP. We assayed active GP by following glycogen breakdown
according to the methods of Ziegler et al. (1979) as modified by Gade and Spring (1989).
Statistics. We tested the changes in percent active GP for statistical differences be-
tween treatments by ANOVA with Tukey's post-tests.

Experiment 3-Responses of Hemolymph Carbohydrates to sRom-CC-I Injections in
Decapitated Adults.

Experimental animals. In August, 1996 we collected adult R. guttata from the
same collection site used in Experiment 1. These R. guttata were kept as described
above for three to six days before experimentation.
Determination of hemolymph carbohydrates. The night before an experiment, we
decapitated the grasshoppers and sealed the exposed orifices with liquid wax. We
measured changes in hemolymph carbohydrates in response to either deionized HO
injection or sRom-CC-I injection identically to Experiment 1, except that we took
hemolymph samples from the coxal membrane.
Statistics. We compared the data using student's t-tests.

Experiment 4-Determination of Hemolymph Lipids in Response to CC Homogenate
Injections through Development.

Experimental animals. We collected R. guttata from April to August, 1994 from the
same collection site used in Experiment 1. We fed these grasshoppers lettuce daily
and Purina Cricket Chow" ad libitum. Grasshoppers were held in the laboratory at
least 48 h prior to use. In all other respects, we maintained these grasshoppers iden-
tically to the grasshoppers used in Experiment 1.
CC homogenates. We prepared CC homogenates by the method of Gade and Spring
(1989). Our CC homogenates contained both Rom-CC-I and Grb-AKH, the predomi-
nant R. guttata CC peptides (Spring and Gade, 1987).
Lipid assays. We measured changes in hemolymph lipids in response to CC prep-
aration injections at six stages of development: instar 3 (= L3); L4; L5; week 1 adults
(= Ad-wl); Ad-w2; Ad-w3. We used the method of Spring and Gade (1987) to determine
if R. guttata's competence to CC homogenates with respect to hyperlipemia changes
through development. We first collected a hemolymph sample from each grasshopper,
and then we injected 5 pl aliquots of test solution (= 0.1 CC-equivalents) intra-abdom-
inally. Second samples were taken 90 min post-injection. We measured total lipids as
vanillin-positive material using vegetable oil standards following the method of Z6ll-
ner and Kirsch (1962).
Statistics. We tested the changes in lipid concentrations for statistical differences
among developmental stages by ANOVA.

Florida Entomologist 81(4)


Experiment 1-Responses of Hemolymph Carbohydrates to sRom-CC-I Injections
through Development.

Changes in hemolymph carbohydrate concentrations after test solution injections
varied widely among the six developmental stages examined (Fig. 1). Except for L5-
d2 (P < 0.05; Tukey's test), injection of sRom-CC-I did not statistically change
hemolymph carbohydrate concentrations in comparison to water injection within any
developmental stage.

Experiment 2-Responses of Active GP to sRom-CC-I Injections through Develop-

Synthetic Rom-CC-I injection activates GP during the developmental period from
L4-d4 to Ad-d9; an ANOVA revealed a statistically significant effect of sRom-CC-I in-
jection (P = 0.009), but no significant effects for developmental stage (P = 0.651) or the
interaction of developmental stage and sRom-CC-I injection (0.417; Fig. 2). In gen-
eral, activation of GP was stronger in the older R. guttata, with competence to sRom-
CC-I appearing to develop by L5-d10.

Experiment 3-Responses of Hemolymph Carbohydrates to sRom-CC-I Injections in
Decapitated Adults

Changes in hemolymph carbohydrate concentrations in decapitated adults that
were injected with water (x = 0.0151 mg/ml; SE = 0.0054; n = 9) did not differ signif-





-20 L

L4-d4 L5-d2 L5-d6 L5-dl0 Ad-d3 Ad-d9

Fig. 1. Changes in hemolymph carbohydrate concentrations in response to injec-
tions at six developmental stages of R. guttata. Dots represent deionized water in-
jected groups (n = 13-20), and squares represent sRom-CC-I injected groups (n = 7-
10). For abbreviations, see text.


December, 1998

Hatle & Spring:AKHs in juvenile Romalea guttata


30 T


20 I

L4-d4 LS-d2 LS-d6 L5-d10 Ad-d3 Ad-d9
Fig. 2. Percent active fat body glycogen phosphorylase in response to injections in
six developmental stages of R. guttata. Dots represent deionized water injected
groups (n = 7-8), and squares represent sRom-CC-I injected groups (n = 13-17). For ab-
breviations, see text.

icantly from concentration changes in decapitated adults that were injected with
sRom-CC-I (x = 0.0091 mg/ml; SE = 0.0038; n = 12).

Experiment 4-Changes in Hemolymph Lipids in Response to CC Homogenate Injec-
tions through Development.
Injections of CC extracts produced a hypolipemic affect in all developmental
stages from L3 to Ad-w3 (Fig. 3). ANOVA revealed no significant differences in the re-
sponses among any of the developmental stages (P = 0.127). In general, adults showed
the strongest hypolipemic responses, and larvae showed the weakest hypolipemic re-


Activation of GP but Absence of Hypertrehalosemia
R. guttata significantly activate GP in response to sRom-CC-I injections (Fig. 2).
The interaction of developmental stage and sRom-CC-I injection was insignificant;
nonetheless, it appears from our data that the competence to AKHs in R. guttata de-
velops in the late fifth instar. Regardless of the developmental moment of the onset of
competence, it is clear that the older R. guttata in our study activated GP in response
to sRom-CC-I injections.
R. guttata has no competence to sRom-CC-I with respect to hypertrehalosemia at
any developmental stage from L4-d4 through Ad-d9 (Fig. 1). For the well-studied L.
migratoria, there have been at least two explanations postulated in the literature for
this lack of hypertrehalosemia concurrent with activation of GP: 1) not enough glyco-

Florida Entomologist 81(4)

December, 1998

developmental stage

L3 L4 L5 Ad-wl Ad-w2 Ad-w3


-0. 15

-0.20 -
M -0.25

Fig. 3. Changes in hemolymph lipid concentrations in response to injections of cor-
pus cardiac preparations at six developmental stages (n = 7-10) of R.guttata. For ab-
breviations, see text.

gen in the fat body (Goldsworthy 1994), and 2) inhibition of hyperglycemia by an un-
specified "head factor" (Loughton and Orchard 1981). Adult R. guttata, especially
those fed daily in the laboratory (as ours were), have sufficient glycogen in the fat
body to produce a hypertrehalosemic effect (Spring and Gade 1987). Second, the lack
of hypertrehalosemia in decapitated adults (see Experiment 3) suggests that, in R.
guttata, a head factor is not necessary for the prevention of hypertrehalosemia.
Alternatively, it may be that the additional quantity of GP activated in our exper-
iments (- 10%) was not sufficient to induce hypertrehalosemia. We do not believe this
to be true for two reasons. First, Gade and Spring (1989) showed stronger activations
of GP (- 30%) but still no hypertrehalosemia. Second, the catalytic nature of enzyme
function requires only a small change in enzyme activation to produce a large change
in metabolite concentrations.
Fifth instar L. migratoria are moderately hypertrehalosemic in response to the en-
dogenous Lom-AKH-I (Van Marrewijk et al. 1984), whereas fifth instar R. guttata are
clearly not hypertrehalosemic in response to sRom-CC-I. In fact, our data suggest
that L5-d2 R. guttata may be hypotrehalosemic in response to sRom-CC-I. We there-
fore hypothesize that the glycogenolysis pathway is not started by activation of GP in
response to sRom-CC-I in R. guttata. R. guttata may activate GP for some function
other than the mobilization of sugars, but this seems highly unlikely. Barring this ex-
planation, the competence to sRom-CC-I in R. guttata appears to be an evolutionary
remnant of the development of flight physiology in last instar Acrididae. Importantly,
our data suggest that the physiology of R. guttata, as well its behavior, is different
from other grasshoppers, and that this difference reflects its flightless, lethargic,
chemically defended life style.

Hatle & Spring:AKHs in juvenile Romalea guttata

Hypolipemic Response to CC Homogenates?

R. guttata are slightly hypolipemic in response to CC homogenate injections at de-
velopmental stages from L3 through Ad-w3. Rom-CC-I and Grb-AKH are the predom-
inant peptides in R. guttata CC homogenates. Hence, at the very least, R. guttata are
not hyperlipemic in response to injections of endogenous AKHs. In contrast, L. migra-
toria develop competence with respect to hyperlipemia to the synthetic endogenous
AKH as L5 (Van Marrewijk et al. 1984). The lack of hyperlipemia in L5 and adult R.
guttata is further evidence that the physiology of these grasshoppers is different from
the physiology of grasshoppers that fly.

Divergent Physiology?

We have shown three ways that R. guttata differs physiologically from L. migrato-
ria. All three of these differences make sense in light of R. guttata's flightless and le-
thargic behavior: 1) absence of hypertrehalosemia in late L5 grasshoppers in concert
with activation of GP; 2) absence of hypertrehalosemia in decapitated adults ofR. gut-
tata; 3) absence of hyperlipemia in L5 and adult R. guttata. Taken together, these data
suggest that R. guttata may have diverged physiologically from grasshoppers that can
fly, and that R. guttata's responses mirror its flightless, lethargic, chemically defended


GADE, G., AND J. H. SPRING. 1986. Presence and preliminary characterization of fac-
tors regulating carbohydrate and lipid metabolism isolated from the corpus
cardiacum of the Eastern lubber grasshopper. pp. 191-194 in Borkovec A. B.
and D. B. Gelman [eds]. Insect Neurochemistry and Neurophysiology. Humana
Press, Clifton, NJ.
GADE, G., AND J. H. SPRING. 1989. Activation of fat body glycogen phosphorylase in the
eastern lubber grasshopper (Romalea microptera) by the endogenous neu-
ropeptides Ro I and Ro II. J. Exp. Zool. 250: 140-149.
GADE, G., C. HILBICH, K. BEYREUTHER, AND K. L. RINEHART. 1988. Sequence analyses
of two neuropeptides of the AKH/RPCH-family from the lubber grasshopper,
Romalea microptera. Peptides 9: 681-688.
GOLDSWORTHY, G. 1994. Adipokinetic hormones of insects: are they the insect gluca-
gons? pp. 486-492 in Davey, K. G., R. E. Peter, and S. S. Tobe. Perspectives in
Endocrinology. National Research Council Of Canada, Ottawa.
LOUGHTON, B. G., AND ORCHARD, I. 1981. The nature of the hyperglycemic factor from
the glandular lobe of the corpus cardiacum of Locusta migratoria. J. Insect.
Physiol. 27: 383-385.
SPIK, G., AND J. MONTREUIL. 1964. Deux causes d'erreur dans les dosages colo-
rimetriques des oses neutres totaux. Bull. Soc. Chim. Biol. 46: 739-749.
SPRING, J. H., AND G. GADE. 1987. Factors regulating carbohydrate and lipid metab-
olism isolated from the corpus cardiacum of the Eastern lubber grasshopper,
Romalea microptera. J. Exp. Zool. 241: 41-50.
SPRING, J. H., AND G. GADE. 1991. Storage and release of neuropeptides from the cor-
pus cardiacum of the Eastern lubber grasshopper, Romalea microptera. J. Exp.
Zool. 258: 60-68.
WHITMAN, D. W. 1986. Laboratory biology of Taeniopoda eques (Orthoptera: Acrid-
idae). J. Entomol. Sci. 21: 87-93.
WHITMAN, D. W., M. S. BLUM, AND D. W. ALSOP. 1990. Allomones: chemicals for de-
fense. pp. 289-351 in Evans, D. L., and J. O. Schmidt [eds]. Insect Defenses.
State University of New York Press, Albany

542 Florida Entomologist 81(4) December, 1998

TH. BEENAKKERS. 1984. Hypertrehalosaemic and hyperlipaemic responses to
adipokinetic hormone in fifth larval instar locusts, Locusta migratoria. Insect.
Biochem. 14: 151-157.
1979. Regulation of glycogen phosphorylase in fat body of Ceropia silkmoth pu-
pae. J. Comp. Physiol. B131: 321- 332.
ZOLLNER, N., AND K. KIRSCH. 1962. Uber die quantitative Bestimmung von Lipoiden
(Mikromethode) mittels der vielen natiirlichen Lipoiden (allen bekannten Plas-
malipoiden) gemeinsamen Sulfophospho-vanillin Reaktion. Z. Ges. Exp. Med.
135: 545-561.

Scientific Notes


European Biological Control Laboratory, United States Department of Agriculture
Agricultural Research Service, Parc Scientifique Agropolis 11,34397 Montpellier
Cedex 5, FRANCE

A problem with coffee consumption is the possible presence of ochratoxin A, a po-
tent toxin produced by Aspergillus ochraceus Wilh. and Penicillium viridicatum
Westl. Several studies have demonstrated the presence of this toxin in green coffee
beans, roasted coffee, and coffee brews, including instant coffee (Levi et al. 1974,
Tsubochi et al., 1984, Micco et al. 1989, Studer-Rohr et al. 1994, Patel et al. 1997). In
the UK, out of 100 retail coffee samples tested for ochratoxin A, 81 tested positive (Pa-
tel et al. 1997). Coffee exported from Brazil to Greece and Lebanon must be tested for
ochratoxin A and levels must be below 20 mg/kg (Milanez et al. 1995). After collecting
coffee beans infected with the coffee berry borer (Hypothenemus hampei (Ferrari), Co-
leoptera: Scolytidae) in Uganda and Benin, as part of a project aimed at finding new
biological control agents against this insect pest, we isolatedA. ochraceus from adult
insects emerging from the beans. In Uganda, of 636 insects emerging from coffee
beans collected in 26 sites, 34 (5.3%) were infected withA. ochraceus. In Benin, out of
564 insects originating in one site, 98 (17.4%) were infected with A. ochraceus. H.
hampei females lay eggs inside the coffee bean where both larval development and
mating occur. If the mother is infected with A. ochraceus while entering the bean, it
is likely that the adult progeny leaving the bean will also be infected, thereby dissem-
inating the fungus. This insect, endemic to Africa, has now spread to most coffee grow-
ing regions in the world; therefore, its potential to serve as a vector for this
cosmopolitan fungus is high. Other insects are known to serve as vectors for toxico-
genic fungi, includingAspergillus flavus Link: Fr. to corn (Dowd 1998). Our finding in-
dicates that H. hampei, in addition to being a direct pest of coffee, could also serve as
a vector forA. ochraceus. Plans aimed at reducing ochratoxin contamination in coffee
beans should take into consideration the presence of the insect in the field.
We are grateful to E. Rosenquist (Office of International Research Programs, USDA)
for sponsoring this work, to C. Lomer (IITA.Benin) for sending coffee beans and P. F.
Dowd (USDA) for comments on the manuscript.


A search for natural enemies of the coffee berry borer Hypothenemus hampei (Fer-
rari) (Coleoptera: Scolytidae) in Uganda and Benin revealed that the insect serves as
a carrier for Aspergillus ochraceus K. Wilh., a fungus known to produce ochratoxin A.
Contamination with this toxin is a serious problem for the coffee industry. Plans
aimed at reducing this problem should take into consideration the presence of this in-
sect in the field.


DOWD, P. F. 1998. The involvement of arthropods in the establishment of mycotoxi-
genic fungi under field conditions. in: K. K. Sinha and D. Bhatnagar [eds.] My-
cotoxins in Agriculture and Food Safety. Marcel Dekker, New York, pp. 307-350.
LEVI, C. P., H. L. TRENK, AND H. K. MOHR. 1974. Study of the occurrence of ochratoxin
A in green coffee beans. J. Assoc Off. Anal. Chem. 57: 866-870.

544 Florida Entomologist 81(4) December, 1998

MIcco, C., M. GROSSI, M. MIRAGLIA, AND C. BRERA. 1989. A study of the contamina-
tion of ochratoxin A of green and roasted coffee beans. Food Add. Contam. 6:
MILANEZ, T. V., M. SABINO, AND L. C. A. LAMARDO. 1995. Comparison of two methods
for the determination of ochratoxin A in green coffee beans. Rev. Microbiol. 26:
PATEL, S., C. M. HAZEL, A. G. M. WINTERTON, AND A. E. GLEADLE. 1997. Survey of och-
ratoxin A in UK retail coffees. Food Add. Contam. 3: 217-222.
rence of ochratoxin A in coffee. Food Chem. Toxicol. 33: 341-355.
TSUBOCHI, H., K. YAMAMOTO, K. HISADA, AND Y. SAKABE. 1984. A survey of occurrence
of mycotoxins and toxicogenic fungi in imported green coffee beans. Proc. Jpn.
Assoc. Mycotoxicol. 19: 14-21.


Florida Entomologist 81(4)

December, 1998


Krista E. M. Galley' and R. Wills Flowers2

1Longleaf Pine Restoration Project, The Nature Conservancy, P.O. Box 875, Niceville,
FL 32588

2Agricultural Research Programs, Florida A&M University, Tallahassee, FL 32307

Ecosystem management programs of the U.S. Forest Service (Hermann et al., in
press) and a cooperative agreement between The Nature Conservancy and the Depart-
ment of Defense have promoted study of the increasingly rare longleaf pine (Pinus
palustris) ecosystem (Biondo 1997) and the arthropods inhabiting it. In the Florida
Panhandle, two research projects on longleaf pine restoration ecology have led to the re-
discovery in Florida of the springtail Sminthurus floridanus MacGillivray (Collembola:
Sminthuridae) and the grasshopper Gymnoscirtetes morse Hebard (Orthoptera: Acri-
didae). Both species have been searched for in Florida in recent years without success.
S. floridanus was described from one specimen collected in "Florida" (MacGillivray
1893). This species was known from that single specimen until Snider (1982) redescribed
S. floridanus from several series collected at the Savannah River Plant, Aiken, South
Carolina. These series were swept from roadside grass beneath tall loblolly pines.
In 1995-1997, S. floridanus was collected on Eglin Air Force Base (EAFB) in north-
west Florida. All specimens were taken in an area subject to frequent fires, character-
ized by a nearly pure stand of longleaf pine, a sparse hardwood midstory, and a dense
groundcover of grasses and forbs, including bluestems (Andropogon spp. and Schiza-
chyrium spp.), low panic grasses (Dichanthelium spp.), pineywoods dropseed (Sporobo-
lus junceus) and wiregrass (Aristida beyrichiana). Collection data are: Florida,
Okaloosa Co., Eglin Air Force Base, T1S-R25W-sec. 30, 31-V-1995, 21-IX-1995, 01-VI-
1996, 16-VI-1997, D-Vac and sweep net, 70% EtOH, Longleaf Pine Restoration Project,
Site 2C-W.

Scientific Notes

A second series of S. floridanus specimens was collected by D-Vac in the Apalach-
icola National Forest (ANF), Florida. Specimens were also taken from longleaf pine-
wiregrass habitats that experience regular fires. Collection data are: Florida, Liberty
Co., Apalachicola Nat. For., Hwy 379 NW Sumatra, Compartment 95, 11-VI-1997, 30-
X-97; Compartment 100, 30-X-1997; Hwy 85 N of Wilma, Compartment 11, 14-VIII-
1997, 30-X-1997.
S. floridanus is a distinctive sminthurid, due to the sharp contrast between the
dark blue dorsum and the yellowish venter, plus the acuminate dorsal protuberance
anteriad of the anal papilla (habitus in Snider [1982: 223] and Borror et al. [1989:
167]). All EAFB and ANF specimens exhibit this dorsal protuberance. Voucher speci-
mens from EAFB and ANF are deposited in the Entomology Museum, Michigan State
University (East Lansing) and Florida State Collection ofArthropods (Gainesville).
Gymnoscirtetes morse Hebard is one of two species in this genus. Both species are
found in the southeastern United States and can be recognized by their small size and
lack of any trace of wings or wingpads in the adult. G. morse was described from De-
Funiak Springs, Florida (Hebard 1918) and has since been found only in the Florida
Panhandle between the ANF and Mobile, Alabama, and in some adjacent Alabama
counties. Collection records in the Florida State Collection ofArthropods and other in-
stitutions, notably the University of Michigan (T. J. Cohn, Museum of Zoology, Uni-
versity of Michigan, personal communication) all date from before the early 1950s.
In 1996, four specimens were collected in June and August on Hurlburt Field, lo-
cated on 6,634 acres in south Okaloosa County west of Mary Esther and Ft. Walton
Beach, Florida. Hurlburt Field contains a variety of habitats, including wet longleaf
pine savannah, where the specimens of G. morse were collected. This collection local-
ity was burned in early 1997. In April, 1997, large numbers of G. morse nymphs were
observed in the burned area. The grasshopper was very common early in the growing
season and became somewhat less abundant as the summer progressed. Voucher
adults were first collected during a visit in June, the last adults of the year were seen
in late October, and nymphs were collected again in April, 1998. G. morse appears to
favor a moist microclimate: grasshoppers jumped up from near the ground to the tops
of grasses and low bushes when approached. One female was observed feeding on
leaves of gallberry (Ilex glabra). Several other adults were confined in plastic bags
with various common plants from their groundcover habitat; only gallberry leaves
showed any evidence of feeding after several days. Collection records for Hurlburt
Field are: Florida, Okaloosa Co., Ft. Walton Beach, Hurlburt Field, 9-VIII-1996; same
locality, S of EOD [Explosive Ordinance Disposal], 27-VI-1997, 11-VII-1997, 17-IV-
Another population of G. morse was found in 1997 on Whitmier Island, a wet prai-
rie on the northern border of EAFB. The groundcover habitat where the G. morse
were taken was very similar to the Hurlburt Field groundcover layer. Collection
records from this locality are: Florida, Santa Rosa Co., Whitmier Island, T1N-R26W-
sec. 19, 23-VIII-1997, Eglin AF Base, W side of RR 717, ex wet prairie. Voucher spec-
imens from Hurlburt Field and EAFB are deposited in the Florida State Collection of
Arthropods (Gainesville).
Gymnoscirtetes morse is distinguishable from the more widespread G. pusilla
Scudder by the external male genitalia (Blatchley 1920); however, descriptions of the
aedeagi of the two species have not been published. There is some evidence that G.
morse may be conspecific with G. pusilla (G. Folkerts, Dept. Zoology and Wildlife, Au-
burn University, personal communication). Both morphospecies are found most com-
monly in the herbaceous groundcover under open pine canopy. Moist flatwoods, where
pitcher plants grow and where gallberry is present, seem particularly favored by

Scientific Notes

Until recently, there has been little study of arthropod communities in fire-main-
tained longleaf pine habitats (Folkerts et al. 1993). This could be due to an erroneous
perception that burned areas have less zoological richness; even entomologists some-
times misunderstood the importance of fire in maintaining the rich biodiversity of
longleaf pine ecosystems (cf. Klots 1951: 33). As arthropod faunas of the longleaf pine
landscape become better known, discoveries of new species and rediscoveries of rare
and uncommonly collected species will become increasingly frequent.
We thank R. J. Snider for identifying S. floridanus and for bringing the status of
this species to KEMG's attention. T. J. Cohn, D. R. Gordon, B. J. Herring, and L.
Provencher provided useful comments on the manuscript.
The Apalachicola National Forest study is funded by U.S. Department of Agricul-
ture grant FLAX 96-35101-3310 to Florida A&M University. Funding for the Hurl-
burt Field survey was provided by contract F862096MS456 to RWF from the
Department of Defense. Effort on Eglin Air Force Base sponsored by the USAMRAA,
U.S. Army Medical Research and Materiel Command, Department of the Army, under
cooperative agreement number DAMD17-98-2-8006. The U.S. Government is autho-
rized to reproduce and distribute reprints for Governmental purposes notwithstand-
ing any copyright notation thereon.
The views and conclusions contained herein are those of the authors and should not
be interpreted as necessarily representing the official policies or endorsements, either ex-
pressed or implied, of the USAMRAA, U.S. Army, U.S. Air Force, or the U.S. Government.


Research on Eglin Air Force Base, Hurlburt Field, and the Apalachicola National
Forest has led to rediscovery in Florida of the springtail Sminthurus floridanus (Col-
lembola: Sminthuridae) and the grasshopper Gymnoscirtetes morse (Orthoptera:
Acrididae). S. floridanus was collected in fire-maintained longleaf pine/wiregrass
stands. G. morse was found in wet areas, including longleaf pine savannahs, flat-
woods, and prairies.


BIONDO, B. 1997. In defense of the longleaf pine. Nature Conservancy 47(4): 10-17.
BLATCHLEY, W. S. 1920. Orthoptera of northeastern America. The Nature Publ. Co.,
Indianapolis, IN. 784 pp.
BORROR, D. J., C. A. TRIPLEHORN, AND N. F. JOHNSON. 1989. An introduction to the
study of insects, 6th ed. Saunders College Publ., Philadelphia, PA. 875 pp.
FOLKERTS, G. W., M. A. DEYRUP, AND D. C. SISSON. 1993.Arthropods associated with
xeric longleaf pine habitats in the southeastern United States: a brief overview,
pp. 159-192 in S. M. Hermann [ed.], Proceedings of the Tall Timbers Fire Ecol-
ogy Conference No. 18, Tall Timbers Research Station, Tallahassee, FL.
HEBARD, M. 1918. New genera and species of Melanopli found within the United
States. Transactions of the American Entomological Society 44: 141-169.
D. R. STRENG, J. L. WALKER, AND R. L. MYERS. 1998. Fire and biodiversity:
studies of vegetation and arthropods pp. 384-401 in G. W. Kelly [ed.], Trans. 63"'
No. Amer. Wildl. and Natur. Resour. Conf. Vol. 63.
KLOTS, A. B. 1951. A field guide to the butterflies. Peterson Field Guide Series (1958
ed.). Houghton Mifflin Co., Boston, 349+xvi pp.
MACGILLIVRAY, A. D. 1893. North American Thysanura. Canadian Entomol. 25: 127-
SNIDER, R. J. 1982. Redescription of Sminthurus floridanus MacGillivray, 1893 (Col-
lembola: Sminthuridae). Florida Entomol. 65: 221-227.

Scientific Notes


'Dow AgroSciences, 2608 S. Dundee Blvd., Tampa, FL 33629

2North Florida Research and Education Center, University of Florida, Route 3 Box
4370, Quincy, FL 32351-9500

The western flower thrips Frankliniella occidentalis (Pergande) is a very serious
worldwide pest of ornamental, vegetable, and fruit crops in the field and greenhouse
(Tommasini and Maini 1995). It is an efficient vector for tomato spotted wilt virus, a
serious disease of a wide variety of plants, including vegetable, flower, and ornamen-
tal crops (Allen et al. 1990). Western flower thrips are difficult to control effectively
with insecticides (Brodsgaard 1994), and resistance has developed to organophos-
phate, carbamate, pyrethroid, and macrocyclic lactone insecticides after repeated ex-
posure (Immaraju et al. 1992).
Spinosad (DowElanco, Indianapolis, IN), a new natural macrocyclic lactone insect
control product with a unique mode of action, was highly efficacious against F. occi-
dentalis in field experiments with pepper conducted in North Florida during 1996 and
1997 (J. E. F, J. S., and S. M. Olson, unpublished data). Another abundant flower
thrips species, F. tritici (Fitch), was not significantly suppressed by spinosad in these
experiments and spinosad did not have detrimental effects on populations of the
minute pirate bug, Orius insidious (Say), a key predator ofF. occidentalis. Thus, spi-
nosad has the potential to be an important new tool for managing F. occidentalis. For
this reason and because Frankliniella species differ in their ability to transmit tomato
spotted wilt virus (Sakimura 1962, 1963), an understanding of the comparative tox-
icity of spinosad to various species of Frankliniella is needed. Knowledge of spinosad
toxicity and the development of an effective bioassay will facilitate resistance moni-
toring for these pests.
Previous resistance or efficacy bioassays for F. occidentalis have employed either
topical application (Robb 1989), detached leaves as a substrate for the insecticide
(e.g., Immaraju et al. 1992), or a residue-on-glass technique (Brodsgaard 1994). Our
objectives were to develop an insecticide bioassay procedure suitable for three com-
mon species of flower thrips in Florida, F. occidentalis, F. tritici, and F. bispinosa (Mor-
gan), and to determine the toxicity of spinosad to these flower thrips species as a
possible explanation for control differences in field plots.
Flower thrips were collected from wild radish, Raphanus raphanistrum L., grow-
ing 50-300 m from pepper fields at the North Florida Research and Education Center
of the University of Florida in Quincy. Collection dates were May 27-29, 1997. Adults
were aspirated into glass tubes (6 mm diam.) and then emptied into individual plastic
diet cups (35 ml) with uncoated paper caps. The cups were provided with sections (20
mm) of snap bean pods sealed at either end with a thin layer of paraffin. After sealing,
bean pod sections were submerged for 30 sec in 10 different concentrations of a 0.24
kg ai/1 suspension concentrate of spinosad (SpinTor 2SC, Dow AgroSciences, India-
napolis, IN) in distilled water, and allowed to air dry for 1 hr. Individual cups were
placed into a sealed plastic rearing container (5.7 liter), the bottom of which was cov-
ered with moist paper toweling. These containers were held in a controlled-environ-
ment chamber maintained at 23 + 2C, 60% RH and a photoperiod of 14: 10 (L: D).
The trial was replicated four times with three diet cups per replicate.

Florida Entomologist 81(4)

December, 1998

Because of the fragile nature of thrips and the difficulty in separating living indi-
viduals into the three species, no attempt was made to standardize the numbers of
each species or the total number of individuals placed in each diet cup. We attempted
to place a minimum of 20 individuals into each cup. Table 1 lists the range and mean
numbers of each species used in this trial. These numbers are representative of the
relative species abundance on wild radish on the collection dates. If less than 5 of any
one species was present in any replicate, that replicate was repeated the following day
using newly prepared solutions.
Although bioassay development will not be dealt with in detail here, a few obser-
vations are relevant. We initially evaluated glass snap-cap vials in addition to plastic
diet cups as bioassay containers. The vials had a small opening which made them
more difficult to use and plastic caps which promoted the formation of excess mois-
ture, thus diet cups were chosen for further evaluation. We evaluated thrips survival
in empty diet cups, in cups with bean sections only, with a small piece of moistened pa-
per toweling only, and with both bean sections and toweling. Thrips survival at 24 hrs
was minimal in empty cups. The paper toweling resulted in excess moisture which
trapped and drowned some thrips. Cups with snap bean sections only resulted in the
highest (virtually 100%) survival. Uncoated paper caps were chosen over wax coated
caps because the latter resulted in excessive moisture in the cups. The sealed plastic
rearing containers with moist paper toweling did not promote excess moisture in the
cups, but moisture from the paper toweling did result in a slight expansion of the pa-
per caps to provide a better seal and less desiccation of bean sections. The ends of bean
sections were coated with paraffin to reduce desiccation and to serve as a barrier to
prevent thrips from crawling inside the bean section. Finally, we chose to treat only
the bean sections and not the cup itself because preliminary trial observations sug-
gested that thrips spent most of their time on the bean sections.
Mortality was evaluated at 24 + 1 hrs. Thrips were considered dead if they were
unable to stand upright and/or move forward when probed. Individuals were segre-
gated into living and dead, placed in alcohol and the respective numbers of each spe-
cies determined under a dissecting microscope. No mortality was observed in
untreated controls for F. occidentalis and F. tritici. For F. bispinosa, mortality (3%)
was observed in only one replicate. Mortality for the various doses were corrected for
control mortality (Abbott 1925). Data were analyzed with analysis of regression using
a log-probit model. The analytical software used was Statgraphics Plus (Manugistics,
Inc., Rockville, MD).
The most common species in our samples was F. bispinosa, while F. occidentalis was
the least common (Table 1). Numbers of F. bispinosa used in the bioassay were >3X
those of F. occidentalis and numbers of F. tritici were >2X those of F. occidentalis. In
contrast, Salguero Navas et al. (1991), also working in North Florida, found that F. oc-


Mean/ Range/
Species Replicate Replicate Total # Tested Sex Ratio F:M

F. occidentalis 8.8 5-21 379 2.14:1.0
F. tritici 18.3 7-42 803 2.00:1.0
F. bispinosa 30.7 14-62 1352 2.21:1.0
All species combined 57.6 34-87 2534 2.13:1.0

Scientific Notes

S/ Fankllnlella ocldentaills /' FrankHni/is tfi

z C D
W, -
80 po
IL s"o. F. tiiApinoas -

s i k i Franklinietll bisplnosa .
20 -b

.01 0,1 1 10 100 1000 10000 .01 0.1 1 10 100 1000 10000l


Fig. 1. Mortality of Frankliniella spp. in response to spinosad; responses ofF. occi
dentalis (A), F. tritici (B), and F. bispinosa (C), and all three species combined (D).
Solid lines are the predicted dose response curves (log-probit model) and dashed lines
represent corresponding 95% confidence intervals. Dose response curves for all three
species of Frankliniella are compared in Figure D.

cidentalis was generally the most abundant species of Frankliniella in tomato flowers
in the spring and F. bispinosa was relatively uncommon in their study. Our results may
represent a host preference of F. bispinosa for wild radish. An unusually warm winter
in 1996-97 may have also contributed to the differences. Sex ratios ofFrankliniella spp.
tested are also given in Table 1. There was roughly a 2 to 1 ratio of females to males
with only minor differences between species. Dose responses of males and females were
not significantly different for any species, so data for the two sexes were combined.
Dose-response curves were similar for all three Frankliniella species (Figure 1).
Regressions of dose/mortality data were highly significant for all three species (R2 =
75-85%, P < 0.00001) (Table 2). The regression slope for the F. occidentalis data was
significantly higher than that for the other two species based on non-overlapping
standard error values. Standard error values around the regression slopes for F. tritici
and F. bispinosa data did overlap, indicating that slopes for these species were not sig-
nificantly different. Although F. tritici was numerically more susceptible than F. bis-
pinosa as indicated by the lower LC,,, LC,,, LC,, and LC,, values, the 95% confidence
intervals around these values for the three species overlapped. Thus, there were no
significant differences among the three species.
Data presented here suggest that spinosad is equally toxic to the three species of
Frankliniella tested and that differential toxicity of spinosad to Frankliniella spp. is
probably not responsible for differences in relative species abundance between non-

Florida Entomologist 81(4)

December, 1998


F. occidentalis F. tritici F. bispinosa

R-Squared (%) 76.35 74.96 85.50
Probability Level < 0.00001 < 0.00001 < 0.00001
Slope (SE) 0.437 (0.040) 0.368 (0.035) 0.383 (0.026)
Intercept (SE) 0.227 (0.132) 0.427 (0.115) 0.124 (0.085)
LC,, (pg/ml) 0.032 0.0096 0.025
95% Confidence Limits 0.0097 0.075 0.0022 0.028 0.011 0.051
LC,, (pg/ml) 0.594 0.31 0.72
95% Confidence Limits 0.293 1.09 0.14 0.61 0.44 1.13
LC,0 (pg/ml) 11.19 10.18 20.59
95% Confidence Limits 6.07 23.03 5.47 20.95 12.73 35.93
LC,, (pg/ml) 122.50 173.98 315.96
95% Confidence Limits 53.80 368.86 72.72 568.82 156.89 760.15

treated and spinosad-treated field plots (J. E. F., J. S., and S. M. Olson, unpublished
data). Field rates of spinosad that have demonstrated activity against Frankliniella
spp. (75-100 g ai/ha) will result in concentrations of 50-200 ppm of spinosad in normal
application volumes. Although laboratory results may not translate directly to field
activity, these concentrations exceed concentrations needed to provide greater than
90% mortality of all three species based on our bioassay. Differences in species abun-
dance in spinosad-treated field plots may thus be due to factors other than differential
toxicity (e. g., migration or competition).
Populations ofF. occidentalis have been shown to have multiple resistance mech-
anisms and have developed cross-resistance to insecticides within the same chemical
group and to those in other classes (Zhao et al. 1995). Consequently, alternating in-
secticides from different classes with different modes of action as a sole resistance
management tactic poses risks. Insecticide selection pressure can be minimized by us-
ing noninsecticidal methods in conjunction with carefully selected insecticides used
only when needed. The efficacy of spinosad demonstrated in our research reported
here, combined with its compatibility with the key natural enemy of flower thrips,
make it a potentially important tool for integrated pest management programs. Fur-
ther, the baseline knowledge of spinosad toxicity reported here will help to develop a
resistance monitoring program to determine the effectiveness of integrated pest man-
agement programs in minimizing resistance development.
This article is Florida Agricultural Experiment Station Journal Series R-05991.


A bioassay technique was developed and used to determine the toxicity of spinosad
to three species of Frankliniella: F. bispinosa, F. occidentalis, and F. tritici. Dose re-
sponse curves for the three species were similar and regressions of dose/mortality
data were highly significant (R2 = 75-85%, P < 0.00001 for all species). 95% confidence
intervals around LClo, LCs, LC,, and LC,, values for the three species overlapped, sug-
gesting that there were no significant differences among the three species tested.

Scientific Notes


ABBOTT, W. S. 1925. A method of computing the effectiveness of an insecticide. J.
Econ. Entomol. 18: 265-267.
ALLEN, W. R, J. A. MATTEONI, AND A. B. BROADBENT. 1990. Susceptibility of cultivars
of florist's chrysanthemum to tomato spotted wilt virus. Can. J. Plant Pathol.
12: 417-423.
BRODSGAARD, H. F. 1994. Insecticide resistance in European and African strains of
western flower thrips (Thysanoptera: Thripidae) tested in a new residue-on-
glass test. J. Econ. Entomol. 87: 1141-1146.
Western flower thrips (Thysanoptera: Thripidae) resistance to insecticides in
coastal California greenhouses. J. Econ. Entomol. 85: 9-14.
ROBB, K. L. 1989. Analysis of Frankliniella occidentalis (Pergande) as a pest of flori-
cultural crops in California greenhouses. Ph.D. Dissertation, University of Cal-
ifornia, Riverside.
SAKIMURA, K. 1962. Frankliniella occidentalis (Thysanoptera: Thripidae), a vector of
the tomato spotted wilt virus, with special reference to the color forms. Ann.
Entomol. Soc. Am. 55: 387-389.
SAKIMURA, K. 1963. Frankliniella fusca, an additional vector of the tomato spotted
wilt virus, with notes on Thrips tabaci, another vector. Phytopathology 53: 412-
MACK. 1991. Seasonal patterns of Frankliniella spp. (Thysanoptera: Thripidae)
in tomato flowers. J. Econ. Entomol. 84: 1818-1822.
TOMMASINI, M. G., AND S. MAINI. 1995. Frankliniella occidentalis and other thrips
harmful to vegetable and ornamental crops in Europe, p. 1- 42. in: A. J. M.
Loomans, J. C. van Lenteren, M. G. Thomasini, S. Maini, and J. Riudavets
(eds.). Biological Control of Thrips Pests. Wageningen Agric. Univ. Papers 95-1,
Wageningen, The Netherlands.
ZHAO, G., W. LIU, J. M. BROWN, AND C. O. KNOWLES. 1995. Insecticide resistance in
field and laboratory strains of western flower thrips (Thysanoptera: Thripidae).
J. Econ. Entomol. 88: 1164-1170.

Florida Entomologist 81(4)

December, 1998


'University of Florida, Fort Lauderdale Research & Education Center,
3205 College Avenue, Fort Lauderdale, Florida 33314

2 Florida Department of Agriculture and Consumer Services, Division of Plant
Industry, P. O. Box 147100, Gainesville, Florida 32614

Two subfamilies of aphids (Aphididae), Pemphiginae and Hormaphidinae, exhibit
intra- and interspecific aggression. For example, Pemphigus betae Doane (Pem-
phigidae) fights duels for feeding sites on Populus angustifolia James, in which two
aphids may kick and shove each other for up to two days (Whitham 1979). Foster
(1996) recently described duels for feeding sites among colony mates ofAstegopteryx
minute van der Goot (Hormaphidinae). Interspecific aggression in these subfamilies
is displayed in the soldier caste of some species (Aoki 1977; Foster 1990).
In its native range in Southeast Asia, the palm aphid, Cerataphis brasiliensis
(Hempel) (Hormaphidinae), alternates between a dicotyledonous tree, Styrax benzoin
Dryand, where colonies form galls, and palms where they live externally on green tis-
sue. The gall-inhabiting colony has a soldier caste that attempts to protect the colony
from predators (Stern et al. 1995).
Cerataphis spp. have been introduced into many tropical areas and survive exclu-
sively on palms where Styrax or other suitable alternate hosts are not present. They
are pests of palms in some countries (Enobakhare 1994; Reinert and Woodiel 1974).
Flat, circular and aleyrodid-like (Fig. 1), all stages bear a pair of minute spikes, or
'horns', on the front of the head. Some species additionally have several pairs of
minute dagger-like setae on the ventral side of the head. The function of these struc-
tures has been presumed to be of an offensive or defensive nature but this presump-
tion has not been confirmed until now.
Palm aphids are common on coconut palm, Cocos nucifera L., and several other
palm species in southern Florida. They usually occur on the unopened frond and the
youngest two or three fronds and sometimes on young fruits. They normally remain
motionless, apparently feeding for long periods. They are associated with ants, and
show the typical mutualistic ant-aphid relationship involving protection by the ants
in exchange for honeydew.
One of our students, Claudia Vanderbilt, called our attention for the first time to
two palm aphids involved in an altercation. Since then, we have observed dueling
palm aphids about 15 times, have videotaped three dueling pairs (running time ap-
proximately 60 minutes), and report the behavior in this note.
The altercations that we observed were on excised palm frond tissue under the mi-
croscope in a laboratory at about 23C. When manipulated slightly with a probe made
of a human hair, the aphids sometimes secrete a small drop of honeydew, as aphids in
general do when ants manipulate them. With insistent probing, they retract their
stylets from the palm tissue and begin to crawl.
Disturbed aphids wander slowly and randomly. Occasionally one aphid encounters
another aphid that is motionless and apparently feeding. The moving aphid some-
times explores the feeding aphid with its antennae and then moves on; at other times
the moving aphid butts the outer margins of the feeding aphid with its horns. After
several butts, the feeding aphid typically rotates a few degrees and appears to with-
draw from feeding. Moving slowly, the aphid that was feeding turns to face the in-

Scientific Notes

Fig. 1. Dueling palm aphids.

truder. Shortly, the aphids engage in a butting duel. To butt another aphid, an aphid
lowers its head, places its horns beneath the head of the other aphid, then snaps its
head upward while simultaneously thrusting forward with the legs. The motion often
lifts the other aphid at its margin. Each of the dueling pair responds to being butted
within a few seconds by butting its opponent. The altercation may last up to 19 min-
utes, the aphids often exchanging blows about 40 times per minute and alternately
resting for intervals of several minutes. We were unable to observe any role of the
minute dagger-like setae mentioned above.
The dueling aphids usually seemed well matched, even when younger nymphs
challenge larger adults. Neither one seemed to be injured by the other, and one gained
ground over the other only after prolonged butting. It was not clear to us what factors
brought about the end of these duels. Often, in what would appear to be the middle of
a well matched duel, the opponents appeared to pause slightly, after which one of
them would climb upon the other, rotate clockwise, remain for a few minutes, then
climb down and walk off; some aphids would then encounter another aphid and begin
a new duel. Some altercations between three aphids simultaneously were observed, in
which case one aphid would be butted from both sides or from front and back. Only
one aphid that displaced a feeding aphid appeared to occupy the loser's feeding site,
as observed by Foster (1996) forA. minute.
The aphids readily engaged each other upon contact, but ignored other small ar-
thropods placed with them, including nymphs and adults of brown citrus aphids (Tox-
optera citricida Kirkaldy) and nymphs of psyllids (Ceropsylla sideroxyli Riley), the
latter which were of similar size and shape as the palm aphids. The palm aphids did

Florida Entomologist 81(4)

December, 1998

not attempt to defend themselves against a small larva of a coccinellid beetle that at-
tacked and consumed several of them.
Dueling among palm aphids may usually be for feeding sites, since this is known
in their close relatives (Foster 1996; Whitham 1979). Aphids generally have highly
specific requirements not only in their host plants, but in the sites on the plant in
which they feed (Dixon 1985). Also, palm tissue is notoriously tough and fibrous. Per-
haps an aphid expends less energy in displacing a feeding aphid than in finding a suit-
able new site and penetrating it.
However, the objective of the dueling was not apparent in our observations. More
extensive investigation of the palm-inhabiting phase of palm aphids in nature may
further elucidate this behavior.
We thank Scott Bryan for technical assistance. This work was partially supported
by a University of Florida Research Project Enhancement Award to FWH. This is
Florida Agricultural Experiment Station Journal Series No. R-06299


Duels between palm aphids, Cerataphis brasiliensis, infesting palm tissue were
observed and videotaped in the laboratory. The objective of this aggression was not
clear. In nature they presumably duel for feeding sites, as do related species. This is
the first report of intraspecific aggression in the palm-infesting phase of palm aphids.


AOKI, S. 1977. Colophina clematis (Homoptera: Pemphigidae), an aphid species with
"soldiers". Kontyfi, Tokyo 45: 276-282.
DIXON, A. F. G. 1985. Aphid Ecology. Glasgow and London: Blackie. 157 p.
Enobakhare, D. A. 1994. Occurence and distribution of Cerataphis palmae (Ghes-
quierei) (Homoptera: Pemphigidae) on Raphia palms in southern Nigeria. In-
sect Science and its Applications 15(1): 101-104.
FOSTER W. A. 1990. Experimental evidence for effective and altruistic colony defence
against natural predators by soldiers of the gall-forming aphid Pemphigus spy-
rothecae (Hemiptera: Pemphigidae). Behavioral Ecology and Sociobiology 27:
FOSTER, W. A. 1996. Aphids: Duelling aphids: intraspecific fighting in Astegopteryx
minute (Homoptera: Hormaphididiae). Animal Behavior 51: 645-655.
REINERT, J. A., AND N. L. WOODIEL 1974. Palm aphid control on MValayan dwarf' co-
conut palms. Florida Entomologist 57(4): 411-414.
STERN, D. L., S. AOKI, AND D. U. KUROSU 1995. The life cycle and natural history of
the tropical aphid Cerataphis fransseni (Homoptera: Aphididae: Hormaphidi-
nae), with reference to the evolution of host alternation in aphids. Journal of
Natural History 29: 231-242.
WHITHAM, T. G. 1979. Territorial behaviour of Pemphigus gall aphids. Nature
279(5711): 324-325.

Scientific Notes


SCenter for Medical, Agricultural and Veterinary Entomology, Agricultural Research
Service, U.S. Department of Agriculture, Gainesville, FL 32604

2 Department of Entomology and Nematology, University of Florida
Gainesville, FL 32611-0630

Several efforts have been made to import natural enemies of Scapteriscus mole
crickets since their arrival in the southeastern USA about 1900. One of these natural
enemies was the sphecid wasp Larra bicolor F. Populations ofL. bicolor from Bolivia
were released between October 1988 and June 1989, and became established in Ala-
chua County, Florida (Frank et al. 1995). Since there is no demonstrated method to
sample for L. bicolor, the nectar-bearing plant Spermacoce verticillata L. was estab-
lished near several release sites so that wasp visits could be observed. Wasps were not
observed until the fall of 1993, and wasps continued to be seen through September
1994. By 1995, it was concluded that L. bicolor had dispersed at least a distance of 4
km from release sites (Frank et al. 1995). Our note documents the collection ofL. bi-
color adults in agricultural fields in northwestern Alachua County that are at least 22
km from the original release sites.
From 18 June to 10 October 1997, white plastic funnel traps ("bucket" or Universal
Moth Traps, International Pheromone Systems, Wirral, Merseyside, England) were
placed in an area planted to over 470 ha. of cotton, Gossypium hirsutum L., to attract
beet armyworm, Spodoptera exigua (Hibner) and fall armyworm, S. frugiperda (J. E.
Smith) (Lepidoptera: Noctuidae). Traps were baited with either commercially-pro-
duced sex pheromones, phenylacetaldehyde (CHCH2CHO, a floral attractant ob-
tained from Aldrich Chemical Co., Milwaukee, WI) in plastic caps (20 mm diameter,
13 mm height; 0.2 or 0.5 ml phenylacetaldehyde per cap), or a combination of phero-
mone and phenylacetaldehyde. Pheromone lures were attached to the bottom of a cork
that was placed in a hole in the canopy of the bucket trap. The phenylacetaldehyde
cap was hot-gun glued (Arrow Fastener Co., Saddle Brook, NJ) to the bottom of the
cork, which was placed in the trap canopy. The combination lure was composed of a
cork with attached cap and the pheromone lure attached to the outside of the cork.
Three tests were conducted in separate cotton fields. The first used Hercon (Hercon
Environmental Corp., Emigsville, PA) pheromone lures for S. exigua, the second used
Scentry (Ecogen, Inc., Langhorne, PA) lures for S. exigua, and the third used Trece
(Trece, Inc., Salinas, CA) lures for S. frugiperda. Four replications of the three treat-
ments were placed within each field along pivot roads or along the field edges. Traps
were observed three times weekly and pheromone and phenylacetaldehyde lures were
replaced every two weeks.
Larra bicolor was collected in 2 of the 3 fields over seven different dates. The first
wasp was collected in the Hercon field 18 June, with subsequent collections in the
Tr6ec field 23 June (1 collected) and 11 July (2 collected). Higher numbers of wasps
were collected in the fall, as 53 wasps were found in late September--early October.
Peak capture was 29 September when 48 wasps were collected over a 5 day period in
6 separate traps. Of the total 57 L. bicolor collected, 32 were found in traps baited
with the pheromone-phenylacetaldehyde combination, 23 were found in the phenylac-
etaldehyde-baited traps, and only 2 were found in the pheromone-baited traps.

Florida Entomologist 81(4)

Nontarget Hymenoptera have been collected in bucket traps placed in field crops
which were baited for several different noctuid species (Adams et al. 1989, Mitchell et
al. 1989, Gauthier et al. 1991), however, this is the first report for collection ofL. bicolor.


Larra bicolor was collected as a nontarget species in white bucket traps baited
with sex pheromones and the floral attractant phenylacetaldehyde in an agricultural
area in northwestern Alachua County, Florida. The first wasp was collected in mid-
June, but larger numbers of wasps were collected in late September and early Octo-
ber. More wasps were collected in traps that had phenylacetaldehyde as a lure. This
collection method may aid researchers in determining the dispersal and effectiveness
of this natural enemy of Scapteriscus mole crickets.


ADAMS, R. G., K. D. MURRAY, AND L. M. LOS. 1989. Effectiveness and selectivity of sex
pheromone lures and traps for monitoring fall armyworm (Lepidoptera: Noctu-
idae) adults in Connecticut sweet corn. J. Econ. Entomol. 82: 285-290.
FRANK, J. H., J. P. PARKMAN, AND F. D. BENNETT. 1995. Larra bicolor (Hymenoptera:
Sphecidae), a biological control agent of Scapteriscus mole crickets (Ortho-
ptera: Gryllotalpidae), established in northern Florida. Fla. Entomol. 78: 619-
AND R. G. ADAMS. 1991. Field bioassay of pheromone lures and trap designs for
monitoring adult corn earworm (Lepidoptera: Noctuidae) in sweet corn in
southern New England. J. Econ. Entomol. 84: 1833-1836.
MITCHELL, E. R., H. R. AGEE, AND R. R. HEATH. 1989. Influence of pheromone trap
color and design on capture of male velvetbean caterpillar and fall armyworm
moths (Lepidoptera: Noctuidae). J. Chem. Ecol. 15: 1775-1784.


December, 1998

Florida Entomologist 81(4)

December, 1998


Center for Medical, Agricultural and Veterinary Entomology, Agricultural Research
Service, U.S. Department of Agriculture, Gainesville, FL 32604

The fall armyworm, Spodoptera frugiperda (J. E. Smith), is a migratory polypha-
gous pest that attacks several important crops (Luginbill 1928). Currently, adult male
populations are monitored using a synthetic blend of sex pheromone components as a
lure (Tumlinson et al. 1986, Mitchell et al. 1989). Chemicals other than sex phero-
mones have been assayed as moth attractants. For instance, floral compounds that at-
tract noctuid moths have been isolated and identified. Baits of phenylacetaldehyde,

Scientific Notes

first isolated from a flower (Araujia sericofera Brothero) (Asclepiadaceae), captured
hundreds of noctuid moths in field traps, including S. frugiperda (Cantelo & Jacobson
1979). Phenylacetaldehyde, benzaldehyde, 2-phenylethanol, and benzyl alcohol were
identified from flowers of the shrub Abelia grandiflora (Andre) (Caprifoliaceae)
(Haynes et al. 1991), a plant that elicits flight responses from cabbage looper, Trichop-
lusia ni (Hibner) (Grant 1971). Benzaldehyde, benzyl acetate, and phenylacetalde-
hyde were collected from night-blooming jessamine Cestrum nocturnum (L.)
(Solanaceae), a plant known to attract looper moths (Noctuidae: Plusiinae) (Heath et
al. 1992).
Phenylacetaldehyde in combination with sex pheromones or blacklights increased
moth trap capture (Smith et al. 1943, Creighton et al. 1973, Cantelo & Jacobson
1979). Both male and female T ni were attracted to phenylacetaldehyde in flight tun-
nel, greenhouse, and screen cage bioassays (Haynes et al. 1991, Landolt et al. 1991,
Heath et al. 1992). Our objective was to determine if phenylacetaldehyde enhances
the attractiveness of sex pheromones to fall armyworm in a flight tunnel bioassay.
Fall armyworms used in the bioassays were reared in the laboratory on a pinto
bean-based artificial diet according to the procedures of Guy et al. (1985). Pupae were
sexed and placed in 163 ml (5.5 oz.) paper cups (Sweetheart, Chicago, IL) that were
placed in 24 x 24 cm screen cages for eclosion. Pupae were maintained under reversed
photoperiod (14: 10, light:dark) in an environmental chamber held at 26C and 70%
RH. Adults had access to cotton balls saturated with distilled water and a honey-
sugar solution. Pupae were transferred daily so that cages contained adult males of a
known age.
The tunnel used was a Plexiglas rectangular box (2.0 by 0.6 by 0.6 m). The floor
had alternating black and white panels (ea. panel 10 cm long). Air was pulled past an
activated charcoal filter and through the tunnel at the rate of 0.22 m/sec by a blower
motor. A cylindrical moth release cage (9 cm by 5.1 cm diameter) and a two-compart-
ment source cage (6 cm long by 2 cm diameter), both made from metal screen, were
hung in the middle at the downwind and upwind portions of the tunnel, respectively.
Distance between the cages was 1.4 m. Room conditions during testing were 26C
and 65-80% RH, and observations were aided by overhead red lights.
Adults were placed in the tunnel room at least one hour before testing and tests
were conducted 1-4 h post-scotophase. The commercial pheromone lures used were
purchased from Scentry (Ecogen, Inc., Langhorne, PA) and from Trece (Trece, Inc.,
Salinas, CA). A standard lure containing 2 mg of the pheromone blend (Z)-9-tetrade-
cen-l-ol acetate (Z9-14: AC) (80.3%), (Z)-ll-hexadecen-l-ol acetate (Z11-16: AC)
(19.2%), and (Z)-7-dodecen-l-ol acetate (Z7-12: AC) (0.5%), loaded in a solvent refined
rubber septum, was prepared by J. H. Tumlinson (USDA-ARS CMAVE). A hexane so-
lution of 10 mg/ml of phenylacetaldehyde (Aldrich Chemical Co., Milwaukee, WI) was
prepared, and 100 pl of this solution was pipetted onto filter paper. The bioassay pro-
tocol was to place an individual moth in the release cage, hold it in the plume for 10
seconds, and then release it down the tunnel. Each moth was observed for 2 min and
upwind flight and contact with the source screen cage was scored. Each replicate con-
tained between 5 and 20 moths, with totals of from 70-150 moths tested per treat-
ment. The experiment was designed as a randomized complete block, and percentage
responses were transformed into arcsine-square roots before analysis of variance
(PROC GLM, SAS Institute 1995). Comparisons between each pheromone lure and
lure plus phenylacetaldehyde combination were tested using the contrast statement
The addition of phenylacetaldehyde in the source cage increased upwind flight and
contact with the lure in all combinations tested. Phenylacetaldehyde in combination

Florida Entomologist 81(4)

December, 1998

with the standard lure increased upwind flight from 68.6% + 5.6 (SE) to 87.4% + 7.3
(df = 1, 28; F = 12.5; P = 0.0014), and increased contact with the source cage from
51.9% + 7.6 to 76.6% + 6.1 (df = 1, 28; F = 8.0; P = 0.0085). The combination of pheny-
lacetaldehyde with a Scentry lure increased upwind flight from 61.1% + 6.3 to 80.0%
+ 9.1 (df= 1, 28; F = 4.3; P = 0.0468, and increased contact from 30.8% + 6.3 to 56.3%
+ 9.0 (df = 1, 28; F = 3.9; P = 0.0584). Similar results were obtained with a Tr6c6 lure,
upwind flight increased from 66.5% + 6.8 to 88.3% + 3.3 (df = 1, 28;F = 9.8; P = 0.0041)
and contact increased from 48.7% + 8.0 to 75.0% + 8.7 (df = 1, 28; F = 10.9; P = 0.0026)
when phenylacetaldehyde was added. No differences were found in upwind flight or
contact among lures lacking phenylacetaldehyde (P > 0.05), and fall armyworm males
did not respond to phenylacetaldehyde alone (n = 50).
Phenylacetaldehyde generally has enhanced trap catch of moths in pheromone
traps or with blacklights (Creighton et al 1973, Cantelo & Jacobson 1979). Our study
is the first to show that a floral compound such as phenylacetaldehyde can increase
attraction of S. frugiperda males to a pheromone source. Pheromone-baited traps for
S. frugiperda have been used to detect seasonal population trends (Tingle & Mitchell
1977, Waddill et al. 1982), document migration patterns over large areas (Mitchell et
al. 1991), and predict larval populations and plant infestation (Silvain & Ti-A-Hing
1985, Silvain 1986, Linduska & Harrison 1986). Phenylacetaldehyde as an enhance-
ment may increase trap capture in the field, thereby improving current uses of pher-
omones and potentially creating new lure and toxicant systems for management of
this pest (Landolt et al. 1991).
The authors thank J. H. Tumlinson (USDA-ARS CMAVE) for preparation of stan-
dard S. frugiperda pheromone lures. Technical support was provided by J. Brady, M.
Lanh and S. Lovvorn. The authors also thank personnel of the Crop Insect Rearing
Staff, USDA-ARS CMAVE for fall armyworm cultures.


More male fall armyworms, Spodoptera frugiperda (J. E. Smith), flew upwind to
combinations of pheromone-treated septa and phenylacetaldehyde than to phero-
mone-treated septa alone in flight tunnel bioassays. No moths flew upwind to pheny-
lacetaldehyde alone at the dose tested. This compound may increase pheromone-
baited trap captures in the field, thereby improving current uses of pheromones and
potentially creating new lure and toxicant systems for management of this pest.


CANTELO, W. W., AND M. JACOBSON. 1979. Phenylacetaldehyde attracts moths to
bladder flower and to blacklight traps. Environ. Entomol. 8: 444-447.
CREIGHTON, C. S., T. L. MCFADDEN, AND E. R. CUTHBERT. 1973. Supplementary data
on phenylacetaldehyde: an attractant for Lepidoptera. J. Econ. Entomol. 66:
GRANT, G. G. 1971. Feeding activity of adult cabbage loopers on flowers with strong ol-
factory stimuli. J. Econ. Entomol. 64: 315-316.
1985. Trichoplusia ni, pp. 487-494. In P. Singh & R. F. Moore [eds.], Handbook
of insect rearing, vol. 2. Elsevier, Amsterdam.
HAYNES, K. F., J. Z. ZHAO, AND A. LATIF. 1991. Identification of floral compounds from
Abelia grandiflora that stimulate upwind flight in cabbage looper moths. J.
Chem. Ecol. 17: 637-646.
HEATH, R. R., P. J. LANDOLT, B. DUEBEN, AND B. LENCZEWSKI. 1992. Identification of
floral compounds of night-blooming jessamine attractive to cabbage looper
moths. Environ. Entomol. 21: 854-859.

Scientific Notes

LANDOLT, P. J., B. LENCZEWSKI, AND R. R. HEATH. 1991. Lure and toxicant system for
the cabbage looper (Lepidoptera: Noctuidae). J. Econ. Entomol. 84: 1344-1347.
LINDUSKA, J. J., AND F. P. HARRISON. 1986. Adult sampling as a means of predicting
damage levels of fall armyworm (Lepidoptera: Noctuidae) in grain corn. Florida
Entomol. 69: 487-491.
LUGINBILL, P. 1928. The fall armyworm. USDA Tech. Bull. 34. 92 p.
MITCHELL, E. R., H. R. AGEE, AND R. R. HEATH. 1989. Influence of pheromone trap
color and design on capture of male velvetbean caterpillar and fall armyworm
moths (Lepidoptera: Noctuidae). J. Chem. Ecol. 15: 1775-1784.
PROSHOLD. 1991. Seasonal periodicity of fall armyworm, (Lepidoptera: Noctu-
idae) in the Caribbean basin and northward to Canada. J. Entomol. Sci. 26: 39-
SAS INSTITUTE. 1995. SAS/STAT guide for personal computers, version 6.11 ed. SAS
Institute, Cary, NC.
SILVAIN, J. F. 1986. Use of pheromone traps as a warning system against attacks of
Spodoptera frugiperda larvae in French Guiana. Florida Entomol. 69: 139-147.
SILVAIN, J. F., AND J. TI-A-HING. 1985. Prediction of larval infestation in pasture
grasses by Spodoptera frugiperda (Lepidoptera: Noctuidae) from estimates of
adult abundance. Florida Entomol. 68: 686-691.
SMITH, C. E., N. ALLEN, AND O. A. NELSON. 1943. Some chemotropic studies withAu-
tographa spp. J. Econ. Entomol. 36: 619-621.
TINGLE, F. C., AND E. R. MITCHELL. 1977. Seasonal populations of armyworms and
loopers at Hastings, Florida. Florida Entomol. 60: 115-122.
1986. Sex pheromone of fall armyworm, Spodoptera frugiperda (J. E. Smith):
identification of components critical to attraction in the field. J. Chem. Ecol. 12:
Seasonal abundance of the fall armyworm and velvetbean caterpillar (Lepi-
doptera: Noctuidae) at four locations in Florida. Florida Entomol. 65: 350-354.


Scientific Notes


Everglades Research and Education Center, P.O. Box 8003, Belle Glade, FL 33430

Plecia nearctica Hardy is the lovebug that motorists frequently encounter as a se-
rious nuisance when traveling in southern states. The insects are smashed against
windshields obscuring the vision of motorists. Cars may overheat when radiators be-
come clogged and the smashed insects damage car paint if the body fluids are not re-
moved soon after contact (Callahan & Denmark 1973). The insect was first described
by Hardy (1940) from Galveston, Texas, who reported it to be widely spread, but more
common in Texas and Louisiana than other Gulf Coast states. It has now progressed

Florida Entomologist 81(4)

December, 1998

to all states bordering on the Gulf of Mexico, as well as Georgia, South Carolina, and
parts of Central America. It was first collected in Florida in 1949 and today is found
throughout Florida (Denmark & Mead 1992).
Several studies have been conducted on adult attractants and adult sampling for
lovebugs. Callahan & Denmark (1973) observed large numbers of lovebug adults con-
gregating at intersections, traffic lights, and filling stations. Their data showed that
lovebugs were attracted to automobile exhaust fumes irradiated with 3600 A UV
light. Whitesell (1974) observed adults flying to heat sources such as recently parked,
warm cars and engines. His data showed that greater numbers of lovebug adults were
caught on a heated box than on sound, exhaust, or control boxes. A mobile trap
mounted on top of a car has been used to measure population density of adult love-
bugs on highways (Sharpe 1974). In field tests, visual observations were used by Lep-
pla et al. (1974) to measure rhythmic activity of adult lovebugs. Thornhill (1976)
marked adult lovebugs with an ultraviolet dust and used large, white sticky traps to
recapture the adults in order to measure the dispersal of the adults. Callahan et al.
(1985) postulated that lovebugs are attracted to highways by automobile exhaust
fumes. They tested irradiated automobile exhaust fumes and their components as at-
tractants for adult lovebugs. Of the five different aldehydes tested, formaldehyde and
heptaldehyde were the most attractive. In this report, I provide data on the attraction
of adult P. nearctica to anethole and the use of anethole baited sticky traps to sample
adult populations of P. nearctica.
Cherry et al. (1996) reported on the attraction of adult beetles ofAnomala margi-
nata (Robinson) to anethole in Japanese beetle traps. During the course of that study,
I observed adult lovebugs hovering in large numbers around Japanese beetle traps
baited with anethole. However, since Japanese beetle traps are designed to capture
heavy-bodied insects such as beetles, I decided to see if lovebugs would be attracted
to anethole in sticky traps which are more suitable for catching smaller insects. Yel-
low sticky traps (Pherocon AM, no bait) made by Trece, Inc. Salinas, California were
used in these tests. Ten pairs of traps (anethole versus control) were set-up at ten dif-
ferent locations on the Everglades Research and Education Center at Belle Glade,
Florida. Traps at each location were 10 m apart and hung one m above the ground on
metal rods. A sponge (3 by 3 by 3 cm) was wedged into each trap. Control traps were
unbaited and each anethole trap had 10 ml of anethole poured into the sponge. The
anethole was obtained from Acros Organics, Fairlawn, New Jersey and was greater
than or equal to 99 percent. Tests were conducted when large numbers of adults were
observed flying at the research center. Six tests were conducted during April-May,
1996 and 1997 (see Table 1). Traps were exposed for 24 h in each test and then covered
with clear cellophane and taken to a laboratory. Lovebug adults on each trap were
counted under microscopic examination. The sex ratio of adults on the traps was de-
termined by scraping 100 adults from different control traps and 100 adults from the
anethole baited traps. These adults were placed in gasoline to dissolve the adhesive
from the trap and then sexed using characters described by Denmark & Mead (1992).
Statistical differences in adult numbers of control versus anethole baited traps in
each of the six tests were determined using paired t-tests (SAS 1996). A two by two
contingency table using Chi-square analysis (Dixon & Massey 1969) was used to de-
termine if the adult sex ratio was significantly different in control traps versus anet-
hole baited traps.
Data in Table 1 show that significantly more adult lovebugs were caught on sticky
traps baited with anethole than unbaited control traps in all six tests. By far, the most
lovebugs caught on any date occurred in both control and anethole traps on May 1,
1997. Reasons for the large catches during that test are not known for sure. However,

Scientific Notes


Control" Anetholeb

Date" Mean SD Range Mean" SD Range

May 16, 1996 3.0 3.5 0-11 186.5 99.0 48-291
May 20, 1996 2.2 2.7 0-7 75.3 50.5 7-153
May 29, 1996 7.5 4.6 0-14 88.7 46.1 14-141
April 15, 1997 11.4 7.5 2-22 113.4 55.6 24-225
April 17, 1997 5.6 2.3 3-11 101.4 34.5 51-151
May 1, 1997 258.8 85.5 123-368 887.8 199.2 648-1240

"Date of start of test. Traps exposed for 24 h.
bAdults per trap. Paired t-test (SAS 1996) showed significantly (P<0.01) more adults were caught on anethole
baited traps than controls during all six testing dates.

field observation indicated large numbers of adults were flying that day probably due
to large populations and calm winds which did not hinder flight. The sex ratio was
49:51 (M:F) in the control traps and 46:54 (M:F:) in the anethole baited traps. Chi-
square analysis showed that there was no significant difference (Chi-square = 0.3, 1
d.f., P > 0.05) in the sex ratio of adults in control versus baited traps. Anethole is an
essential oil found in plants (Morrison & Boyd 1973) and adult lovebugs are known to
feed on different plants (Hetrick 1970). Previous studies have shown anethole to be
attractive to diverse insects such as bees (Ladd & Tew 1983), scarabs (Cherry et al.
1996), and wireworms (Lehman 1932).
Florida Agricultural Experiment Station Journal Series Number R-06302.


Significantly more adult lovebugs were caught on sticky traps baited with anethole
than unbaited control traps in six tests. These data show that sticky traps baited with
anethole can be used as a simple and efficient sampling tool for adult P. nearctica.


CALLAHAN, P. S., AND H. A. DENMARK. 1973. Attraction of the "lovebug" Plecia nearc-
tica (Diptera: Bibionidae) to UV irradiated exhaust fumes. Florida Entomol.
56: 113-119.
CALLAHAN, P. S., T. C. CARLYSE, AND H. A. DENMARK. 1985. Mechanism of attraction
of the lovebug, Plecia nearctica to southern highways: further evidence for the
IR-dielectic waveguide theory of insect olfaction. Applied Optics 24: 1088-1093.
CHERRY, R. H., M. G. KLEIN, AND W. S. LEAL. 1996. Attraction of adult Anomala
marginata (Coleoptera: Scarabaeidae) to anethole. J. Agric. Entomol. 13: 359-364.
DENMARK, H. A., AND F. W. MEAD. 1992. Lovebug, Plecia nearctica Hardy (Diptera:
Bibionidae). Fla. Dept. Agric. Consumer Serv. Entomol. Circ. No. 350.
DIXON, W. J., AND F. J. MASSEY. 1969. Introduction to statistical analysis. McGraw-
Hill, New York.
HARDY, D. E. 1940. Studies in New World Plecia (Bibionidae: Diptera). Part 1. Kansas
Entomol. Soc. 13 (1): 15-27.

562 Florida Entomologist 81(4) December, 1998

HETRICK, L. A. 1970. Biology of the "love-bug", Plecia nearctica (Diptera: Bibionidae)
Florida Entomol. 53: 23-26.
LADD, T. L., AND J. E. TEW. 1983. Attraction of honey bees (Hymenoptera: Apidae) to
traps baited with lures for Japanese beetles (Coleoptera: Scarabaeidae)
J. Econ. Entomol. 76: 769-770.
LEHMAN, R.S. 1932. Experiments to determine the attractiveness of various aromatic
compounds to adults of the wireworms. J. Econ. Entomol. 25: 949-958.
Rhythmic activity of Plecia nearctica. Environ. Entomol. 3: 323-326.
MORRISON, R. T., AND R. N. BOYD. 1973. Organic chemistry, Allyn and Bacon Inc.
SAS INSTITUTE. 1996. SAS Systems for Windows. Version 6.12 SAS Institute, Cary, N.C.
SHARPE, H. 1974. Love bugs are drawn to heat. Sunshine State Agric. Res. Rep. Vol.
19, No. 3-4. Pages 3-4.
THORNHILL, R. 1976. Dispersal of Plecia nearctica (Diptera: Bibionidae). Florida En-
tomol. 59: 45-53.
WHITESELL, J. J. 1974. Heat, sound, and engine exhaust as "lovebug" attractants
(Diptera: Bibionidae: Plecia nearctica) Environ. Entomol. 3: 1038-1039.


Florida Entomologist 81(4)

December, 1998


Department of Entomology, Clemson University, Clemson, SC 29634

1Current address: Bureau of Water, South Carolina Dept. of Health and
Environmental Control, 2600 Bull St., Columbia, SC 29201

The larvae of antlions (Neuroptera: Myrmeleontidae) are renowned for their pred-
atory tactic: the construction of funnel-shaped pitfall traps in sandy substrate, be-
neath which they wait for prey. Pit-building behavior, however, is limited to the tribe
Myrmeleontini (New 1986) and is characteristic of the genus Myrmeleon (Lucas &
Stange 1981). The lie-in-wait predation strategy suggests that various prey will be en-
countered by the antlion larva. Plasticity of predatory behavior should increase the ef-
ficiency by which an opportunistic predator subdues and processes different types of
prey. Therefore, I asked the question: does the behavioral response of a pit-building
antlion, Myrmeleon mobilis Hagen, differ among prey types? In this study I charac-
terize the predatory behaviors ofM. mobilis and compare the sequence and frequency
of these behaviors in response to three prey types.
Thirty late first- and second-instar M. mobilis larvae were collected from shel-
tered, sandy areas in Clemson, Pickens County, South Carolina, on 8 October, 1995.
Larvae were placed individually in containers with 3 cm of sterilized sand, and held
at 25 + 1C, 65 + 5%RH, and a photoperiod of 12:12 (L:D). Each larva was allowed to
construct a pit and then fed a maintenance diet of earwigs, Euborellia annulipes Lu-

Scientific Notes

cas (Carcinophoridae); rearing continued for 12 days, until all individuals had
reached late second instar. The three experimental prey species (length + SD/max.
width + SD) were the termite Reticulitermes flavipes Kollar (Rhinotermitidae) [5.7
+ 0.85 x 1.2 + 0.08 mm], the ant Prenolepis imparis Say (Formicidae) [4.22 + 0.20 x 1.5
+ 0.17 mm], and the beetle Alphitobius diaperinus Panzer (Tenebrionidae) [6.13
- 0.82 x 2.72 0.13 mm].
Behavioral trials were conducted at 23-25C. Each M. mobilis larva was presented
with one individual of a randomly selected prey species. Prey was dropped into the
center of the pit, to standardize introduction (Griffiths 1980), and the resulting inter-
action was videotaped at a distance of ca. 7 cm. Recording began with prey introduc-
tion and ended when the prey either escaped, or was consumed, and the larva
returned to the pre-introduction 'ready position' (jaw set). Ten trials of each prey spe-
cies were recorded and no larva was used in more than one trial. Descriptions of pred-
atory behaviors were based on videotaped trials and direct observation. Each trial
was reviewed, and sequence and frequency of behaviors noted. Significant behavioral
transitions (p = 0.05) were identified using a first order, preceding-following, behav-
ioral transition matrix (after Willey et al. 1992). Flow diagrams of significant transi-
tions were constructed.
The following 12 discrete predatory behaviors were identified in the behavioral
catalog of Myrmeleon mobilis:

1. Attack.

The head is moved rapidly forward while closing the mandibles, and is often
flicked rapidly back, expelling sand from the pit.

2. Holding.

The prey is gripped securely in the mandibles.

3. Submergence.

Holding prey, the larva moves down and back into the substrate until the entire
larva and at least part of the prey are not visible.

4. Emergence.

Holding prey, the larva moves up and forward until the entire prey and at least
part of the larva's head/mandibles is visible.

5. Prey Beating.

Holding prey, the larva rapidly flicks its head up and down (4-5 beats per bout)
(Fig. 1.), often drumming the prey on the substrate.

6. Feeding.

While at least one mandible tip is inserted, fluids are extracted from the prey, often
alternating with mandibular probing and manipulation of the prey.

7. Pit Clearing.

The head is moved laterally, accumulating sediment on the dorsal surface, then
flicked rapidly back, expelling sediment.

Florida Entomologist 81(4)

December, 1998

Fig. 1. Prey-beating behavior exhibited by Myrmeleon mobilis with beetle prey,Al-
phitobius diaperinus.

8. Head Roll.

The head is raised and swept in a circular motion along the pit wall, accumulating
sediment in the pit center.

9. Prey Clearing.

The mandibles are used to position prey on the dorsal head surface, then the head
is flicked rapidly back, expelling prey.

10. Grooming.

The tip of one mandible is moved along the groove on the inside edge of the oppos-
ing mandible.

11. Quiescence.

Larva remains motionless, without prey, for 7+ seconds.

12. Jaw Set.

The larva pulls beneath the sand, while fully opening the mandibles. The eyes, an-
tennae and mandible tips remain visible.
Sequences for all prey types typically followed a core pattern of behaviors (Fig. 2),
starting with attack and holding, followed by submergence, emergence, and feeding.
After feeding ended, maintenance behavior generally occurred (prey clearing, pit
clearing, head roll, and grooming) and, finally, jaw set. The major difference in behav-
ioral sequence was prey beating behavior: 90% of the beetle prey-trials resulted in

Scientific Notes




JA W SET --- 0.10
Fig. 2. Flow diagram of predatory behavior for M. mobilis, showing sequence of sig-
nificant behavioral transitions (p = 0.05) and transition frequency (n = 10 trials) when
presented with ant prey, Prenolepis imparis.

Florida Entomologist 81(4)

December, 1998

prey beating, compared to 20% of the ant trials, and 10% of the termite trials. The
mean frequency of prey-beating bouts for the beetle (42.40 + 12.59SE) was signifi-
cantly different (p < 0.005) from that for both the termite (2.00 + 2.00SE) and the ant
(8.90 + 7.57SE); the latter two were not significantly different (Tukey's, p > 0.05).
My field observation in areas ofM. mobilis habitation revealed that taxa including
Hymenoptera, Coleoptera, Orthoptera, and non-insect arthropods are consumed by
antlion larvae. In the laboratory, M. mobilis larvae ate both soft and hard-bodied prey.
However, trials with highly sclerotized 1. . I ..-.I-. I differed significantly from those
with softer prey (ants and termites) in sequence and frequency of prey-beating behav-
ior, demonstrating that the predatory response of M. mobilis varies with prey type.
Prey-beating behavior may be an adaptation to facilitate mandibular penetration (in
beetles, this usually occurred in a coxal joint or between tagma), or to disorient and
subdue vigorously struggling prey. Griffiths (1980) described behavior similar to prey
beating for 'difficult' prey in the feeding biology ofMorter obscurus Rambur, and noted
that treatment of hard and soft-bodied ant prey varied with respect to mandibular in-
sertion. In addition, previous research suggests that phylogeny may be reflected by
behavior (Mansell 1988, Matsura & Murao 1994). An interspecies comparison of pred-
atory behavior in Myrmeleontidae may prove worthwhile in relating behavioral dif-
ferences to phylogeny.
Many thanks to P. H. Adler for his help with numerous aspects of this project, L.
A. Stange, for reviewing the manuscript and confirming antlion identifications, and to
J. McCreadie, K. van den Meiracker, and C. Buchanan-Beane. This study was sup-
ported by Clemson University and an E. W. King Endowed Memorial Grant.


The predatory behavior of a pit-making antlion, Myrmeleon mobilis, is character-
ized. Behavioral sequences among three prey types were similar, when compared via
flow diagrams. A significant difference in behavioral frequency existed between hard-
bodied and soft-bodied prey types.


GRIFFITHS, D. 1980. The feeding biology of ant-lion larvae: prey capture, handling and
utilization. J. Anim. Ecol. 49: 99-125.
LUCAS, J. R., AND L. A. STANGE. 1981. Key and descriptions to the Myrmeleon larvae
of Florida (Neuroptera: Myrmeleontidae). Florida Ent. 64: 207-216.
MANSELL, M. W. 1988. The pitfall trap of the Australian ant-lion Callistoleon illustris
(Gerstaecker) (Neuroptera: Myrmeleontidae): an evolutionary advance. Aus-
tralian J. Zool. 36: 351-356.
MATSURA, T. A., AND T. MURAO. 1994. Comparative study on the behavioral response
to starvation in three species of antlion larvae (Neuroptera: Myrmeleontidae).
J. Insect Behav. 7: 873-884.
NEW, T. R. 1986. A review of the biology of Neuroptera Planipennia. Neuroptera In-
ternational, Suppl. 1: 1-57.
WILLEY, M. B., JOHNSON, M. A., AND P. H. ADLER. 1992. Predatory behavior of the ba-
silica spider, Mecynogea lemniscata (Araneae, Araneidae). Psyche 99: 153-167.

Scientific Notes


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

'USDA-APHIS, Gainesville Plant Protection Station, P. O. Box 147100,
Gainesville, FL 32614-7100

2Instituto de Ecologia, A. C. 91000 Xalapa, Veracruz, Mexico

3USDA-APHIS, Methods Development, 12 Calle 6-96 Zona 10, Guatemala City,

4MOSCAMED, Guatemala City, Guatemala

Augmented releases of tephritid parasitoids have suppressed populations of both
the Mediterranean fruit fly (Ceratitis capitata [Weidemann]) and the Caribbean Fruit
fly, (Anastrepha suspense [Loew]) (Wong et al. 1991, Sivinski et al. 1996). Typically,
braconid parasitoids of larvae, such as Diachasmimorpha longicaudata (Ashmead),
are employed. However, parasitoids that attack fruit fly pupae might be useful addi-
tions to such programs since they are able to attack flies that might otherwise escape
parasitism. Flies developing within large fruits are less likely to be parasitized by
braconids; parasitoids are less able to reach them with their ovipositors (e.g., Sivinski
1991, Sivinski et al. 1997). Since tephritid larvae typically leave fruits to pupate in
the soil, fruit size is less important to pupal parasitoids foraging for hosts.
The diapriid Coptera haywardi (Ogloblin) is a widespread native of Latin America,
where it has been collected from the pupae of several Anastrepha species (Loiacono
1981). It appears to attack only species of Tephritidae (Sivinski et al. 1998). Unlike
many common, ectoparasitic, pteromalid pupal parasitoids of cyclorrhaphous
Diptera, C. haywardi develops as an endoparasitoid (Sivinski et al. 1998). This more
intimate relationship with its host may result in greater specialization and a nar-
rower host range (see Godfray 1994). Specialized parasitoids are particularly valu-
able in augmentative releases since they are less likely to harm beneficial insects and
more likely to focus their foraging on declining numbers of target pests.
There is a possibility of adding pupal parasitoids to existing braconid mass-rear-
ing programs. For example, fruit fly larvae could be exposed to a braconid parasitoid,
and after host pupation the pupae could then be placed with pupal parasitoids. Un-
parasitized pupae would typically be available for the second parasitoid since only a
few braconid rearing operations consistently reach parasitism levels of 50%, and
some, such as early efforts with raising Diachasmimorpha tryoni (Cameron), average
as low as 20% (pers. observ. of the authors). This scheme would be most effective with
a pupal parasitoid that, 1) would not hyperparasitize the primary braconid parasi-
toid, and 2) was able to develop in flies whose maturation was disrupted by radiation.
Irradiation of larvae prior to parasitization is used in mass-rearing programs in Flor-
ida, Mexico, and Guatemala to prevent mixed lots of parasitoids and fertile flies (Siv-

Florida Entomologist 81(4)

December, 1998

inski and Smittle 1990; Sivinski personal observation). Previous studies have found
no indication of C. haywardi hyperparasitism ofD. longicaudata developing inA. sus-
pensa (Sivinski et al. 1998).
In order to determine if C. haywardi would develop in pupae that had been formed
by irradiated larvae, we provided the diapriid with irradiated and unirradiated pupae
of A. suspense in the following manner. Mixed lots of unparasitized pupae and pupae
parasitized by D. longicaudata were obtained from the Florida Division of Plant In-
dustry, Gainesville, Florida (see Sivinski et al. 1996). At the start of the experiment,
A. suspense had been in colony for ~ 9 years (150 generations). D. longicaudata had
been in colony ~ 6 years and C. haywardi had been colonized on A. suspense for ~ 1
year. These lots had been previously derived from late 3"d instar larvae that had been
either irradiated in a Cesium 137 source (Nordion International Inc., Model M; Ka-
nata, Ontario, Canada) at 6 kR or left unirradiated. Depending on availability, either
10 ml (~400 pupae) or 3 ml (-120 pupae) of 1-day old pupae were placed with 15 un-
sexed individuals of C. haywardi in 250 ml cardboard cups containing moist vermic-
ulite, honey, and water. The cups were covered with a fine-mesh cloth and left at 26 (-
1)C and ambient humidity for 1 week. At the end of this period pupae and adult par-
asitoids were separated and the pupae held at 290C and 70% humidity for 1 month.
At this point the adult insects that had emerged were identified and counted. Une-
merged pupae were dissected to determine their contents. There were 8 replicates of
3 ml and 7 replicates of 10 ml cups of pupae. Each replicate consisted of 5 cups of ir-
radiated and 5 cups of unirradiated pupae, so that a total of 470 ml of irradiated and
470 ml of unirradiated pupae were exposed to a total of 1125 diapriids each. Since the
results of the experiment were consistent and unambiguous, the data from cups con-
taining different amounts of pupae were pooled.
There was not a single successful development of C. haywardi in pupae formed by
irradiated A. suspense larvae. In the unirradiated lots, a total of 9772 A. suspense
closed. In the irradiated lots, 1A. suspense and 0 C. haywardi closed. Parasitism by
D. longicaudata was 31% (SE = 7%) in unirradiated lots and 29% (SE = 7%) in irradi-
ated lots. Parasitism by C. haywardi of unirradiated pupae averaged 12% (SE = 0.3%).
A similar experiment, examining the development of C. haywardi in irradiated pu-
pae of C. capitata was conducted in the "Aurora" USDA-APHIS/MOSCAMED facility
in Guatemala City, Guatemala. At the time of the experiment, C. haywardi had been
reared for ~5 generations on C. capitata pupae. Coptera haywardi was presented with
pupae of C. capitata formed from larvae either irradiated at 14.5 kR with a Cobalt 60
source at the MOSCAMED rearing facility at El Pino, Guatemala or left unirradiated.
Lots of 1180 pupae, either irradiated or unirradiated, were exposed for a period of
three days to ~ 800 unsexed individuals of C. haywardi housed in a 1 m by 1 m plexi-
glass cage. Pupae then were removed and held at 26oC and 60-70% humidity for one
month. There were six replicates, so that a total of 7080 irradiated and 7080 unirra-
diated pupae were exposed to parasitism.
As in the case of irradiated A. suspense, there was no emergence of adult C. hay-
wardi from irradiated C. capitata. Unirradiated pupae yielded 231 C. haywardi (~ 4%
Thus, C. haywardi lacks the useful attribute of being able to exploit irradiated
fruit flies in those mass-rearing programs which expose larvae to high levels of radi-
ation. This does not preclude its use in other types of mass-rearing programs. For ex-
ample, in a parasitoid mass-rearing system previously used in Hawaii there was a
sufficient difference in the developmental periods of the braconid Diachasmimorpha
tryoni (Cameron) and its host C. capitata to allow separation of the adults of the two
species without the use of radiation. (Wong and Ramadan 1992). In this instance, the

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